LEBANON Lebanon
ECONOMIC Sinking
MONITOR
(To the Top 3)
Spring 2021
Middle East and North Africa Region
Lebanon
Economic Monitor
Lebanon Sinking (To the Top 3)
)لبنان يغرق (نحو أسوأ ثالث أزمات عاملية
Le Naufrage du Liban (Top 3 des pires crises mondiales)
Spring 2021
Global Practice for Macroeconomics, Trade & Investment
Middle East and North Africa Region
LEBANON ECONOMIC
MONITOR Document of the World Bank
The Delibera
Depressi
TABLE OF CONTENTS
Acronyms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
الموجز . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv
Résumé . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix
1. The Policy Context . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2. Recent Macro-Financial Developments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Output and Demand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3
Fiscal Developments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
The External Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Money and Banking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11
3. Global Crises Comparators: Looking for the Minimum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Per Capita Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Depreciation-Inflation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .22
Fiscal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23
Debt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
External Position . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24
Overall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4. Outlook and Risks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27
The FX Subsidy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .28
Large Scale Interruptions to Vital Public Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .28
Special Focus I: FX Subsidy Reform in the Deliberate Depression . . . . . . . . . . . . . . . . . . . . . . . .31
Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31
Macroeconomic Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Design a Broad-Coverage Subsidy Reform Compensation Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .36
iii
Implementing Broad-Coverage Subsidy Reform Compensation Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . .38
Complementarities between the BC-CT Program and other SSN Programs . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Special Focus II: Public Service Delivery in the Deliberate Depression . . . . . . . . . . . . . . . . . . . . 41
The Electricity Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .41
Water and Sanitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .43
Annex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Forecasts of Lebanon’s Real GDP Growth Using MIDAS Regressions: An Update for 2020 and 2021 47
Dynamic Response of Inflation to Currency in Circulation in Lebanon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Unit Root Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
VAR Variables in Difference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .52
VAR Variables in (Log) Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
The Cointegrated VAR: Vector Error Correction Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Global Financial Crises Episodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .58
The Argentinian Banking Crisis of 1980 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
The Philippines Financial Crisis of 1981 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
The Mexican Debt Crisis of 1981–82 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
The Chilean Banking Crisis of 1981 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .61
The Venezuelan Banking Crisis of 1994 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
The Argentinian Financial Crisis of 2001–02 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
The Uruguayan Banking Crisis of 2002 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
The Greek Financial Crisis 2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
List of Figures
Figure 1 While the Contraction in Real GDP Commenced in 2018, it Accelerated Sharply in 2020 . . . . . . .4
Figure 2 Net Exports are Expected to Be the Sole Positive Contributor to Real GDP . . . . . . . . . . . . . . . . . . .4
Figure 3 Firm Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5
Figure 4 Large Shortfalls in Revenues Will Induce a Significant Deterioration in the Fiscal Position . . . . . 10
Figure 5 Valuation Effects from Exchange Rate Depreciations Will Pressure the Debt-to-GDP Ratio . . . . .10
Figure 6 A Steady Depletion in the Gross Foreign Exchange Position at BdL. . . . . . . . . . . . . . . . . . . . . . . . . 11
Figure 7 Ratios of C1, C2b, Luxury and Other Imports Were Stable Until the Period
Leading to the Crisis… . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Figure 8 …when Ratios of C1 and C2b Imports Rose at the Expense of those for Luxury
and Other Goods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Figure 9 A Sharp Depreciation in the Exchange Rate along with Surging Inflation and Narrow Money 13
Figure 10 Inflation in Basic Items is a Key Driver of Overall Inflation, Hurting the Poor and
the Middle Class . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Figure 11 Food Expenditure Shares by Deciles for Lebanese Nationals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13
Figure 12 Heavy Deleveraging of Assets (Private Loans) and Liabilities (Private Deposits)
in Financial Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Figure 13 A Steady and Sharp Deterioration in Credit Performance as Measured
by NPL Ratio for Banks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
iv LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
Figure 14 Lebanon’s Real GDP is a More Accurate Reference Point for the Start of
the Financial Crisis than Real GDP/Capita . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21
Figure 15 Real GDP Per Capita for G8 Plus Lebanon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .22
Figure 16 Real GDP Per Capita for G8 Plus Lebanon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .22
Figure 17 CPI Indices for G8 Plus Lebanon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Figure 18 Overall Fiscal Balance for G8 Plus Lebanon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Figure 19 CPI Indices for G8 Plus Lebanon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Figure 20 Overall Fiscal Balance for G8 Plus Lebanon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Figure 21 Gross Debt for G8 Plus Lebanon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Figure 22 Current Account Balance for G8 Plus Lebanon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25
Figure 23 A Steady Depletion in the Gross Foreign Exchange Position at BdL . . . . . . . . . . . . . . . . . . . . . . . . 33
Figure 24 Upon Removal of FX Subsidy, Direct Effects on Inflation Are Substantial . . . . . . . . . . . . . . . . . . . . 35
Figure 25 Proposed Monthly Benefit Amount (US$), and Coverage (%), by Year . . . . . . . . . . . . . . . . . . . . . . 37
Figure 26 Compensation Program Outlay and Net Savings from Phase I Subsidy Withdrawal, US$ mn 37
Figure 27 Evolution of High Frequency Indicators Used to Nowcast and Forecast Lebanon’s
Real GDP Growth in 2020 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Figure 28 Growth of High Frequency Real Economy Indicators Used to Nowcast and
Forecast Lebanon’s Real GDP Growth in 2021 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .50
Figure 29 Time Series Dynamics of the Variables in Changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Figure 30 Responses to a 1 Percent Increase in the Growth of Currency in Circulation . . . . . . . . . . . . . . . . 54
Figure 31 Time Series Dynamics of the Variables in Log Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .55
Figure 32 Responses to a 1 Percent Increase in the (Log of) Currency in Circulation . . . . . . . . . . . . . . . . . . 56
Figure 33 Responses to a 1 Percent Increase in (the Log of) Currency in Circulation from a VECM . . . . . .57
List of Tables
Table 1 Average Change in Full-Time Employees by Gender for All Firms Surveyed. . . . . . . . . . . . . . . . . . .6
Table 2 Summary of Fiscal Accounts Showing Actual Numbers, WB Estimates and
Government Budgets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Table 3 Crisis Severity: Percent Decline in Per Capita GDP, Duration of Contraction,
and Years to Full Recovery in 25 of the Worst Systemic Banking Crises, 1857–2013 . . . . . . . . . 20
Table 4 Selected Macroeconomic Indicators for Lebanon; 2016–2021 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Table 5 Cost of FX Import Subsidy and Impact of its Removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Table 6 Real GDP Growth Forecasts for 2020 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Table 7 Candidate Predictor Variables for the Real High Frequency Indicators . . . . . . . . . . . . . . . . . . . . . .49
Table 8 Forecasts of Real GDP Growth for 2021 Using Real Activity Indicators . . . . . . . . . . . . . . . . . . . . . .49
Table 9 Forecasts of Real GDP Growth for 2021 Using Real Activity and Financial Indicators . . . . . . . . . 51
Table 10 Unit Root Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Table 11 Cumulative Effect of a 1 Percent Increase in Currency in Circulation on Inflation . . . . . . . . . . . . . 52
Table 12 Cumulative Effect of a 1 Percent Increase in Currency in Circulation on Inflation . . . . . . . . . . . . . 54
Table 13 The Johansen (1988) Trace Statistic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .56
Table 14 Cumulative Effect of a 1 Percent Increase in Currency in Circulation on Inflation . . . . . . . . . . . . . 57
List of Boxes
Box 1 The Impact of Multiple Crises on Formal Firms and the Labor Market in Lebanon . . . . . . . . . . . . . 5
Box 2 Lebanon’ 2021 Budget Draft Proposal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Box 3 Impact of Crises on Poverty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13
Box 4 The Most Recent Financial Operation by BdL with Commercial Banks . . . . . . . . . . . . . . . . . . . . . .14
TABLE OF CONTENTS v
ACRONYMS
AER Average Exchange Rate LBP Lebanese Pound
BdL Banque du Liban LEM Lebanon Economic Monitor
BoP Balance of Payments MIDAS Mixed-Data Sampling
CA Current Account MoET Ministry of Economy and Trade
CBP Central Bank of the Philippines MTO(s) Money Transfer Operator(s)
CBV Central Bank of Venezuela NGO(s) Non-governmental Organizations
CD(s) Certificate of Deposit(s) NID National ID
CFF Compensatory Financing Facility NIR Net International Reserve
CPI Consumer Price Index NPL(s) Non-performing Loan(s)
CSI Crisis Severity Index NPTP National Poverty Targeting Program
DBP Development Bank of the Philippines NSSF National Social Security Fund
ECB European Central Bank O&M Operation and Maintenance
EdL Electricité du Liban PEP(s) Politically Exposed Person(s)
EFF Extended Fund Facility PISA Programme for International Student
EFF Extended Fund Facility Assessment
EMU European Monetary Union PMI Purchasing Manager’s Index
ES Enterprise Survey PNB Philippines National Bank
ESSN Emergency Crisis and COVID-19 PoB Port of Beirut
Response Social Safety Net Project POS Point-of-Sale
FCV Fragility, Conflict, and Violence Pp Percentage Points
FI(s) Financial Institution(s) PSI Private Sector Initiative
FSBS Fund for the Stability of the Banking RDNA Rapid Damage and Needs Assessment
System SBA Stand-by Arrangement
FX Foreign Exchange SDC(s) Social Development Center(s)
GDP Gross Domestic Product SSN Social Safety Net
GFSM International Government Accounting TB(s) Treasury Bond(s)
Standards TD(s) Time Deposit(s)
GNI Gross National Income (per capita) UN United Nations
GNP Gross National Product US$ United States Dollar
GOL Government of Lebanon WE(s) Water Establishment(s)
GRM Grievance Redress Mechanism WSS Water Supply and Sanitation
IADB Inter-American Development Bank xM 2020 First x months of 2020
IMF International Monetary Fund yoy Year over Year
vii
PREFACE
T
he Lebanon Economic Monitor provides (Senior Economist). Special Focus II: Public Service
an update on key economic developments Delivery in the Deliberate Depression, has been led by
and policies over the past six months. It also Wissam Harake (Senior Economist), Sameh Mobarak
presents findings from recent World Bank work on (Senior Energy Specialist), Amal Talbi (Lead Water
Lebanon. The Monitor places them in a longer-term Resources Management Specialist), Sally Zgheib
and global context and assesses the implications of (Senior Water Supply and Sanitation Specialist), and
these developments and other changes in policy on Nathalie Lahire (Senior Economist). The Lebanon
the outlook for Lebanon. Its coverage ranges from Economic Monitor has been completed under the
the macro-economy to financial markets to indicators guidance of Christos Kostopoulos (Lead Economist),
of human welfare and development. It is intended for Eric Le Borgne (Practice Manager) and Saroj Jha
a wide audience, including policy makers, business (Country Director). Zeina Khalil (Communications
leaders, financial market participants, and the Officer) is the lead on communications, outreach and
community of analysts and professionals engaged in publishing.
Lebanon. The findings, interpretations, and conclusions
The Lebanon Economic Monitor is a product expressed in this Monitor are those of World Bank
of the World Bank’s Lebanon Macroeconomics, staff and do not necessarily reflect the views of the
Trade and Investment (MTI) team. It was prepared Executive Board of The World Bank or the govern-
by Wissam Harake (Senior Economist), Ibrahim ments they represent.
Jamali (Consultant) and Naji Abou Hamde For information about the World Bank and its
(Economic Analyst) with contributions from Lars activities in Lebanon, including e-copies of this publi-
Jessen (Lead Debt Specialist), Haocong Ren (Senior cation, please visit www.worldbank.org.lb
Financial Sector Economist), Zeina El Khoury To be included on an email distribution list
(Private Sector Specialist), Angela Elzir Assy (Labor for this Lebanon Economic Monitor series and
Market Specialist), Ganesh Kumar Seshan (Senior related publications, please contact Alain Barakat
Economist), Bilal Malaeb (Economist), Fahmina (abarakat@worldbank.org). For questions and com-
Rahman Dutta (Social Protection Specialist) and ments on the content of this publication, please
Haneen Ismail Sayed (Lead Operations Officer). contact Wissam Harake (wharake@worldbank.org)
Special Focus I: FX Subsidy Reform in the Deliberate or Christos Kostopoulos (ckostopoulos@worldbank.
Depression, has been led by Haneen Ismail Sayed org). Questions from the media can be addressed to
(Lead Operations Officer), and Wissam Harake Zeina Khalil (zelkhalil@worldbank.org).
ix
EXECUTIVE SUMMARY
T
he Lebanon financial and economic crisis is less due to knowledge gaps and quality advice
is likely to rank in the top 10, possibly and more the result of a combination of (i) a lack of
top three, most severe crises episodes political consensus over effective policy initiatives;
globally since the mid-nineteenth century. and (ii) political consensus in defense of a bankrupt
This is a conclusion of the Spring 2021 Lebanon economic system, which benefited a few for so long.
Economic Monitor (LEM) in which the Lebanon crisis In the face of these challenges, Lebanon lacks a fully-
is contrasted with the most severe global crises functioning executive authority and is currently in the
episodes as observed by Reinhart and Rogoff (2014)1 process of forming its third Government in a little over
over the 1857–2013 period. In fact, Lebanon’s GDP a year. This debilitating institutional void has lasted
plummeted from close to US$ 55 billion in 2018 to over 8 months so far.
an estimated US$ 33 billion in 2020, with US$ GDP/ The social impact of the crisis, which is
capita falling by around 40 percent. Such a brutal already dire, could rapidly become catastrophic;
and rapid contraction is usually associated with more than half the population is likely below
conflicts or wars. Even prior, the World Bank has long the national poverty line. Those paid in Lebanese
identified Lebanon as a Fragility, Conflict & Violence Lira—the bulk of the labor force—are seeing potent pur-
(FCV) State, and as such, the dire socio-economic chasing power declines. Phone surveys conducted in
conditions risk systemic national failings with regional the end of 2020 by the World Food Program found
and potentially global consequences.2 This illustrates that 41 percent of households reported challenges
the magnitude of the economic depression that the
country is enduring, with sadly no clear turning point
on the horizon, given the disastrous deliberate policy 1
Reinhart, Carmen M. and Kenneth S. Rogoff (2014),
inaction. Recovery from Financial Crises: Evidence from 100
In the Fall 2020 LEM, Lebanon’s economic Episodes, American Economic Review: Papers &
crisis was termed The Deliberate Depression. Proceedings 2014, 104(5): 50–55.
For over a year, Lebanese authorities countered an
2
In Amin’s Maalouf’s Le Naufrage des Civilisations,
translated as The Wreckage of Civilizations, the
assailment of compounded crises—namely, the coun-
Lebanese author highlights how the failings of the
try’s largest peace-time economic and financial crisis, Levant are propagating into a failing of all civilizations.
COVID-19 and the Port of Beirut explosion—with delib- Maalouf, Amin (2019), Le Naufrage des Civilisations,
erately inadequate policy responses. The inadequacy French & European Publications, Inc., April 2, 2019.
xi
in accessing food and other basic needs. The share Real GDP growth is estimated to have con-
of households having difficulties in accessing health tracted by 20.3 percent in 2020, on the back of a
care rose from 25 percent (July-August) to 36 percent 6.7 percent contraction in 2019. In a large part due
(Nov-Dec). The unemployment rate also rose among to COVID-19, the tourism sector has been particularly
the respondents, from 28 percent in February (pre- hit; tourist arrivals fell by 71.5 percent, (yoy), over the
COVID) to nearly 40 percent in Nov-Dec. first five months of 2020 (5M-2020). Meanwhile, con-
Lebanon, with a history of civil war and con- struction permits and cement deliveries (proxies for
flicts, faces realistic threats to its already fragile the construction and real estate) suffered respective
social peace. As previously argued (World Bank, declines of 26.9 percent (yoy) and 44.7 percent (yoy)
20163), the key overarching constraints to develop- over the first 10 months of 2020 (10M-2020).
ment in Lebanon are i) elite capture hidden behind the An ostensible improvement in some fiscal
veil of confessionalism, and ii) conflicts and violence, indicators (as a percentage of GDP) masks an
with these two having a symbiotic relationship: they actual deterioration. Revenues are estimated to
feed and strive on each other. Demonstrations, while have declined sharply as a result of the severe eco-
more modest in numbers, have recently grown angrier, nomic contraction, with the ratio to GDP falling further
erupting in cities across Lebanon to protest against due to an inflation-driven increase in nominal GDP.
the dire economic conditions; vital routes are being However, this is more than offset by a larger decline
cut off causing significant disruptions to mobility and in current expenditures, which benefited from: lower
livelihood; increased crime rates threaten personal interest payments (due to the Eurobond default and
security; national fragmentation can allow infiltration a favorable arrangement with BdL on domestic debt);
of sinister groups with grave security implications. cuts in transfers; and also a denominator-led GDP
Hence, there is growing wariness of potential triggers effect. Hence, while the 2020 overall fiscal balance is
to social unrest. estimated to have improved by 0.7 percentage points
(pp) to reach –4.9 percent of GDP, the primary bal-
ance deteriorated by 2.3 pp to –2.8 percent of GDP.
Recent Economic Developments The sharp economic contraction implied
a commensurate drop in imports, and conse-
Monetary and financial turmoil are driving crisis quently, an anticipated narrowing of the current
conditions, more palpably through interactions account deficit. During 10M-2020, merchandize
between the exchange rate, narrow money and imports shrank by 45 percent, which drove a 54.8
inflation. Acute exchange market pressures in percent decrease in the trade-in-goods deficit. We
Lebanese markets are reflected by heavy fluctuations estimate that the current account deficit fell by 10 pps
in the US$ banknote exchange rate, which temporarily to reach 11 percent of GDP in 2020, compared to a
breached LBP 15,000/US$, before falling back down. medium-term (2013–2019) average of 22.5 percent of
This is within the context of a multiple exchange rate GDP. Nonetheless, the sudden stop in capital inflows
system, which includes the official exchange (LBP has implied a steady depletion of foreign exchange
1,507.5/US$) as well Banque du Liban’s (BdL) (FX) reserves at BdL, which exacerbates constraints
platform rate set at LBP 3,900/US$. Overall, the World on imports.
Bank Average Exchange Rate (AER) depreciated by The burden of the ongoing adjustment/
129 percent in 2020. Exchange rate pass through deleveraging in the financial sector is highly
effects on prices have resulted in surging inflation, regressive, concentrated on smaller depositors,
averaging 84.3 percent in 2020. Meanwhile, the stock the bulk of the labor force and smaller busi-
of currency in circulation increased by 197 percent, nesses. De facto lirafication and haircuts on dollar
even as broad money supply (which includes bank
deposits) declined, with the latter weighed down by 3
World Bank (2016) “Lebanon Systematic Country
deleveraging in the financial sector. Diagnostic”.
xii LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
deposits are significant despite BdL’s and banks’ public services: electricity, water supply, sanita-
official commitment to safeguarding deposits. The tion and education. The Deliberate Depression has
burden of the ongoing adjustment/deleveraging is further undermined already weak public services via
regressive and concentrated on the smaller deposi- two effects: (i) it has significantly increased poverty
tors, who lack other source of savings, the local labor rates expanding the demography that is not able to
force, that is paid in lira, and smaller businesses. The afford private substitutables (the way citizens had pre-
banking sector is advocating for mechanisms that viously adapted to abysmal quality of public services),
incorporate state owned assets, gold reserves, and and are thus more dependent on public services; and
public real estate in order to overhaul their impaired (ii) threatens financial viability and basic operability
balance sheets. This constitutes a bailout of the of the sector by raising its costs and lowering its
financial sector and is inconsistent with restructuring revenues. Specifically,
principles that protect taxpayers and depositors.
These principles include bail in solutions based on • Severe shortage of foreign currency threatens
a hierarchy of creditors, starting with banks share- termination of private sector contracts for
holders. Government can also apply a wealth tax (on power plant maintenance and temporary power
financial and real assets) as a tool to progressively generation. Meanwhile, Electricité du Liban
restructure the financial sector. (EdL) revenues, which are in Lebanese pound,
are shrinking because of increasing technical,
commercial and collection losses. EdL is likely to
Special Foci increase rolling blackouts to manage its cashflow
shortfalls.
The financial and economic crisis has intensified • In 2020, the Water Establishments (WEs)
Lebanon’s fragility and fragmentation, increasing witnessed serious depletions in supplies,
the risk to social and civil unrest. The crisis revenues, and financial and human resources,
has exacerbated long-term national deficiencies while affected by an upward spiral in costs. Due
including institutional weaknesses, failed economic to reduced water supply from the WEs in 2020,
and social policy, and dismal public service delivery. people have had to rely more on other costlier
In such an environment, there is growing weariness and less convenient water alternatives, such as
of triggers for social unrest. In this LEM, we highlight water tankers and bottled water, whose prices
two potential economic triggers that are under have surged.
increased scrutiny, and which can have significant • The breakdown in sanitation services risks
social implications. intensifying the spread of water-borne diseases,
The first Special Focus examines Lebanon’s adversely impacting an already vulnerable public
FX subsidy for critical and essential imports, health.
which offers a serious political and social chal- • Compounded crises have placed Lebanon’s
lenge. On the one hand, the current FX subsidy education sector under severe strain. The
is both distortionary, expensive and regressive. It increase in poverty rates is leading to an exodus
exerts considerable stress on Lebanon’s balance of of students from private to public schools—this
payments. On the other hand, the subsidy prevents year alone, 54,000 students (11 percent of public
the prices of these products from increasing, which sector students)—as well as higher student drop-
would exacerbate inflationary-depreciation pressures, outs, especially from the most marginalized
further striking at residents purchasing power. households. Further, the most recent school
The second Special Focus discusses the closures due to the COVID-19 pandemic have
impact that the crises are having on four basic effectively cost students a “lost year” of learning.
Executive Summary xiii
الموجز
ية .وقد أظهرت
مسوحات أجراها برنامج األغذية العاملي عرب الهاتف يف أواخر 2020أن
دية وس ّ
وتهم الرشائ ّ الساحقة للقوى العاملة — تراجعاً بالغاً يف ق ّ
41يف املئة من األرس يصعب عليها الحصول عىل املواد الغذائ ّ
عل األزمة االقتصاديّة واملالية التي ترضب لبنان من بني األزمات
يا منذ أواسط ً
وربا من بني األزمات الثالث ،األكرث حدة عامل ًّ
القرن التاسع عرش .إنّها إحدى خالصات تقرير مرصد االقتصاد
العرشّ ، ل
ّ
حاجاتها األساسيّة األخرى .وارتفعت نسبة األرس التي تواجه صعوبات يف اللبناين لربيع ،2021الذي يقارن أزمة لبنان مع األزمات العامليّة األكرث
ية من 25يف املئة (متوز/يوليو-آب/أغسطس) الحصول عىل الرعاية الصح ّ ة وفق رينارت Reinhartوروغوف Rogoff (2014)4خالل الحقبة حد ً
إىل 36يف املئة (ترشين الثاين/نوفمرب — كانون األول/ديسمرب) .كام أن .2013–1857يف الواقع ،تراجع إجاميل الناتج املحيل يف لبنان من حواىل
دل البطالة ارتفع يف صفوف املشمولني باملسح ،فانتقل من 28يف املئة مع ّ 55مليار د.أ .يف العام 2018إىل حواىل 33مليار د.أ .يف العام ،2020
يف شباط/فرباير (ما قبل كوفيد) إىل حواىل 40يف املئة يف ترشين الثاين/ مع تراجع إجاميل الناتج املحيل للفرد بالدوالر األمرييك بنسبة حواىل 40
نوفمرب — كانون األول/ديسمرب. با ما يُعزا مثل هذا االنقباض القايس والرسيع إىل نزاعات يف املئة .غال ً
ية،يواجه لبنان ،وهو ذو تاريخ حافل بالنزاعات والحرب األهل ّ نف البنك الدويل لبنان عىل أنّه أو حروب .وحتى قبل ذلك ،لطاملا ص ّ
الهش أصالً .فوفق تقرير التشخيص ية تُهدّ د سلمه االجتامعي ّ مخاطر واقع ّ دد الظروف االقتصاديّة دولة هشاشة ،ونزاع ،وعنف ،وبالتايل ،قد تُه ّ
املنهجي للبنان (البنك الدويل ،) 2016 ،يعترب العنرصان التاليان من القيود
6
واالجتامعيّة املرتدية بانهيار وطني منهجي تكون له انعكاسات محتملة
ية التي ت ُعيق عملية التنمية يف لبنان ( )iالحكم الطائفي ،أي تويل األساس ّ جسد حجم الكساد االقتصادي مم يُ ّ ي االقليمي والعامليّ ،5 عىل املستويَ ْ
الحكم من قبل طبقة نخبوية تستخدم ذريعة الطائفية قناعاً لها و()ii الذي يشهده البلد ،من دون بارقة أمل بتغيري تلوح يف األفق ،نظرا ً إىل
النزاعات والعنف الناجامن جزئياً عن رصاعات واسعة النطاق يف منطقة التقاعس املتعمد عن اتخاذ السياسات املالمئة.
الرشق األوسط ،وهام عنرصان عىل عالقة متكافلة مع بعضهام البعض، وصفت األزمة يف تقرير مرصد االقتصاد اللبناين لخريف ُ ،2020
دة غضب التظاهرات يتغذيان وينموان من بعضهام البعض .وازدادت ح ّ مد» .عىل مدى رض لها لبنان بأنّها «الكساد املتع ّ االقتصاديّة التي يتع ّ
جا عىل الظروف االقتصاديّة املرتدية وإن الناشئة يف املدن اللبنانيّة احتجا ً أكرث من عام ،كانت السلطات اللبنانية تواجه وابل األزمات املتعاقبة
مم أعاق عا؛ وقُطعت الطرق الحيويّةّ ، كانت أعداد املتظاهرين أكرث تواض ً يام أكرب أزمة مالية واقتصاديّة يشهدها البلد يف زمن السلم، — ال س ّ
دد أمن دل الجرمية يُه ّ التنقّالت وكسب لقمة العيش؛ كام أن ارتفاع مع ّ وكوفيد– ،19وانفجار مرفأ بريوت — بسياسات غري مالمئة عمدا ً .وال
تعود هذه االستجابات غري املالمئة إىل نقص يف املعلومات أو إىل توجيهات
Reinhart, Carmen M. and Kenneth S. Rogoff (2014), Recovery 4 دة ،منها( )iغياب اإلجامع خاطئة ،بل هي نتيجة توليفة من عوامل ع ّ
from Financial Crises: Evidence from 100 Episodes, American عالة؛ و( )iiاالجامع السيايس يف الدفاع السيايس بشأن مبادرات سياسات ف ّ
.Economic Review: Papers & Proceedings 2014, 104(5): 50–55 دا .ويف وجه هذه مفلس ،أفاد البعض لفرتة طويلة ج ً عن نظام اقتصادي ُ
5يف كتاب أمني معلوف Le Naufrage des Civilisationsالذي تُرجم يف التحديات ،يفتقر لبنان إىل سلطة تنفيذيّة تعمل بشكل كامل وهو يف طور
االنجليزيّة إىل ،The Wreckage of Civilizationsيُسلّط الكاتب اللبناين الضوء تشكيل حكومته الثالثة يف أكرث من عام واحد بقليل .ويستمر هذا الفراغ
عىل فشل الرشق املتفيش الذي يؤدي إىل فشل الحضارات كلّها.
& Maalouf, Amin (2019), Le Naufrage des Civilisations, French
وق من أكرث من 8أشهر وحتى هذا التاريخ. املؤسيس املع ّ
.European Publications, Inc., April 2, 2019 قد يُصبح األثر االجتامعي لألزمة ،الصعب أصالً ،مأساويًّا
.”World Bank (2016) “Lebanon Systematic Country Diagnostic 6 جح أن يكون أكرث من نصف السكان دون خط الفقر الوطني. برسعة؛ يُر َّ
(البنك الدويل ( )2016التشخيص املنهجي يف لبنان). يشهد الذين يتقاضون رواتبهم وأجورهم باللرية اللبنانية — أي الغالبية
xv
در أن يكون عجز الحساب الجاري قد تراجع 10نقاط مئويّة ليبلغ 11 نُق ّ الناس الشخيص؛ وقد يسمح التفكّك الوطني باندساس مجموعات ذات
دل عجز ة مع مع ّ يف املئة من إجاميل الناتج املحيل يف العام ،2020مقارن ً نوايا سيئةّ ،
مم يولّد انعكاسات خطرية عىل مستوى األمن .وبالتايل يزداد
متوسط األجل ( )2019–2013نسبته 22.5يف املئة من إجاميل الناتج القلق ،يف لبنان ،من العوامل التي ميكن أن تؤدي إىل اضطرابات اجتامعية.
املحيل .ومع ذلك ،أدّى التوقّف املفاجئ يف التدفقات الرأسامليّة الوافدة
إىل تراجع مضطرد الحتياطي مرصف لبنان من العمالت األجنبيةّ ،
مم
يفاقم القيود عىل الواردات. التط ّ
ورات املاكرو-اقتصادية واملالية األخرية
عتب عبء التكيف الجاري /تقليص ميزانية القطاع املرصيف ي َُ
ية الساحقة يا بشكل كبري ،يُركّز عىل املودعني الصغار ،والغالب ّ تراجع ًّ تضح ذلك بشكل ية والنقديّة ظروف األزمة ،و ي ّ تقود االضطرابات املال ّ
عتب ترصيف الودائع إىل اللرية ُ َ ي الصغرية. واملؤسسات العاملة، للقوى يةّ النقد والكتلة ّم، خ والتض الرصف، سعر بني التفاعل ملموس من خالل
اللبنانيّة واالقتطاع من الودائع بالدوالر األمرييك أمرا ً واقعاً ،بالرغم من مبعناها الضيق .تنعكس الضغوط الحادة لسوق الرصف عىل األسواق
اإللتزام الرسمي من قبل املصارف ومرصف لبنان بحامية الودائع .يُ َ
عتب ية عىل شكل تقلّبات كبرية يف سعر رصف الدوالر األمرييك مقابل اللبنان ّ
عبء التكيف الجاري /تقليص ميزانية القطاع املرصيف تراجعيًّا ويُركّز اللرية اللبنانيّة ،الذي تجاوز بشكل مؤقت 15ألف لرية لبنانيّة ،قبل أن
عىل املودعني الصغار ،الذين يفتقرون إىل مصادر ادخار أخرى ،والقوى ددة تشمل يرتاجع مجددا ً .ويندرج ذلك يف سياق نظام أسعار رصف متع ّ
العاملة املحلية التي تتقاىض أجورها ورواتبها باللرية اللبنانيّة ،واملؤسسات سعر الرصف الرسمي ( 1د.أ 1507.5 = .ل.ل ،).باإلضافة إىل سعر رصف
يات تشمل أصوالً متلكها الصغرية .يدعو القطاع املرصيف إىل وضع آل ّ منصة مرصف لبنان البالغ 1د.أ 3900 = .ل.ل ،.وبشكل عام ،تراجع
الدولة ،واحتياطي الذهب ،والعقارات العامة من أجل إصالح ميزانيّاتهم متوسط سعر الرصف الذي يحتسبه البنك الدويل بنسبة 129يف املئة يف
مم يُشكّل عملية إنقاذ من القطاع العام للقطاع املايل ،ال الضعيفةّ ، مم أدّى إىل زيادة العام .2020وأث ّرت تقلّبات سعر الرصف عىل األسعارّ ،
تتامىش مع مبادئ إعادة الهيكلة التي تحمي دافعي الرضائب .وتشمل التضخّم ليبلغ 84.3يف املئة يف العام .2020ويف موازاة ذلك ،ازداد مخزون
ية الدائنني، هذه املبادئ حلول إنقاذ مبشاركة داخل ّية عىل أساس هرم ّ العملة املتداولة بنسبة 197يف املئة ،حتى بعد تراجع الكتلة النقديّة
يكن للحكومة أن تفرض رضيبة ءا من املساهمني يف املصارف .كام ُ بد ً مبعناها الواسع (التي تشمل الودائع املرصفيّة) ،والتي تأث ّرت بدورها
ية) كأداة من أجل إعادة ّ املال واألصول ةّ ي العقار األصول (عىل الرثوات عىل بتقليص ميزانية القطاع املرصيف.
هيكلة القطاع املايل بشكل تقدمي. يُقدَّ ر تقلّص منو إجاميل الناتج املحيل الحقيقي بنسبة 20.3
يف املئة يف العام ،2020إثر تقلّص بنسبة 6.7يف املئة يف العام .2019
رض قطاع السياحة إىل رضبة قويّة بشكل خاص ،ناجمة إىل حد وقد تع ّ
مجاالت الرتكيز الخاصة بعيد عن كوفيد – 19؛ وتراجع عدد املسافرين الوافدين بنسبة 71.5يف
املئة (من سنة ألخرى) ،عىل مدى األشهر الخمسة األوىل من العام 2020
ية من هشاشة لبنان وتفكّكه ،مام زاد من زادت األزمة املالية واالقتصاد ّ يات تسليم ( .)5M-2020ويف موازاة ذلك ،شهدت تراخيص البناء وعمل ّ
خطر االضطرابات املدنية واالجتامعية .وزادت األزمة من أوجه الخلل االسمنت ،وهي من مؤرشات أنشطة قطاعي البناء والعقارات ،تراج ً
عا
ية ية ،والسياسة االجتامع ّ الطويلة األمد ،مبا يف ذلك مكامن الضعف املؤسس ّ سنويًا قدره 26.9و 44.7يف املئة عىل التوايل ،خالل األشهر العرشة األوىل
واالقتصاديّة الفاشلة ،والخدمات العامة السيئة .ويف ظل هذا السياق، من العام .)10M-2020( 2020
يزداد القلق ،يف لبنان ،من العوامل التي ميكن أن تؤدي إىل اضطرابات تحسن ظاهري يف بعض املؤرشات املالية يخفي ّ يف الواقعُ ،
َي ي محتمل ْ اجتامعية .يف هذا التقرير ،نسلّط الضوء عىل محركّني اقتصاديَّ ْ در تراجع را فعل ًّ
يا .يُق َّ (كنسبة مئويّة من إجاميل الناتج املحيل) تدهو ً
ية ملحوظة. يكن أن يكون لهام انعكاسات اجتامع ّ قيد التدقيق املتزايدُ ، االيرادات بشكل حاد نتيجة االنقباض االقتصادي الشديد ،مع تراجع
ينظر مجال الرتكيز األول يف دعم الرصف األجنبي يف لبنان النسبة إىل إجاميل الناتج املحيل بشكل أكرب بسبب زيادة إجاميل الناتج
ياً. يا وسياس ًّ
يا جدّ ّ يا اجتامع ًّ مة ،والذي يشكّل تحد ً ية وامله ّللواردات األساس ّ املحيل اإلسمي الناجمة عن التضخّم .لكن ،يُقابل ذلك تراجع أكرب يف
عتب دعم الرصف األجنبي الحايل تشويهيًّا ،ومكلفًا ،وتراجعيًّا. من جهة ،يُ َ مم ييل :انخفاض تسديدات الفوائد النفقات الجارية ،التي تستفيد ّ
فهو يشكل ضغوطًا كبرية عىل ميزان املدفوعات يف لبنان .ومن جهة أخرى، (بسبب عدم سداد اليوروبوندز وبسبب ترتيب مالئم مع مرصف لبنان
مم يزيد من ضغوط يحول الدعم دون زيادة أسعار هذه املنتجاتّ ، د من التحويالت؛ باإلضافةهنا أيضاً إىل أثر بشأن الدين املحيل)؛ والح ّ
ية للمواطنني. ّ رش
ائ ال د أكرث من القدرة التضخّم-انخفاض قيمة العملة ،فيح ّ ارتفاع إجاميل الناتج املحيل (الذي يؤدي إىل انخفاض النسبة) .وبالتايل،
رق مجال الرتكيز الثاين إىل وقع األزمات عىل أربع خدمات ويتط ّ تحسن امليزان املايل الكيل للعام 2020بقدر 0.7نقطة مئوية ليبلغ مع ّ
ية :الكهرباء ،وإمدادات املياه ،والرصف الصحي ،والتعليم. عامة أساس ّ -4.9يف املئة من إجاميل الناتج املحيل ،تراجع الرصيد األويل 2.3نقطة
مد بشكل أكرب الخدمات العامة الضعيفة أصالً فقد أضعف الكساد املتع ّ مئويّة إىل -2.8يف املئة من إجاميل الناتج املحيل.
فتوسع بشكل ملحوظ عدد ّ الفقر الت د
ّ مع زيادة ريْن)i( : من خالل أث َ أدى االنقباض االقتصادي الحاد إىل تراجع متناسب يف الواردات، ّ
السكان غري القادرين عىل تكبّد كلفة الخدمات البديلة الخاصة (السبل وبالتايل ،من املنتظر أن يؤدي إىل تقلّص يف عجز الحساب الجاري .خالل
ية الخدمات العامة يف مع نوع ّ التي اعتمدها املواطنون يف السابق للتك ّ األشهر العرشة األوىل من العام ،2020تقلّصت واردات السلع بنسبة 45
املرتدية) ،فباتوا يعتمدون بالتايل بشكل أكرب عىل الخدمات العامة؛ و()ii عا يف عجز التجارة يف السلع قدره 54.8يف املئة. مم ولّد تراج ً يف املئةّ ،
xvi )LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3
التكاليف .وبسبب تراجع إمدادات املياه من مؤسسات املياه يف العام تهديد االستمراربة املالية للقطاع وعمله األسايس من خالل زيادة
ة وأقل ،2020كان عىل املواطنني االعتامد عىل بدائل أخرى أكرث كلف ً تكاليفه وخفض إيراداته.
مالءمةً ،عىل غرار صهاريج املياه وقوارير املياه ،التي ارتفعت أسعارها. وبشكل خاص،
دد تراجع خدمات الرصف الصحي بزيادة انتشار األمراض املنقولة •يُه ّ
با عىل الصحة العامة الهشّ ة أصالً. ً سل ّر ثيؤ مم
ّ املياه، يف ية بإنهاء عقود القطاع الخاص •يُه ّ
دد االفتقار الحاد للعمالت األجنب ّ
•شكّلت األزمات املتعاقبة ضغوطًا كبرية عىل قطاع التعليم يف لبنان. لصيانة محطات توليد الطاقة الكهربائيّة وتوليد الطاقة املؤقت .ويف
دالت الفقر إىل نزوح جامعي للطالب من املدارس تؤدي زيادة مع ّ الوقت نفسه ،تتقلّص إيرادات رشكة كهرباء لبنان ،التي هي باللرية
الخاصة إىل املدارس الرسمية — بلغ هذا العام وحده 54000 اللبنانيّة ،بسبب الخسائر الفنية والتجاريّة املتزايدة وتلك املرتبطة
دل تلميذ ( 11يف املئة من تالمذة القطاع العام) – باإلضافة إىل مع ّ بالجباية .ويُتوقَّع أن تزيد مؤسسة كهرباء لبنان من فرتات التأنني يف
رسب مدريس أعىل ،ال سيّام من األرس األكرث تهميشً ا .إىل ذلك ،خرس ت ّ التغذية بالتيار الكهربايئ إلدارة القصور يف تدفقاتها النقديّة.
را
ً مؤخ أبوابها املدارس إقفال مع ّم ل التع من ا م
ً عا ً ا فعلي التالمذة •يف العام ،2020افتقرت مؤسسات املياه إىل اإلمدادات وااليرادات
بسبب جائحة كوفيد.-19 واملوارد البرشية واملاليّة بشكل خطر ،يف ما شهدت زيادة مضطردة يف
الموجز xvii
RÉSUMÉ
L
a crise économique et financière qui sévit an, les autorités libanaises ont été assaillies par une
au Liban est probablement l’une des dix, série de crises successives — à savoir, la plus grande
voire l’une des trois pires crises que le crise financière et économique du pays en temps de
monde ait connu depuis le milieu du XIXe siècle. paix, la COVID-19 et l’explosion du Port de Beyrouth
Telle est la conclusion de l’édition Printemps 2021 de — auxquelles ils ont apporté des réponses politiques
l’Observatoire économique du Liban (ou LEM, pour délibérément inadéquates. Inadéquations qui mal-
Lebanon Economic Monitor), dans laquelle la crise au heureusement ne proviennent pas d’un manque
Liban est considérée comme l’une des crises mondiales d’informations ou de mauvaises directives mais d’une
les plus sévères au regard de la liste établie par Reinhart combinaison (i) d’un manque de consensus politique
et Rogoff (2014)7 durant la période 18572013. En effet, à l’égard d’initiatives de politiques effectives ; et
le PIB du Liban a chuté de 55 milliards de dollars en (ii) d’un consensus politique qui défend un système
2018 à environ 33 milliards de dollars en 2020, avec économique en faillite — un système qui a profité à
une baisse d’environ 40 % du PIB par habitant. Une certains pendant longtemps. Face à ces défis, le pays
telle contraction, si brutale et rapide, est d’habitude est actuellement doté d’un pouvoir exécutif qui n’est
attribuée à des conflits ou des guerres. La Banque pas pleinement fonctionnel, attendant la formation
mondiale a longtemps considéré le Liban, et ce, bien de son troisième gouvernement en un peu plus d’un
avant la crise, comme un pays en situation de fragilité, an, et est paralysé par un vide institutionnel qui dure
conflit et violence (FCV). À ce titre, il est possible que depuis plus de 8 mois.
les conditions socio-économiques difficiles que connaît L’impact social de la crise, déjà désastreux,
le pays risquent d’entraîner une faillite systémique à peut rapidement devenir catastrophique; plus de
l’échelle nationale, avec d’éventuelles conséquences la moitié de la population serait en-dessous du
mondiales.8 Ceci illustre la magnitude de la dépression
économique que le pays subit, et, malheureusement, 7
Reinhart, Carmen M. and Kenneth S. Rogoff (2014),
le fait qu’aucune perspective de changement ne soit Recovery from Financial Crises: Evidence from 100
à l’horizon eu égard à l’inaction désastreuse mais Episodes, American Economic Review: Papers &
Proceedings 2014, 104(5): 50–55.
néanmoins choisie des politiques. 8
Dans l’ouvrage d’Amin Maalouf (2019), Le Naufrage des
Dans l’édition Automne 2020 du LEM, la Civilisations, l’Académicien et auteur libanais montre
crise économique que traverse le Liban a été qua- comment les échecs du Levant se transforment en un
lifié de Dépression Délibérée. Pendant plus d’un échec de toutes les civilisations.
xix
seuil national de pauvreté. Ceux qui sont payés en et l’augmentation rapide de la masse monétaire,
livres libanaises — soit la majorité de la main-d’œuvre sont des conditions génératrices de crises.
— ont vu et voient leur pouvoir d’achat se rétrécir Les sévères pressions du marché de change sur
comme une peau de chagrin. Selon des sondages les marchés libanais sont illustrées par de lourdes
téléphoniques menés fin 2020 par le Programme fluctuations du taux de change de la livre face au
alimentaire mondial (PAM), 41 % des ménages ont dollar américain, qui a temporairement dépassé
rapporté des difficultés à accéder à la nourriture LBP 15 000/US$, avant de baisser à nouveau. Ceci
et à d’autres besoins fondamentaux. La part des s’inscrit dans le cadre d’un système à taux de change
ménages connaissant des difficultés à accéder aux multiples, qui comprend le taux de change officiel à
services de santé est passée de 25 % en juillet-août à LBP 1 507,5/US$, ainsi que celui de la plateforme
36 % en novembre-décembre. Le taux de chômage a de la Banque du Liban (BdL) à LBP 3 900/US$. En
également augmenté parmi les répondants, passant général, le taux de change effectif moyen calculé
de 28 % en février (avant la COVID) à environ 40 % en par la Banque mondiale (AER) a baissé de 129 %
novembre-décembre. en 2020. L’impact des fluctuations des taux de
Le Liban, historiquement touché par les change sur les prix a provoqué une recrudescence
conflits et la guerre civile, est confronté à des de l’inflation, enregistrant une moyenne de 84,3 % en
menaces réalistes qui mettent en danger sa 2020. En parallèle, le stock de monnaie en circulation
paix sociale déjà fragile. Comme il a été mis en a augmenté de 197 %, même si la masse monétaire
évidence par le passé (Banque mondiale, 20169), les au sens large (qui comprend les dépôts bancaires)
principales contraintes au développement au Liban s’est réduite en raison du désendettement du secteur
sont i) l’accaparement des ressources par l’élite, financier.
sous couvert de confessionnalisme et ii) les conflits La croissance du PIB réel a connu une
et la violence — ces deux facteurs entretenant une contraction de 20,3 % en 2020, aggravant celle
relation symbiotique, l’un se nourrissant de l’autre déjà importante de 6.7 % observée en 2019. Le
pour se développer mutuellement. La colère des secteur du tourisme a été particulièrement touché,
manifestations, bien que plus modestes en nombre, en grande partie en raison de la COVID-19 ; les
est récemment montée d’un cran, émergeant dans arrivées de touristes ont baissé de 71,5 %, (d’une
des villes à travers le Liban pour protester contre année sur l’autre), durant les cinq premiers mois de
les conditions économiques exsangues. De ce fait, 2020 (5M-2020). Dans le même temps, les permis
l’accès à des routes vitales est coupé, entraînant des de construction et les livraisons de ciment (moyens
interruptions prolongées qui entravent la mobilité et intermédiaires pour la construction et l’immobilier)
réduisent plus encore les moyens de subsistance ; ont baissé de 26,9 % (d’une année sur l’autre) et de
l’augmentation des taux de criminalité menace la 44,7 % (d’une année sur l’autre) durant les dix pre-
sécurité des personnes ; et la fragmentation nationale miers mois de 2020 (10M-2020).
peut permettre l’infiltration de groupes menaçants, ce Une amélioration apparente de certains
qui aurait des implications graves sur la sécurité du indicateurs fiscaux (en pourcentage du PIB)
pays. Les politiciens-timoniers du Liban, bloquant la masque une détérioration effective et notoire.
barre du pays, le rapprochent ainsi inexorablement Les revenus ont baissé de manière significative
de récifs meurtriers et d’un naufrage tragique. en raison d’une sévère contraction économique,
le ratio au PIB se détériorant davantage à cause
d’une augmentation du PIB nominal due à l’inflation
Les récents développements galopante. Toutefois, cette situation est largement
économiques compensée par une baisse plus importante des
Les turbulences monétaires et financières, et 9
World Bank (2016) “Lebanon Systematic Country
plus concrètement le taux de change, l’inflation Diagnostic”.
xx LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
dépenses courantes, ayant bénéficié des éléments des principes de restructuration visant à protéger les
suivants : des paiements d’intérêts inférieurs (en contribuables et les déposants. Ces principes com-
raison du défaut de paiement des Eurobonds et d’un prennent des solutions de cautionnement basées
arrangement favorable, bien qu’illusoire au niveau du sur une hiérarchie de créditeurs, à commencer par
secteur publique consolidé, avec la BdL concernant les actionnaires des banques. Le gouvernement peut
la dette intérieure) ; des baisses de transferts ; et d’un également imposer un impôt sur la fortune (actifs
PIB sur la base d’un dénominateur. Ainsi, alors que le financiers et biens immobiliers) comme moyen de res-
solde budgétaire global de 2020 a progressé de 0,7 tructurer de manière progressive le secteur financier.
point de pourcentage (pp) pour atteindre –4,9 % du
PIB, le solde primaire a régressé de 2,3 pp, atteignant
–2.8 % du PIB. Les champs d’intérêt particuliers
La contraction économique accrue signifie
une baisse conséquente des importations, et La crise économique et financière a accentué la
donc, une réduction anticipée du déficit du compte fragilité et la fragmentation du Liban, augmentant
courant. Durant les dix premiers mois de 2020 ainsi le risque de mécontentement social et civil.
(10M-2020), les importations de marchandises ont La crise a exacerbé les déficiences nationales à
régressé de 45 %, provoquant une baisse de 54,8 % long terme, y compris les fragilités institutionnelles,
du déficit lié aux échanges de biens. Nous estimons la politique socioéconomique tenue en échec
que le déficit du compte courant a baissé de 10 et la mauvaise prestation des services publics.
points de pourcentage pour atteindre 11 % du PIB en Dans un tel contexte, le pays connaît une lassitude
2020, à comparer avec la moyenne à moyen terme grandissante face à d’éventuels facteurs susceptibles
(2013–2019) de 22.5 % du PIB. Toutefois, l’arrêt sou- de déclencher un mécontentement social. Dans cette
dain et soutenu des flux de capitaux entrants signifie édition du LEM, nous mettons l’accent sur deux
un épuisement continu des réserves de change (FX) éventuels éléments déclencheurs qui sont de plus en
à la BdL, accentuant par là-même les contraintes sur plus l’objet d’une attention particulière, et qui auraient
les importations. des implications sociales de grande envergure.
Le fardeau de l’ajustement en cours/du Le premier champ d’intérêt particulier
désendettement dans le secteur financier est concerne la bonification des réserves étrangères
particulièrement régressif, concentré sur de au Liban pour les importations essentielles
petits déposants, la plus grande majorité de la et critiques, ce qui constitue un véritable défi
maind’œuvre, et des entreprises de taille plus sociopolitique. D’une part, en plus d’être onéreuse
modeste. De facto, la « lirafication » et les « haircuts » et régressive, la bonification des réserves étrangères
(les « ponctions ») sur les dépôts en dollars sont a présentement un effet de distorsion et exerce en
importants, malgré l’engagement officiel de la Bdl et outre des pressions considérables sur la balance des
des banques à sauvegarder les dépôts. Le fardeau paiements du Liban. D’autre part, elle empêche une
de l’ajustement en cours/du désendettement dans augmentation du prix de ces produits, ce qui accen-
le secteur financier est particulièrement régressif tuerait les pressions générées par l’inflation et la
et concentré sur de petits déposants qui n’ont pas dépréciation et porterait un coup au pouvoir d’achat
d’autres sources d’épargne, une maind’œuvre locale de la population.
rémunérée en livres libanaises et des entreprises de Le deuxième champ d’intérêt particulier
taille plus modeste. Le secteur bancaire appelle à la porte sur l’impact que les crises ont sur quatre
mise en place de mécanismes qui tiennent compte services publics de base : l’électricité, l’eau, les
des biens de l’État, des réserves en or et des actifs services sanitaires et l’éducation. La Dépression
immobiliers publics afin de rétablir l’équilibre de Délibérée a davantage fragilisé les services publics
ses bilans déficients. Il s’agit d’un renflouement du déjà vulnérables de deux façons : (i) elle a augmenté
secteur financier qui ne s’inscrit pas dans le cadre de manière significative les taux de pauvreté,
Résumé xxi
élargissant ainsi la portion de la population n’ayant en eau, de revenus et de ressources financières
pas les moyens de se payer des biens et services et humaines, qui est venue s’ajouter à la
privés de substitution (ce qui était la manière dont la flambée des coûts. En 2020, en raison de cet
population s’était auparavant adaptée à la prestation approvisionnement limité, la population a été
de services publics de mauvaise qualité) et étant, par obligée de recourir à d’autres alternatives plus
conséquent, davantage dépendante des services onéreuses et moins pratiques, telles que les
publics ; et (ii) elle menace la viabilité financière et citernes d’eau et l’eau en bouteille, dont les prix
l’opérabilité fondamentale du secteur en augmentant se sont envolés.
ses coûts et en baissant ses revenus. L’on retient plus • La dégradation des services d’assainissement
particulièrement les éléments suivants : risque d’intensifier la propagation de maladies
transmises par l’eau, ce qui affectera de manière
• Une sévère pénurie de devises étrangères négative la santé publique déjà vulnérable.
menace la résiliation des contrats du secteur • Les crises successives auxquelles le Liban a été
privé pour l’entretien des centrales électriques confronté ont exercé de sévères pressions sur
et la génération provisoire d’électricité. Dans le le secteur de l’éducation. En raison des taux de
même temps, les revenus d’Électricité du Liban pauvreté élevés, les élèves ont abandonné les
(EdL), qui sont en livres libanaises, ont été réduits écoles privées au profit des écoles publiques —
considérablement en raison des pertes techniques cette année seulement, 54 000 élèves (11 % des
et commerciales et du manque de paiements élèves du secteur public). À noter également :
collectés. Il se pourrait donc qu’EdL augmente un taux d’abandon scolaire plus élevé,
ses coupures de courant rotatives afin de pouvoir particulièrement au sein des ménages les plus
gérer son manque de flux de trésorerie. marginalisés. En outre, la récente fermeture des
• En 2020, les Offices des Eaux ont connu une écoles en raison de la COVID-19 a fait perdre aux
grave pénurie en matière d’approvisionnement élèves une année d’apprentissage.
xxii LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
1
THE POLICY CONTEXT
L
ebanon faces a dangerous depletion of the local labor force and smaller businesses. De
resources, including human capital since facto Lirafication and haircuts on dollar deposits are
brain drain has become an increasingly ongoing despite BdL’s and banks’ official commit-
desperate option. Over a year into the financial crisis, ment to safeguarding deposits. The burden of the
Lebanon has yet to identify, least of all embark upon, a ongoing adjustment/deleveraging is regressive
credible path toward economic and financial recovery. and concentrated on the smaller depositors who
In fact, Lebanon lacks a fully-functioning executive lack other source of savings, the local labor force
authority and is currently in the process of forming that is paid in LBP, and smaller businesses. The
its third Government in a little over a year. Meanwhile, banking sector is advocating for mechanisms that
social discontent has spilled over to street action incorporate state-owned assets, gold reserves, and
even under COVID-19 conditions; internal political public real estate in order to overhaul their impaired
discord and fragmentation continues; and geopolitical balance sheets. This constitutes a bailout of the
tensions complicate solutions. In consequence, highly financial sector and is inconsistent with the restruc-
skilled labor is increasingly likely to take up potential turing principles that protect taxpayers. These
opportunities abroad, constituting a permanent social principles include bail in solutions based on a hier-
and economic loss for the country. archy of creditors, starting with banks’ shareholders.
Lebanese authorities and the IMF began Government can also apply a wealth tax (on financial
discussions in May 2020. The discussions even- and real assets) as a tool to progressively restructure
tually stalled as differences and inconsistencies the financial sector.
emerged within the Lebanon team regarding the Lebanon urgently needs to adopt and
Government’s financial recovery program. IMF dis- implement a credible, comprehensive and coor-
cussions await the formation of new Government. dinated macro-financial stability strategy, within
The burden of the ongoing adjustment/ a medium-term macro-fiscal framework. This
deleveraging in the financial sector is highly strategy would be based on: (i) a debt restructuring
regressive, concentrated on smaller depositors, program that would achieve short-term fiscal space
1
and medium-term debt sustainability; (ii) comprehen- confidence and stability in the exchange rate; (iv) a
sively restructuring the financial sector in order to phased fiscal adjustment aimed at regaining confi-
regain solvency of the banking sector; (iii) adopting dence in fiscal policy; (v) growth enhancing reforms;
a new monetary policy framework that would regain and (vi) enhanced social protection.
2 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
2
RECENT MACRO-
FINANCIAL
DEVELOPMENTS
Output and Demand is the lowest PMI recorded since it was first published
in 2013 (see Box 1 for the impact on firms and labor
The compounded crises, namely, the financial market). Meanwhile, the real estate sector has been
crisis, COVID-19 and the Port of Beirut (PoB) subject to two offsetting factors; on the one hand, con-
explosion, have had staggered impacts on struction permits and cement deliveries—considered
output and with differentiated magnitudes. Due to to be indicators of future and ongoing construction,
insufficient high frequency data, precise identification respectively—suffered respective declines of 26.9
of each of those impacts is a challenging task. In percent (yoy) and 44.7 percent (yoy) over 10M-2020.
order to draw empirical conclusions, we resort to a On the other hand, throughout 2020 real estate
combination of methodologies and models. To gauge sales thrived as some depositors sought means to
the impact of financial crisis along with COVID-19 utilize their otherwise untransferable bank deposits.13
effects, we use Mixed-Data Sampling (MIDAS)
methods to assess the state of the economic cycle 10
World Bank (2020), Beirut Rapid Damage and Needs
using available high frequency measures of economic Assessment, August 2020.
activity (See Annex A). The World Bank had earlier
11
According to the RDNA, the disaster event is estimated to
cause (i) up to 0.4 and 0.6 percentage point (pp) declines in
estimated the economic impact of the PoB explosion
the growth rate of real GDP in 2020 and 2021, respectively,
through a Rapid Damage and Needs Assessment due to losses in the stock of physical capital; plus potentially
(RDNA).10,11 (ii) import constraints that could subtract an additional of 0.4
Real GDP is estimated to have contracted and 1.3 pps from growth in 2020 and 2021, respectively.
by 20.3 percent in 2020 (Figure 1).12 High frequency 12
This represents a slight downward revision from the Fall
indicators support a substantial contraction in eco- 2020 LEM which projected a 19.2 percent contraction in
real GDP growth for 2020.
nomic activity. BLOM’s monthly Purchasing Manager’s 13
The financial sector facilitated real estate purchases using
Index (PMI), which captures private sector activity, pre-October 2019 dollar deposits under conditions of capital
averaged 41.1 over 2020—PMI values below 50 rep- controls (and therefore, lack of alternatives to get those
resent a contraction in economic activity. In fact, this deposits out), leading to an increase in such purchases.
3
While the Contraction in Real GDP
FIGURE 1 • Net Exports are Expected to be the
FIGURE 2 •
Commenced in 2018, it Accelerated Sole Positive Contributor to Real GDP
Sharply in 2020
Real GDP Components
Real GDP Growth (%) 15
20 10
16.4
5
15
11.3
10.8
0
10.2
9.1
9.3
10
8.1
–5
8.0
7.5
2.7
6.4
–10
3.9
3.8
3.8
5 3.9
3.4
2.5
2.5
1.7
1.5
1.5
–0.8
1.1
0.9
0.9
–15
0.2
0
–20
–1.9
–5 –25
–6.7
–10 –30
–35
–15 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020e
–20
Private Consumption Government Consumption
2020 –20.3
–25 Gross Fixed Capital Investment Net Exports
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
Statistical Discrepancy Real GDP Growth
Sources: CAS and WB staff calculations. Sources: CAS and WB staff calculations.
Real estate registration fees increased by 104.8 Fiscal Developments
percent in 2020. The retail sector suffered siz-
able losses, due to a combination of the financial An ostensible improvement in fiscal indicators
crisis and the COVID-19 lockdown measures. BTA (as a percentage of GDP) masks an actual
Fransabank retail trade index (in real terms) declined deterioration. While both revenue and expenditure
by 73.1 percent over the first nine months of 2020 sharply contracted as a ratio of GDP due to the
(9M-2020). crisis, the net impact on the overall fiscal balance
On the demand side, net exports was the was positive. Revenues are estimated to have
sole contributor to growth in 2020, for the second reached 11.5 percent of GDP in 2020, down by a
year in a row (Figure 2). According to Custom’s data, staggering 9.1 pp of GDP from the previous year.
the merchandise trade deficit was reduced by 54.8 This deterioration was the result of a two-pronged
percent (yoy) over the first 11 months of 2020 (11M- development. First, depressed economic activity
2020), benefitting from a 45.4 percent (yoy) retraction induced significant reductions in tax and non-tax
in imports and despite a 4.2 percent (yoy) decrease revenues in nominal terms. Second, a substantial
in exports. The improving merchandise trade balance increase in nominal GDP—as a sharp increase in
has been partially offset by a deterioration in the the GDP deflator more than offset the contraction in
trade in services balance, as a result of COVID-19’s real economic activity—created a denominator effect,
substantial impact on the tourist sector; tourist arrivals adding to the sharp decline of revenues as a percent
decreased by 71.5 percent (yoy) in the first five of GDP. Similarly, expenditures are estimated to have
months of 2020, and hotel occupancy rate (published fallen to 16.4 percent of GDP in 2020, down by a
by Ernst & Young) averaged a mere 16.6 percent mammoth 14.8 pp of GDP compared to 2019. This
over 9M-2020. Similarly, private consumption, which fall is primarily attributed to a decrease in current
averaged 92.3 percent of GDP over 2015–2018, is expenditures, which benefited from: lower debt-
estimated to have taken a severe blow in 2020; Byblos servicing costs due to the default on foreign debt;
Bank/AUB’s consumer confidence index declined a favorable arrangement with BdL on its holding
by 65.1 percent in 9M-2020, compared to the same of Treasury Bonds (TBs); lower transfers; and a
period in 2019. denominator-led GDP effect. As a result, overall fiscal
4 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
BOX 1: THE IMPACT OF MULTIPLE CRISES ON FORMAL FIRMS AND THE LABOR MARKET IN LEBANONa,b
Impact on Firms
The World Bank conducts Enterprise Surveys regularly in active member countries. It conducted one such survey in Lebanon in 2019/2020;
this happened to be a pre-crisis survey, as the country would be subsequently hit by compounded crises—the financial crisis, COVID-19 and
the August 4th explosion at the Port of Beirut (PoB). To assess the impact on firms from these crises, a follow up survey was conducted in
Lebanon in November 2020.
From the surveys, we find that almost one out of five firms originally surveyed are confirmed or assumed permanently closed,c
manufacturing firms that remain open are operating at 35 percent of capacity, and almost half of the firms have been affected by the PoB
explosion. As illustrated by Figure 3, 16.6 percent of firms surveyed are confirmed or assumed permanently closed, while 72 percent
have been temporarily closed at some point since the beginning of the pandemic due to Covid-19, with a 10 week average closure period.
Further, 46 percent of all surveyed firms—both in Beirut and elsewhere—have been affected by the August 4 explosion either directly or
indirectly. In fact, 22 percent of firms had to close temporarily because of the explosion and 17 percent experienced interruption of their
supply chain.
Four out of five firms say their sales declined significantly together Firm Status
FIGURE 3 •
with demand. Since the beginning of the financial crisis (October
2019), 79 percent of firms surveyed reduced sales (nominal) 100%
14%
by an average of 69 percent, with small firms bearing a greater 90% 19% 15% 18%
31% 29% 29%
impact. In reflection, firms reported a 74 percent drop in (real) 80%
demand for their products and services. The lack of demand is 70%
affecting the manufacturing sector the most severely; 81 percent 60% 62% 62%
60% 63%
of surveyed manufacturing firms reported a drop in sales that is 50% 45%
equivalent to 70 percent on average. Exporters were also hit 40% 56% 66%
hard. One out of four firms reported an average 70 percent drop 30%
4% 6%
in exports. However, 20 firms were able to increase their exports; 20% 5% 7% 3%
21% 3%
half of these firms were in the manufacturing sector, mainly food 10% 17% 16% 16% 19%
13% 1%
and machinery, but also in retail and ICT. This increase could be 0% 4%
All
Manufacturing
Wholesale &
Retail
Other
Services
Small
Medium
Large
linked to the improved competitiveness resulting from the severe
unofficial depreciation of the currency, or the drop in domestic
demand due to the crisis, forcing firms to look for customers
overseas.
Share of Firms with Unknown Status or Refused to Answer
More than half of firms surveyed are experiencing liquidity Share of Firms Open Share of Firms Temporarily Closed
challenges. Since the beginning of the financial crisis, 55 Share of Firms Confirmed and Assumed Closed
percent of firms experienced a decline in their cashflow and
around 75 percent of firms decreased sales and purchases on Source: 2021 Follow up to the Enterprise Survey.
credit. Only 13 percent of firms reported relying on banks or
financial institutions to finance their liquidity shortfalls. Instead,
28 percent of firms are financing their needs through equity, and 8 percent through delayed payments.d
The future looks bleak and uncertain. One out of four firms does not think recovery is possible. Twenty-four percent of firms expect to fall
(further) into arrears and default on liabilities in the next 3 months, projecting their survival at less than 7 months or 27 weeks, given current
costs. In sum, and according to firms reported expectations, one out of four firms will have to exit by June 2021.
COVID-19 adaptation is relatively low, with few firms ramping up online operations. Around 6 percent of firms have either started or
increased their online activities, while 8 percent increased online delivery. A relatively higher share of firms (18 percent) have started to
increase or has already increased remote work. There is a significant gap between small and large firms, with the latter showing more
capacity to adapt.
Firms surveyed rank credit guarantees, tax deferrals and exemptions for social security contributions as the most helpful policy potential
responses. But preferences vary by firm characteristics and by gender ownership. When asked “What would be the most effective policy
to help firms cope with the crisis?” 30 percent of firms indicated credit guarantees, 18 percent pointed to tax deferrals and 16 percent
of firms expressed a preference for exemptions to National Social Security Fund (NSSF) contributions. Small firms preferred credit
guarantees, with tax deferrals in second place, while large firms preferred NSSF exemptions at a slightly higher rate than tax deferrals.
Meanwhile, 34 percent of male owned firms ranked credit guarantees as more effective, while 43 percent of female owned firms ranked
NSSF exemptions as their top choice.
(continued on next page)
Recent Macro-Financial Developments 5
BOX 1: THE IMPACT OF MULTIPLE CRISES ON FORMAL FIRMS AND THE LABOR MARKET IN LEBANONa,b
(CONTINUED)
Impact on Jobs
The multiple crises exacerbated unemployment. One in five workers lost their jobs since October 2019, while 61 percent of firms surveyed
decreased the number of permanent workers by 43 percent on average. Medium-size and large firms laid-off a larger number of workers:
76 percent of large firms surveyed downsized by an average of 37 percent, while 70 percent of medium firms downsized by 43 percent.
That smaller firms shrunk relatively less could be due to the different nature of the relationship between managers and workers, or the fact
that small firms may have less flexibility in scaling back production. These numbers cover only formal firms. The impact is expected to be
worse amongst informal firms and micro-sized formal firms, which are not included in the survey.
Only 16 percent of workers in the survey sample are women; in large firms, the female share of workers is only 10 percent (Table 1). The
crisis appears to have hit female employees relatively less than male employees. The net job loss was 2 percentage points (pp) lower for
women than men on average (5 pp less in the case of small firms, but 1 pp more in large firms). Although the crises can impact women’s
access to economic opportunities in different ways, the pandemic-related school closures are likely to have made it particularly difficult
for women to juggle work and care responsibilities. In addition, the firm survey shows that 10 percent of business affected by the PoB
explosion were female-owned, out of which more than half (54 percent) were of small size and 29 percent were medium sized. This loss of
women-led businesses diminishes future employment opportunities for women. According to a study conducted by Stand for Women, half
the businesses that closed will not be able to reopen without some form of assistance.e
TABLE 1 • Average Change in Full-Time Employees by Gender for All Firms Surveyed
Average percentage
change in full-time
employees
Size Number of Average percentage of females to total
(in full-time employees) firms Average size full-time employees Male Female
5–19 219 8.4 16% –25% –20%
20–99 123 37.1 20% –30% –27%
100 or more 37 171.3 10% –22% –23%
All firms 379 24.8 16% –27% –25%
Relatively few firms surveyed have resorted to other labor adjustment measures, such as reducing salaries, benefits, or working hours.
Only, 13 percent of firms have decreased salaries by around 45 percent, while 29 percent of firms increased salaries of their employees
by around 40 percent. Note that these nominal increases were in fact negative in real terms, as inflation reached triple digits during this
period. Additionally, according to the survey, 23 percent of firms have decreased benefits to their workers, by 79 percent on average,
while 41 percent of firms have decreased working hours by 36 percent on average (the other 59 percent of firms reported no change to
workers’ hours).
a
This Box was prepared by Zeina El Khoury (Private Sector Specialist, FCI) and Angela Elzir Assy (Labor Market Specialist, Jobs Group, SPJ); with inputs from Elizabeth
Ruppert Bulmer (Lead Economist, Jobs Group), and data analysis by Reyes Aterido (Consultant). Survey design was supported by Patricia Haydamous (Economist, ETC). The
survey was implemented in collaboration with DEC Enterprise Analysis Unit. The survey and analysis are funded through the Lebanon PROSPECTS Partnership.
b
Unless otherwise stated, the analysis is based on the November 2020 follow-up survey to the standard Enterprise Surveys (ES). This follow-up survey aims to measure the
impact of multiple crises (the financial and COVID-19 crises and the explosion at the PoB) on the private sector, by combining the baseline ES data collected with the follow-up
data. The survey topics cover the following dimensions: sales, production, labor, finance, preferred support policies, future prospects and business closures. Of the 532
originally surveyed firms, only 379 were reached between November and December 2020. Surveys were phone-based, and multiple contact attempts were made.
c
There are two estimates of the share of firms that have closed, the share of firms that were confirmed to have closed since the COVID-19 pandemic was declared (3.6
percent), and another wider definition of closed firms, which includes the firms that closed since the baseline ES, and also the firms that could not be contacted during fieldwork
[either those with non-functioning phone lines (2.8 percent), or did not answer multiple calls (10.2 percent)] and therefore are assumed to have closed (16.6 percent).
d
Fifty-two percent of firms experiencing liquidity shortfalls reported that they used “other” sources of cash flow shortages. Given that all formal sources of financing
operations are included in the response options, we assumed that a large proportion of firms did not finance their shortages but rather either reduced operation costs or
financed through informal channels like family or friends.
e
UN Women, 2020. Gender Alert on COVID-19 Lebanon.
6 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
deficit is estimated to have shrunk to 4.9 percent of a large part of the macro adjustment, even if not by
GDP, a narrowing of 5.6 pp of GDP from 2019, and design from policy makers, is de facto taking place
compared to a medium-term average of 8.6 percent through (1) an inflation tax on the revenue side (for
(Figure 4). However, the primary balance is estimated items other than those where the tax is set in fixed Lira
to have worsened in 2020, reaching –2.8 percent terms such as most fees);16 and (2) a real cut on the
of GDP, compared to –0.5 percent in 2019, and a expenditure side. Both are blunt instruments, but with
medium-term average of 0.5 percent. This widening high inflation they generate a huge and fast adjust-
is surprising for countries that are in sovereign ment. Hence, as is illustrated by the 2021 Budget
default as financing shortages usually forces them to proposal (Box 2), keeping the overall deficit at the
run a primary balance. In the case of Lebanon, BdL is same nominal number implies a sharp consolidation.
providing the required financing for the government, Public debt ratios, which were already noto-
notwithstanding its own financial difficulties. riously unsustainable, are further aggravated by
Fiscal data available over the first eight the economic crisis. Debt-to-GDP is estimated to have
months of 2020 confirm fiscal distress, despite reached 174 percent by the end of 2020, increasing
the apparent improvement in the overall bal- by 3 pp from the 2019 ratio. The sharp depreciation
ance. Revenues were down by 20.2 percent (yoy) in the local currency has implied a significantly lower
over 8M-2020, with across the board decreases in dollar value for domestic debt in 2020, lowering the
tax and non-tax revenues. VAT and custom revenues dollar value for total debt (the numerator in the debt-
were highly affected by the contraction in economic to-GDP ratio); this is, however, more than offset by a
activity, registering respective decreases of 49.7 per- significantly lower denominator, GDP in US$, due
cent (yoy) and 34.5 percent (yoy); telecom revenues also to the currency depreciation, leaving a slightly
were also down by 56.5 percent (yoy) in 8M-2020. larger debt-to-GDP ratio.17 The expected worsening
Meanwhile, total expenditures declined by 18.4 of the exchange in 2021 will further exacerbate this
percent (yoy), over the same period. This was partly dynamic (Figure 5). So, whereas the surge in inflation
a result of a 58.9 percent (yoy) decrease in interest
payments driven by (i) a 87.9 percent (yoy) decline 14
The agreement entails non-payment of coupon
in foreign debt interest payments, an implication of obligations on Treasury bonds held by BdL.
the Eurobond default; and (ii) a 41.3 percent (yoy) 15
This is driven by a combination of lower oil prices, falling
decrease in domestic debt interest payments due to demand due to economic crisis, and increased power
a preferential agreement between BdL and Ministry cuts as tool for savings by EdL.
16
This effect on the revenue side in Lebanon, however, is
of Finance.14 Decreases of 37.1 and 39 percent (yoy)
weakened by a combination of (i) a large chunk of the
in transfers to EdL,15 and municipalities, respectively, tax base is on imported goods, which for tax reasons are
also contributed to the total decline in expenditures. still valued at the official, thereby generating artificially
Hence, while the overall fiscal deficit narrowed by lower revenues; (ii) a collapse in compliance; and (iii) the
13.1 percent over 8M-2020, the primary balance Tanzi effect—the reduction in the volume of tax collection
deteriorated by 409 percent, flipping from a surplus and a deterioration of real tax proceeds being collected
resulting from high inflation in a country.
of LBP 555 billion to LBP –1,713 billion.
Tanzi, Vito (1977), Inflation, Lags in Collection, and the
High inflation rates drive a sharp fiscal Real Value of Tax Revenue, Staff Papers, vol. 24, March
adjustment. For nominally fixed fiscal items, more 1977, IMF, pp. 154–167.
typically expenditure items, high inflation induces not 17
Taken at the official exchange rate, the share of foreign
only a reduction of the ratio to GDP, but also a reduc- currency-denominated debt to the total outstanding
tion in real terms. For fiscal items that are indexed to stock of debt was 36 percent by end-2020 (amounting to
42 percent of GDP), mostly held by domestic banks; for
inflation, more typically revenue items, high inflation
illustration, if one were to use instead a simulated exchange
will induce a nominal increase in these items (in rate of LBP 3,555/US$, foreign currency-denominated
principal maintaining real value), thereby partially debt would surge to 57 percent of the total outstanding
offsetting the denominator-led GDP effect. Hence, stock of debt (amounting to 100 percent of GDP).
Recent Macro-Financial Developments 7
BOX 2: LEBANON’ 2021 BUDGET DRAFT PROPOSALa
The 2021 budget draft proposal targets an overall fiscal deficit equivalent to 2.4 percent of GDP, and a primary deficit of 0.8 percent of
GDP.b This marks a consolidation when compared to 2020 estimates for the overall and primary deficits of 4.9 and 2.8 percent of GDP,
respectively (Table 2). Nominally, total revenues and expenditures, and thus the overall fiscal balance, in Budget 2021 are largely unvaried
from those in Budget 2020; only the primary balance deteriorates, signaling the re-allocation of savings from interest payments to primary
spending.
In the 2021 draft, total revenues are projected to continue contracting to 7.3 percent of GDP, compared to 11.9 percent in 2020. The
change is mainly due to a denominator-led effect, resulting from a rise in nominal GDP; in fact, revenues in both 2020 and 2021 budgets
are almost unchanged (LBP 14,176 billion in 2020 and LBP 14,140 billion in 2021): the carry-over of the depressed economic activity in
2020 and the continued sharp fall in 2021 are dragging down revenue collection, notwithstanding the inflation-creep “tax” that is common
in rapid and high inflation environments.
Tax revenues represent the main component of total revenues, budgeted to reach 5.4 percent of GDP in 2021, compared to 8.4
percent of GDP in 2020 budget; in nominal terms the 2021 budget marks a nominal increase in tax revenues—LBP 10,493 billion
compared to LBP 9,966 billion, which is, however, well below the projected 100 percent in inflation for 2021, so tax collection is falling
in real terms.
The budget draft proposes new measures, including a newly introduced “national solidarity tax” and an adjustment to tax on interest profits.
The “national solidarity tax” imposes a (i) 1 percent tax on deposits between US$1 million and US$20 million; (ii) 1.5 percent on deposits
greater than US$20 million but less than US$50 million; and (iii) 2 percent on deposits greater than US$50 million. This tax applies to
equivalent deposits in LBP, calculated at the official exchange rate of LBP 1,507.5/US$, as well as to deposits in other foreign currencies.
Summary of Fiscal Accounts Showing Actual Numbers, WB Estimates and Government
TABLE 2 •
Budgets
2019 2020 2021 2020 2021
Central Government Finance (in percent of GDP) Actual WB estimates Budget Budget draft
Revenue (including grants) 20.6 11.5 6.9 11.3 6.3
o/w. tax revenues 15.5 8.1 4.9 8.4 4.9
Total expenditure and net lending 31.2 16.4 12.2 15.4 8.5
Current 29.9 15.2 10.9 14.9 8.2
o/w interest payment 10.0 2.1 1.4 4.0 1.4
Capital & net lending (excluding foreign financed) 1.3 1.2 1.3 0.4 0.3
Overall balance (deficit (–)) –10.5 –4.9 –5.3 –4.1 –2.2
Primary balance (deficit (–)) –0.5 –2.8 –3.9 –0.1 –0.7
(in LBP bln)
Revenue (including grants) 16,678 13,706 14,750 13,396 13,572
o/w. tax revenues 12,533 9,593 10,500 9,966 10,493
Total expenditure and net lending 25,180 19,492 26,132 18,232 18,259
Current 24,152 18,090 23,356 17,738 17,524
o/w interest payment 8,067 2,504 2,949 4,695 3,106
Capital & net lending (excluding foreign financed) 1,028 1,402 2,776 493 735
Overall balance (deficit (–)) –8,502 –5,786 –11,382 –4,836 –4,687
Primary balance (deficit (–)) –435 –3,282 –8,433 –141 –1,581
(continued on next page)
8 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
BOX 2: LEBANON’ 2021 BUDGET DRAFT PROPOSALa (CONTINUED)
World Bank estimates suggests that, as of December 2020, this tax can generate revenues in the amount of around LBP 165 billion on LBP
deposits, and a US$750 billion on dollar deposits.
Another measure is a tax on interest income paid on bank deposits, certificates of deposit and treasury bills, which currently stands at 10
percent. The adjustment proposed imposes a 30 percent tax on the part of interest rates that exceeds 3 percent and 5 percent on US$
and LBP accounts, respectively. This effectively targets the financial sector’s interest earned from government bonds. Bank customers’
deposits will largely be exempt since interest rates on US$ and LBP deposits have generally fallen below these thresholds.
Expenditures in 2021 budget draft are also projected to drop, to 9.4 percent of GDP, compared to an estimated 16.4 percent of GDP in
2020. In nominal terms, expenditures would rise by a mere 0.2 percent compared to the 2020 budget, but would fall by 84.1 percent in real
terms (when considering the official aggregate Consumer Price Index). Sharp nominal cuts in spending (33.8 percent) arise in debt service
thanks to the Eurobond default and favorable arrangement with BdL. Capital expenditures are projected at 0.4 percent of GDP, marking a
49 percent increase in nominal terms, from those in the 2020 budget.
Notable new expenditure items include: (1) the rescheduling arrears due to the National Social Security Fund (NSSF) to be paid in equal
installments in Treasury Bonds over the next 20 years, with the first such payment due in September 2021; (2) in a bid to encourage
employment of Lebanese in firms, the Government will pay, on their behalf, their due contributions to NSSF for a period of two years, as long
as (i) they are hired before December 31, 2021; (ii) they were either unemployed or entering the job market for the first time; and (iii) each
employee’s salary does not exceed LBP 18 million per year; (3) the amount of bank deposit guaranteed by the National Deposit Guarantee
Institution increases to LBP 300 million, from a previous LBP 75 million; and (4) an obligation for banks to pay new foreign currency
deposits in the currency of the deposit, upon the request of the involved person, in a bid to attract new deposits in foreign currencies.
Relatedly, new deposits in FX will be exempt from tax on interest income, in hope of attracting FX deposits to the banking sector.
Irrespective of new expenditure items, the proposed 2021 budget allocations to primary expenditures constitute a decrease of around 12
percent compared to what are estimated for 2020. We note that our projections suggest the persistence of exceptionally high inflation
rates in 2021. This predisposition can either (i) degrade the proposed budget’s creditability, due to expected social pressures and real
costs resulting from the high inflationary environment; or, if forced through, (ii) further entrench the severe decline in purchasing power for
another year.
a
The budget draft proposal has yet to be discussed by Cabinet and will need to be ratified by Parliament. Hence, measures and numbers in the budget draft are subject to
revisions by Cabinet and Parliament before the budget is passed as a law.
b
It is important to note that these numbers exclude a transfer to EdL in the amount of LBP 1,500 billion (0.7 percent of GDP), since this transfer is considered by the Ministry
of Finance as a treasury advance, and is therefore excluded from Central Government fiscal balance. Since EdL has no capacity and poor prospects of ever repaying this and
previous Treasury advances, under international government accounting standards (GFSM), they should be recorded as transfers.
is rapidly eroding the real value of domestic debt, the current account deficit took place. Specifically, over
sharp depreciation of the currency continues to make 11M-2020, a 44 percent decline in merchandize im-
Lebanon’s sovereign debt burden unsustainable. ports more than offset a 5 percent drop in exports,
shrinking the trade in goods deficit by a drastic 54.8
percent in one year. Net remittances is estimated to
The External Sector have increased from 6.1 percent of GDP in 2019 to
9.7 percent of GDP in 2020. The increase is a result
As foreign financing of the current account (CA) of (i) a sharp decline in US$ GDP (a denominator
deficit came to a sudden stop in late 2019, a effect); (ii) large decreases in remittances outflows,
massive contraction of the CA took place in 2020 as foreign workers in Lebanon suffer from the eco-
driven by a sharp contraction in imports. Fol- nomic contraction; and (iii) some remittances inflows
lowing the de facto introduction of capital controls, incentivized by the well-documented “insurance”18
the sovereign default, capital inflows into Lebanon
stopped. While BdL made use of its limited foreign 18
The “insurance” behavior suggests that diaspora
exchange reserves in 2020, a forced and massive increases remittances back home in case of natural
adjustment/re-sizing of the previously massive disasters.
Recent Macro-Financial Developments 9
Large Shortfalls in Revenues Will
FIGURE 4 • FIGURE 5 • Valuation Effects from Exchange Rate
Induce a Significant Deterioration in Depreciations Will Pressure the Debt-
the Fiscal Position to-GDP Ratio
Fiscal Aggregates (% of GDP) Gross Public Debt
5 100 250
3 90
0 80 200
Percent of GDP (%)
–3 70
–5
US$ Billion
60 150
–8 50
Percent (%)
–10 40 100
–13 30
–15 20 50
–18 10
–20 0 0
–23
1991
1994
1997
2000
2003
2006
2009
2012
2015
2018
2021
–25
–28
External Public Debt (US$ bln) Domestic Public Debt (US$ bln)
1993
1995
1997
1999
2001
2003
2005
2007
2009
2011
2013
2015
2017
2019
External Public Debt, at an Gross Public Debt (US$ bln)
Illustrative Exchange Rate, Gross Public Debt as
Overall Fiscal Balance as a Percentage of GDP (rhs, %) Percentage of GDP (rhs, %)
Primary Fiscal Balance, Excluding Interest Payments External Public Debt, at Official Rate, as a Percentage of GDP (rhs, %)
Sources: Lebanese authorities and WB staff calculations. Sources: Lebanese authorities and WB staff calculations.
and other countercyclical19 behaviors observed in not publish net reserves, which are estimated to be
countries with large diasporas. Nominally, howev- significantly negative.
er, remittances inflows are estimated to have been A high import ratio for the consumption
negatively impacted by an impaired banking sector— basket, along with the shortage of dollars in the
the traditional conduit for remittances—and the CO- market suggest an implicit tradeoff between
VID-19 global impact. Overall, we expect the current (i) importation of goods and services and
account deficit in 2020 to contract, falling by almost (ii) BdL’s stock of foreign exchange reserves. This
10 pps to reach 11 percent of GDP, compared to a compelled authorities to prioritize imports. First, and
medium-term (2013–2019) average of 22.5 percent early on in the crisis, BdL identified a list of highly
of GDP. critical goods (denoted as C1 goods)—namely, fuel,
The sudden stop in capital inflows, medicine and wheat—to be backed by its stock of
coupled with a smaller but still large current foreign exchange reserves at the official exchange
account deficit, has steadily depleted BdL’s rate.20 The Government followed suit in July 2020 with
foreign exchange (FX) reserves (Figure 6). By a list of other critical goods, issued by the Ministry
January 2021, gross FX reserves at BdL (excluding of Economy and Trade (MoET), which BdL agreed
gold reserves) reached $23.5 billion, declining by to back up at the platform exchange rate (LL 3,900
$13.8 billion since end-2019. BdL’s gross position
includes US$5 billion in Lebanese Eurobonds and 19
During economic hardships in the home country,
an unpublished amount lent out to banks since expatriates can also boost transfers back home in
October 2019. Much of the remainder is required support of family.
reserves on banks’ customer FX deposits, which is
20
BdL set up a mechanism via commercial banks whereby
importers of highly critical goods can exchange LBP for
estimated at US$16.7 billion. Critically, BdL’s gross
dollars at the official exchange rate for 85 to 90 percent
position differs widely from its net reserves (i.e., gross of the cost of their imports, while sourcing the remaining
FX reserves at the central bank net of FX liabilities to 15 to 10 percent from the market at the US$ banknote
others); contrary to other central banks, BdL does rate.
10 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
A Steady Depletion in the Gross
FIGURE 6 • Ratios of C1, C2b, Luxury and other
FIGURE 7 •
Foreign Exchange Position at BdL Imports Were Stable until the Period
Leading to the Crisis…
Gross Reserves at BdL
50,000 Critical and Essential Imports
45,000 100%
40,000
90%
35,000
30,000 80%
US$ mln
25,000 70%
% of Total Imports
20,000
60%
15,000
10,000 50%
5,000 40%
0 30%
Jan-13
Apr-13
Jul-13
Oct-13
Jan-14
Apr-14
Jul-14
Oct-14
Jan-15
Apr-15
Jul-15
Oct-15
Jan-16
Apr-16
Jul-16
Oct-16
Jan-17
Apr-17
Jul-17
Oct-17
Jan-18
Apr-18
Jul-18
Oct-18
Jan-19
Apr-19
Jul-19
Oct-19
Jan-20
Apr-20
Jul-20
Oct-20
Jan-21
20%
10%
Foreign Currencies Foreign Securities 0%
Gross FX Reserves Compulsory FX Reserve 2011 2012 2013 2014 2015 2016 2017 2018 2019
Sources: BdL and WB staff calculations. Critical 1 Critical 2b Luxury Other
Note: Compulsory FX reserves are World Bank estimates based on published data, and
a 15 percent required reserve ratio on FX deposits in commercial banks. Sources: Customs, MOET, BdL and WB staff calculations.
per US$). This list was updated in November 2020 to FIGURE 8 • …when Ratios of C1 and C2b Imports
include almost half the number of items that were in Rose at the Expense of those for
Luxury and Other Goods.
the original list. Hereafter, the original list will be called
C2a while the latest will be called C2b. Critical and Essential Imports
An examination of historical and recent 100%
trends for C1 and C2b imports are revealing. 90%
The shares of C1, C2b, luxury and other imports to 80%
total imports are presented annually from 2011 to 70%
% of Total Imports
2019 and monthly over 2019–20 in Figures 7 and 8, 60%
respectively. Figure 7 illustrates relative stability in the 50%
respective shares until 2019. Meanwhile, as foreign 40%
exchange constraints became binding, the ratios 30%
of C1 imports, and to a lesser extent C2a and C2b 20%
imports, rose at the expense of those for luxury and 10%
other goods (Figure 8). The average monthly values 0%
over the January 2019 to October 2020 period for C2a
Jan-19
Feb-19
Mar-19
Apr-19
May-19
Jun-19
Jul-19
Aug-19
Sep-19
Oct-19
Nov-19
Dec-19
Jan-20
Feb-20
Mar-20
Apr-20
May-20
Jun-20
Jul-20
Aug-20
Sep-20
Oct-20
and C2b are US$180 and US$161 million, indicating
unsubstantial savings by the second MoET list. Critical 1 Critical 2b Luxury Other
Importers from across the economy com-
Sources: CAS and WB staff calculations.
pete for access to these FX-backed facilities.
However, less transparent demand is also well-docu-
mented in countries with capital controls and multiple
exchange rate systems; specifically, corruption and Money and Banking
misclassification of imports to benefit from cheap
foreign exchange. As importers adapt to capital con- Monetary and financial turmoil are driving crisis
trols and more depreciated black-market rates, this conditions, specifically through interactions
incentive will grow. between the exchange rate, narrow money and
Recent Macro-Financial Developments 11
inflation. Acute exchange market pressures in As inflation has surged during that same period,
Lebanese markets are reflected by heavy fluctuations real interest rates in the country are highly negative
in the US$ banknote exchange rate,21 which breached across the board.
LL 15,000/US$, before falling back down. This is Since the eruption of the financial crisis,
within the context of a multiple exchange rate system, BdL has been almost an exclusive policy maker,
which includes the official exchange (LBP 1,507.5/ with the exception of a brief period in which
US$) as well BdL’s platform rate set at LBP 3,900/ Government defaulted on its Eurobond obliga-
US$. Overall, the World Bank calculated Average tions and unsuccessfully proposed its Financial
Exchange Rate (AER)22 depreciated by 129 percent Recovery Plan. The Fall 2020 LEM listed in detail
in 2020 (Figure 9). the slew of BdL circulars, which formalized BdL’s
Limited economic utility for electronic dol- crisis management strategy. The main BdL policy
lars, along with scarcity of dollar banknotes, and
23
initiatives/updates since then include (i) a new
minimum incentives to save in LBP, all rendered financial operation; (ii) an announcement by BdL
the economy heavily cash-based in local currency. to allow commercial banks to conduct currency
In 2020, the stock of currency in circulation increased exchanges at the market rate; and (iii) the expiration
by 197 percent (yoy), even as M2 increased by a of a deadline for commercial banks to meet Circular
mere 6.3 percent, while M3 declined by 1.4 percent. 154 provisions.
The latter two money supply measures reflect the In March 2021, BdL announced a new
deleveraging that is going in the financial sector (see financial operation. Details of this operation and
below). an illustrative example are presented in Box 4. In
Exchange rate pass through effects on this most recent operation, BdL offers banks FX in
prices have resulted in surging inflation, hitting exchange for LBP, increasing its liabilities in FX and
the poorest and most vulnerable the most. The
12-month inflation rate has risen steadily and sharply
from 10 percent in January 2020, to 89.7 percent in 21
The US$ banknote market has become a main supply
June 2020, 120 percent in August 2020, and most channel for dollars for both real and financial activity,
recently, to 157.9 percent in March 2021. Inflation is as commercial banks heavily restricted withdrawals and
a highly regressive tax, disproportionally affecting transfers of customers’ dollar deposits.
the poor and vulnerable (Box 3), and more generally, 22
The AER is derived by applying consumption-based
people living on fixed income like pensioners. This is weights on the official, the platform and the US$ banknote
exchange rates. For a detailed derivation please refer to:
especially so in Lebanon’s case where basic items of
World Bank (2020), The Lebanon Economic Monitor:
the consumption basket are primary drivers of overall The Deliberate Depression, Fall 2020. Since then, we
inflation. In fact, the main contributors of inflation make adjusted the AER per the following:
are food and non-alcoholic beverages, followed by We account for a reduced MoET list of subsidized goods
clothing and footwear, and then furnishing and house- starting in November 2020.
hold equipment (Figure 10); prices for these basic Beginning in August 2020, 75 percent of the non-
imported services is linked to the official exchange
consumption items have surged by 254.3, 289.8 and
rate (compared to 100 percent prior), while 25 percent
389.7 percent, respectively, in 2020. is linked to the platform rate. The reason being is
The severe restrictions on capital outflows that, around that time, certain prominent hospitals
have given the monetary authorities room and universities announced that they will start billing
to lower interest rates. From October 2019 to according to the BdL platform rate.
December 2020, BdL lowered interest rates on
23
This refers to dollar deposits from prior October 2019,
which are subject to strict capital controls and can
banks’ LBP and dollar deposits by 639 and 567
generally be withdrawn only in LBP at the platform rate
basis points (bps), respectively. Banks’ lending rates (LBP 3,900/US$), a significant haircut compared to the
in LBP and US$ have mirrored this effect, falling by value of the dollar being traded in the parallel banknote
342 and 332 bps, respectively over the same period. market.
12 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
FIGURE 9 • A Sharp Depreciation in the Exchange Inflation in Basic Items is a Key
FIGURE 10 •
Rate along with Surging Inflation and Driver of Overall Inflation, Hurting
Narrow Money the Poor and the Middle Class
Exchange Rate-Currency in Circulation-Inflation Contributions to Overall Inflation in 2020
5.0 4.0 100.0
4.5
80.0
4.0
3.0 60.0
3.5
Percent
3.0
Percent
Percent
40.0
2.5 2.0
2.0 20.0
1.5 0.0
1.0
1.0
0.5 –20.0
0 0
Headline Inflation growth Food & Non-alcoholic Beverages
Aug-19
Sep-19
Oct-19
Nov-19
Dec-19
Jan-20
Feb-20
Mar-20
Apr-20
May-20
Jun-20
Jul-20
Aug-20
Sep-20
Oct-20
Nov-20
Dec-20
Jan-21
Feb-21
Mar-21
Clothing & Footwear Furnishings, Household Equipment
Transportation Alcoholic Beverages & Tobacco
Communication Water, Electricity, Gas, & Other Fuels
US$ Bank Note Rate WB Average Exchange Rate Owned Occupied Health
Inflation Rate (rhs) Currency in Circulation (rhs) Actual Rent Education Other
Sources: CAS, BdL and WB staff calculations. Sources: CAS and WB staff calculations.
BOX 3: IMPACT OF CRISES ON POVERTYa
The multiplicity of crises affecting Lebanon is likely to have Food Expenditure Shares by Deciles
FIGURE 11 •
an escalating impact on poverty. After remaining in the single for Lebanese Nationals
digits for over a decade, average annual inflation ballooned to
84.3 percent in 2020, while average food inflation alone grew 40%
37.4%
by a record 254 percent over 2020. The annual rates do not
fully reflect the rapid rise in monthly inflation wherein the year- 35%
on-year inflation rate reached 146 percent in December 2020
Food Share
with a corresponding food inflation rate of over 400 percent. 30%
The growth in food inflation is particularly concerning, as food
consumption forms a larger proportion of the expenses incurred 25%
by poorer households (Figure 11).
20%
The deteriorating purchasing power is causing households 20.2%
to struggle in making ends meet. Phone surveys conducted 15%
during November and December 2020 found that 41 percent 1 2 3 4 5 6 7 8 9 10
of households reported challenges in accessing food and other Per-Capita Expenditure Deciles (HBS, 2012)
basic needs. The share of households facing difficulties in
Source: Staff calculation using HBS 2012.
accessing health care was at 36 percent, up from 25 percent
in July-August 2020. Unemployment rate also rose among the
respondents, from 28 percent in February 2020 (pre-COVID) to nearly 40 percent in the Nov-Dec period.
The absence of timely and relevant data makes it very difficult to reasonably predict the state of poverty in the country. The last household
budget survey was completed in 2011–2012 under conditions very different from the state of the country today. The pandemic has further
heightened the vulnerability of the poor whose living, working and health conditions place them at greater risk compared to wealthier
segments of the population. Under highly inflationary conditions, poorer households have limited means of preserving their purchasing
power and are likely to resort to a variety of last-resort coping strategies, including ceasing their medications, borrowing at extortionary
rates and selling of assets to simply put food on the table. Tentative projections using the older data suggest that well over half the
population is likely to be under the national poverty line.
The economic crisis and resulting rising in poverty raise an urgent need for social assistance. High levels of poverty can have a long-lasting
impact on Lebanon’s human development and increases vulnerabilities across the lifecycle. Adequate social assistance will therefore be critical
both in the short term to provide emergency relief, and in the medium-long term to improve resilience to shocks among vulnerable Lebanese.
a
This box has been prepared by Ganesh Kumar Seshan (Senior Economist, EMNPV), Bilal Malaeb (Economist, EMNPV), Fahmina Rahman Dutta (Social Protection Specialist,
HMNSP) and Haneen Ismail Sayed (Lead Operations Officer, HMNSP), under the guidance of Johannes G. Hoogeveen (Practice Manager, EMNPV).
Recent Macro-Financial Developments 13
BOX 4: THE MOST RECENT FINANCIAL OPERATION BY BDL WITH COMMERCIAL BANKS
The most recent financial operation employed by BdL involves the following steps:
1. Banks discount medium- to long-term LBP Time Deposits (TDs) and/or Certificates of Deposits (CDs) at a discount rate of yield plus
up to 1 percent.
2. With the proceeds from the above operation, banks buy US$ at LBP 1514/US$ from BdL and deposit this amount at their US$
current account with BdL, receiving zero interest rate.
3. Banks are also required to unwind LBP financial engineering structures at least equivalent to the US$ amount bought (at least 1 to 1
ratio). The unwinding process consists of discounting at par LBP TDs as well as their linked LBP Repos.
4. Banks would sign an undertaking to sell back to BdL the amount US$ bought at the electronic platform rate upon the first demand
from BdL, through discounting US$ TDs and CDs.
5. The amount of US$ the bank is committed to sell back to BdL will be reduced by the amount of US$ the bank sells to BdL under
Circular 151, from January 2021 until the call back date (as per Central Decision Council amendment).
For illustrative purposes, we present a simplified example.
Steps 1–3 are Spot Transactions
Suppose step 1 involves discounting LBP 1,514,000 in face value divided evenly between TDs and CDs. Long term average interest rates
for LBP TDs and CDs are around 3.5 and 8.5 percent, respectively. Hence the average yield for step 1 is 6 percent (= 0.5*3.5 + 0.5*8.5).
Moreover, we assume the average maturity for these TDs and CDs is 10 years. Hence, the present value for this operation is calculated by
discounting (i) coupon payments, which we are assuming annual; and (ii) the face value payment at maturity, such that:
6%*(1,514,000) 1,514,000
∑
10
+ = LBP 1,514,000 = US $ 1000 (at LBP 1,514 /US $)
n=1
(1+ 6%)n (1+ 6%)n
Hence, banks use the proceeds to purchase US$1,000 from BdL, which is deposited at a non-yielding current account with the central
bank.
Further, per step 3, banks unwind LBP 1,514,000 of financial engineering TDs and repos by discounting at par. These TDs were created
as part of BdL’s financial engineering operations in which BdL lent banks LBP at 2 percent annual rates on the condition they were placed
as TDs with BdL earning 10.5 percent annually. Hence, the net return on these TDs for banks was 8.5 percent annually. Hence, as earlier,
we discount (i) the annual net return on these TDs; and (ii) the repayment of loan principal at maturity; this generates a present value of
LBP 1,514,000.
Steps 4 and 5 are Future Transactions
At future dates, banks are obliged to sell back US$ 1000 minus whatever dollars they sell under Circular 151, herein denoted by X. The
dollar value is exchanged to LBP at the platform rate of the time, denoted by E, and which is currently at LBP 3,900/US$.
The combination of steps constitutes a call option for BdL to buy back the US dollars, albeit, with undefined terms, due to uncertainty over
key parameters, such as X, E, and time to expiration of the call option. Hence, for simplification purposes in this illustrative example, we
assume away the accounting and pricing implication of the call option for BdL, and symmetrically for Banks.
Hence, according to the numbers used in this illustrative example, a balance sheet analysis suggests (with spot transactions in light green
background and future transactions in white):
BdL Banks
Assets Liabilities Assets Liabilities
LBP –1,514,000 LBP –1,514,000
US$ +1000 US$ +1000
LBP –1,514,000 LBP –1,514,000 LBP –1,514,000 LBP –1,514,000
US$ –(1000-X) US$ –(1000-X)
LBP +[($1000-$X)*E] LBP +[($1000-$X)*E]
(continued on next page)
14 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
BOX 4: THE MOST RECENT FINANCIAL OPERATION BY BDL WITH COMMERCIAL BANKS (CONTINUED)
For BdL:
At spot
• BdL’s liabilities are lower by LBP 3,028,000 (=1,514,000+1,514,000) but higher by US$1000.
• BdL’s assets are lower by LBP 1,514,000
• Net impact on BdL’s capital would thus be
• Positive (i.e., capital increases) if the exchange rate is lower than LBP 1,514/US$
• Negative (i.e., capital decreases) if the exchange rate is higher than LBP 1,514/US$
Over time
• BdL’s liabilities are lower by LBP [1,514,000+1,514,000-($1000-$X)*E] but higher by US$X.
• BdL’s assets are lower by LBP 1,514,000
• Net impact on BdL’s capital would thus be
• Positive (ie. capital increases) if the exchange rate is lower than
LBP (1,514,000)−(1000−X)*E per US$
X
• Negative (ie. capital decreases) if the exchange rate is more than
LBP (1,514,000)−(1000−X)*E per US$
X
For Banks:
At spot
• Bank’s assets are lower by LBP 3,028,000 but higher by US$1000.
• Banks’ liabilities are lower by LBP 1,514,000
• Net impact on banks’ capital would thus be
• Negative (i.e., capital decreases) if the exchange rate is lower than LBP 1,514/US$
• Positive (i.e., capital increases) if the exchange rate is greater than LBP 1,514/US$
Over time
• Banks’ assets are lower by LBP [1,514,000+1,514,000-($1000-$X)*E] but higher by US$X.
• Banks’ liabilities are lower by LBP 1,514,000
• Net impact on Banks’ capital would thus be
• Negative (ie. capital decreases) if the exchange rate is lower than
• LBP (1,514,000)−(1000−X)*E per US$
X
• Positive (ie. capital increases) if the exchange rate is more than
• LBP (1,514,000)−(1000−X)*E per US$
X
From an income perspective, and according to the numbers used in this illustrative example, in the absence of this operation, banks (BdL)
would be earning (paying) a return of
• an average of 6 percent from LBP TDs and CDs;
• an average of 8.5 percent from LBP TDs and repos;
for a total of 14.5 percent on LBP instruments. Meanwhile, in this operation
• 14.5 percent earnings on LBP instruments for banks are eliminated; hence BdL saves 14.5 percent;
• BdL does not pay interest on Banks’ dollar deposits;
Overall, from an income perspective, and according to the numbers used in this illustrative example, BdL saves (banks lose) 14.5 percent
annually.
Recent Macro-Financial Developments 15
decreasing it in LBP, with the exchange occurring • This would be based on the current “Sayrafa”
at the official exchange rate. It then compels banks platform, which is now set at LBP 3,900/US$.28
to sell this FX back to BdL (whether directly or via Indeed, after the Baabda Palace announcement,
the ongoing Lirafication scheme24) at the platform BdL sent a “Request” to banks and exchange
rate, which is currently set at LBP 3,900/US$. With dealers to register on Sayrafa.
the caveat that a comprehensive assessment of this • This is not meant to completely replace the
operation is more accurate from a hindsight per- parallel market.
spective, as it is able to better capture interactions • It is not clear how close the Sayrafa rate would be
and incentives that might not be clear at this point, to US$ banknote rate; or how responsive it would
we present some preliminary observations: be to changes in the banknote rate.29
• Transactions would be restricted to exchange of
• As a spot transaction, the operation will cancel currency notes only (LBP and US$).
out LBP assets and liabilities that resulted from • On the demand side (of FX): the announced
BdL’s previous financial operations.25 intention is to allow access to only those engaged
• As a spot transaction, BdL will incur an FX liability in “legitimate economic transactions”, (i.e.,
to banks and hence worsen its net international traders/industrialists/businesses, for trade and
reserve (NIR) position. Since the FX offered by business activities).
BdL to banks is not real liquidity, or “fresh dollars,”
its use is limited to closing banks’ FX positions
with BdL (i.e., banks’ deleveraging schemes such 24
BdL has regulated that dollar deposits that existed
as Lirafication). prior to October 2019 can be withdrawn only in LBP
• As a spot transaction, this operation will worsen at the platform rate, thereby effecting Lirafication and
(improve) BdL’s (banks’) capital position. a haircut (up to 70 percent) on these deposits. More
• Over time, all things equal, this operation will formally, Circular 151, as part of a sequence of circulars
(148, 151, 549, 565), allowed the withdrawal of pre-
interact with Circular 151, with a likelihood that,
crisis deposits at exchange rates that are higher than
on the net, BdL will increase its FX liabilities. the official rate, but lower than the US$ banknote rate.
Commercial banks’ deleveraging, as incentivized This rate eventually settled at the platform rate, which is
by Circular 151,26 would be occurring currently set at LBP 3,900/US$.
irrespective of this operation and will not be 25
In previous financial engineering schemes, BdL incurred
further incentivized by it. Hence Step 5 in Box 4. LBP liabilities, which were offered as an incentive to
commercial banks to purchase Eurobonds from BdL,
The most recent financial operation by BdL with
allowing BdL to raise its gross foreign exchange reserves
commercial banks27 suggests that this operation by tapping into commercial banks’ foreign assets.
would prevent a larger overall reduction to BdL 26
These deleveraging schemes have resulted in reductions
FX liabilities that would have occurred in the in customers FX deposits with banks, which effected
absence of this operation. commensurate reductions in banks’ FX deposits with
• Over time, and after banks have sold back this BdL, and hence, reductions in BdL’s FX liabilities, and
improvements in BdL’s NIR.
FX liquidity to BdL, BdL would have replaced the 27
Step 5 states: the amount of US$ the bank is committed
LBP liability with currency in circulation. to sell back to BdL will be reduced by the amount of
US$ the bank sells on an annual basis to BdL under
Also in March 2021, a Baabda Presidential Circular 151.
Palace announcement stated that BdL will start 28
In fact, the Sayrafa platform was initially set up by Circular
allowing commercial banks to conduct currency 149 (April 3, 2020), in which BdL announced a special
unit comprising BdL, banks and exchange bureaus that
transactions similar to legal exchange dealers, at
exchanges foreign currencies. The Sayrafa was intended
a rate to be set by the Central Bank’s electronic as electronic board that publishes exchange rates.
platform. While information on this has been scarce, 29
Minister of Finance, in an interview to a TV channel,
notable factors include: mentioned the rate LL10,000/US$.
16 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
• On the supply of FX: the sources for US$ FIGURE 12 • Heavy Deleveraging of Assets (Private
banknotes to be exchanged through this platform Loans) and Liabilities (Private
Deposits) in the Financial Sector
are not clear. Options or combinations thereof
include: Bank’s Deposits and Loans (% yoy change)
• Commercial banks’ own FX liquidity. 30
20
• A new set-up where exporters are incentivized
10
to repatriate FX dollars they have been
Percent (%)
0
depositing in foreign banks. –10
• BdL’s own FX reserves. –20
–30
Ultimately, if BdL wants to control/influence –40
–50
the US$ banknote rate, then it has to find a suf-
Jan-15
Apr-15
Jul-15
Oct-5
Jan-16
Apr-16
Jul-16
Oct-16
Jan-17
Apr-17
Jul-17
Oct-17
Jan-18
Apr-18
Jul-18
Oct-18
Jan-19
Apr-19
Jul-19
Oct-19
Jan-20
Apr-20
Jul-20
Oct-20
Jan-21
ficient supply of FX to inject at a rate lower than
the banknote market rate, so as to attract buyers LBP Credit Outstanding FX Credit Outstanding
of dollars. Essentially, this would act as a more to Private Sector to Private Sector
Private Sector Deposits in LBP Private Sector Deposits in FX
traditional FX intervention in the market in defense
of the exchange rate. In such a case, it can be used Sources: CAS and WB staff calculations.
as a monetary tool to mop up currency in circulation
from the market, as more orthodox monetary tools are
ineffective.30 However, even if currency in circulation Conditions in the financial sector continue
is reduced, this operation would be inconsistent with to deteriorate, while a consensus among key
other policies (Lirafication and monetization of the stakeholders on the burden-sharing of losses has
deficit), making the overall monetary stance of BdL proven to be elusive, and is further complicated
unclear. This can weaken the impact of a reduction in by the lack of progress in government formation.
currency in circulation on inflation. A larger question Customer deposits at commercial banks declined
would be, in absence of a more comprehensive and by US$19.8 billion (or 12.6 percent) in 2020, further
sustainable solution, would BdL/banks spend from deepening the liquidity shortage. Liquidity needs in
their very valuable (i.e., limited and dwindling) FX the banking system have been met mainly through
resources? deleveraging and reduction in net foreign assets.
Officially, BdL did not extend the deadline for Deleveraging had resulted in banks’ domestic credit
Lebanese commercial banks to meet new param- portfolio shrinking by US$12.1 billion (or 27.5 percent)
eters set by BdL via Circular 154; that deadline during the same period (Figure 12), mainly in foreign
was February 28, 2021. Key stipulations of Circular currency loans, although it partially reflected cus-
154 include: raising bank capital by 20 percent; banks tomers settling loans backed by real estate collateral,
to place funds in correspondent banks amounting to using deposits locked up in bank accounts.
a minimum of 3 percent of customers’ FX deposits; Lending from BdL has allowed Lebanese
banks to convince customers to repatriate 15 percent commercial banks to pay off liabilities to
of deposit outflows above US$500,000 since end-
2017; banks’ shareholders and politically exposed
persons (PEPs) to repatriate 30 percent of deposit 30
In more conventional monetary and financial conditions,
outflows above US$500,000 since end-2017. BdL and increasing policy rates can be used to effect monetary
the Banking Control Commission are reported to be tightening. However, this is no longer an option as the
banking system is dysfunctional and the monetary
currently undertaking a bank by bank examination
transmission channel impaired. In addition, under
of compliance. There is a lack of clarity on follow up capital controls, monetary authority has more flexibility
and next steps, and there have been no subsequent in not increasing interest rates as capital is captive even
initiatives since the passing of the deadline. with hugely negative real interest rates.
Recent Macro-Financial Developments 17
correspondent banks in order to retain linkages FIGURE 13 • A Steady and Sharp Deterioration in
to the global financial system. As of December Credit Performance as Measured by
NPL Ratio for Banks
2020, commercial banks’ placements in and liabilities
for non-resident financial institutions (FIs) amounted NPLs by Sector
to US$4.7 and US$6.6 billion, respectively, compared Total Credit Portfolio
to US$6.8 and US$8.8 billion in December 2019.
Consumption
Foreign correspondent banks have significantly Housing
tightened conditions and reduced lines to Lebanese Financial Intermediation
banks. As a condition on continuing to transact via Retail Trade
Wholesale Trade
correspondent banks, commercial banks have had to Contracting and Construction
pay down liabilities to these banks, partially financed Processing Industries
by lending from BdL. This allowed a marginal 0% 10% 20% 30% 40% 50% 60%
improvement in the net position of commercial banks Dec-19 Dec-20
at non-resident FIs, which nonetheless remained
Sources: BdL and WB staff calculations.
negative at –US$1.8 billion in end-2020, compared
to –US$2.1 billion in end-2019.
Large losses associated with the sover-
eign exposures of the banking sector remain 19.6 percent a year earlier. NPL ratio for construction,
unresolved. Exposures to the sovereign amount to processing industries and wholesale and retail trade
70 percent of total banking assets or roughly 250 reached 51 percent, 50 percent and 43 percent,
percent of 2019 GDP. These include US$110 billion respectively (Figure 13). Provisioning coverage was
in BdL instruments, US$10 billion in Lebanese gov- at just over 50 percent as of end-2020. Continued
ernment Eurobonds and US$11 billion in domestic deterioration in the quality of the remaining credit
government securities at the official exchange rate. portfolio (US$40 billion at the official exchange rate
The credit portfolio of the banking sector and 61 percent denominated in US$) would be
has substantially deteriorated during recent expected, given the lack of progress in government
months. The non-performing loan (NPL) ratio—that formation and necessary restructuring and reform.
is, gross NPLs including unearned interests as a Losses arising from the rapid deterioration of assets
percentage to total loans—stood at 33.0 percent (37.7 quality, although paling in comparison to the losses
percent for FX loans) as of end-2020, compared to from sovereign exposures, will need to be addressed
13.3 percent at end-June 2019 before the crisis and in a comprehensive balance sheet cleanup.
18 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
3
GLOBAL CRISES
COMPARATORS: LOOKING
FOR THE MINIMUM
W
e continue to monitor the Lebanon of crisis or the year of the financial crisis. There
financial crisis in the context of Global are cases where the downturns start earlier but
Crisis Comparators.31 In this LEM, we these are less frequent.
compare the Lebanon crisis with the most severe • A substantial number of these crises are not
global crises episodes as observed by Reinhart “pure” banking crises in that these unfold
and Rogoff (2014), henceforth referred to as R&R.32 alongside a currency crash and often involve a
Based on the most extensive financial crises database sovereign default as well.
available,33 R&R calculate a crisis severity index (CSI)
for a sample of 100 crisis episodes over the 1857– 31
The Fall 2020 LEM, entitled The Deliberate Depression,
2013 period. The CSI is computed based on (i) the
compares Lebanon’s macroeconomic fundamentals in
depth of the crisis episode—the peak-to-trough decline the lead-up to the crisis to two groups of global crises
in real GDP per capita, and (ii) its duration—the number comparators: the Asian crisis countries of 1997–98, and
of years it takes to reach the prior peak in real per a more eclectic set of crises that occurred in the 2000’s
capita income. Specifically, for each crisis episode: [Argentina (2001), Greece (2008), Ireland (2008),
Iceland (2008) and Cyprus (2012)]. We conclude that,
leading up to the crisis point, Lebanon’s macroeconomic
CSI = −1*(peak-to-trough percentage change) +
fundamentals were weak compared to these global
number of years from peak to recovery to prior peak crises comparators, suggesting that the adjustment
process will be more painful and will take longer, even
R&R’s 25 most severe crises and associated with optimal policy measures in place.
results are presented in Table 3. Some notable find- 32
Reinhart, Carmen M. and Kenneth S. Rogoff (2014),
ings include: Recovery from Financial Crises: Evidence from 100
Episodes, American Economic Review: Papers &
Proceedings 2014, 104(5): 50–55.
• A typical recurring pattern across time and space 33
Reinhart, Carmen M., and Kenneth S. Rogoff (2009),
in these cycles is one in which economic activity This Time Is Different: Eight Centuries of Financial Folly.
reaches a peak either the year before the onset Princeton: Princeton University Press.
19
Crisis Severity: Percent Decline in Per Capita GDP, Duration of Contraction, and Years to Full
TABLE 3 •
Recovery in 25 of the Worst Systemic Banking Crises, 1857–2013 (Reinhart and Rogoff, 2014)
Peak to Trough Peak to Trough Peak to Recovery Double Dip
Rank Year Country % change # years # years CSI Yes or No
1 1926 Chile –46.6 3 16 62.6 Y
2 1931 Spain (civil war) –34.6 9 26 60.6 Y
3 1983 Peru –32 11 25 57 Y
4 1931 Uruguay –36.1 3 17 53.1 Y
5 1893 Australia –28 8 20 48 Y
6 1929 Mexico –31.1 6 16 47.1 Y
7 1921 Italy –25.5 3 21 46.5 Y
8 1890 Brazil –21.7 4 21 42.7 Y
9 1923 Canada –30.1 4 10 40.1 N
10 1890 Uruguay –21 2 19 40 Y
11 1981 Philippines –18.8 3 21 39.8 Y
12 1980/1985 Argentina –21.8 11 18 39.8 Y
13 1929 India –8.2 9 31 39.2 Y
14 1929/1933 US –28.6 4 10 38.6 Y
15 1994 Venezuela –24.2 11 14 38.2 Y
16 1939 Netherlands –16 6 21 37 Y
17 2009 Greece –24 6 12 36 Ya
18 1931/1934 Argentina –19.4 3 15 34.4 Y
19 1931 Poland –24.9 4 9 33.9 N
20 1929/1931 Austria –23.4 4 10 33.4 N
21 1981 Mexico –14.1 7 17 31.1 Y
22 1920 UK –18.7 3 11 29.7 Y
23 2001 Argentina –20.9 4 8 28.9 N
24 1980 Chile –18.9 2 8 26.9 N
25 2002 Uruguay –18.9 4 8 26.9 N
Average –24.3 5 16 40.5
a
This is listed as N in Reinhart and Rogoff (2014), since until its publication, Greece had not yet experienced its double dip which subsequently occurred in 2016.
• Out of the 100 episodes, 63 were in advanced Lebanon’s real GDP per capita has been on a
economies and 37 in emerging economies. continuous decline since the onset of the Syria war
• While emerging market peak-to-trough average in 2011, with a much sharper drop commencing in
output declines are about 5 percent larger than 2018 (Figure 14). Prior to 2018, Lebanon’s per capita
those in the advanced economies, they are not GDP declined as a direct consequence of the war in
statistically significantly more protracted.
• A double dip34 is observed for 45 percent of the whole 34
A double dip is when the post-trough, but pre-recovery
sample and two-thirds of the most severe crises. growth is interrupted with a real GDP contraction.
20 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
neighboring Syria; both real economic activity slowed FIGURE 14 • Lebanon’s Real GDP is a More
appreciably,35 and population increased significantly Accurate Reference Point for the
Start of the Financial Crisis than
due to the refugee influx.36 R&R calculations of crisis
Real GDP/capita
depth and duration, and hence, the CSI, depend on the
identification of peak year—the year in which the per Real GDP per Capita
15%
capita real GDP reached a pre-crisis peak. However,
10%
due to the impact of the Syrian crisis, and in order 5%
to more accurately gauge the depth and duration of
Percent (%)
0%
the Lebanon financial crisis, it would be unbalanced –5%
to identify the peak year in Lebanon by the pre-crisis –10%
peak in real GDP per capita as in R&R. Instead for the –15%
–20%
Lebanon episode, we identify the peak year by the pre-
–25%
crisis peak in real GDP, which is 2017.
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
The Lebanon financial crisis is likely to rank
Real GDP Growth Population Growth
in the top 10, possibly three, most severe crises Real GDP/Capita Growth
episodes globally since 1900. In order to compute
Sources: CAS and WB staff calculations.
a CSI for the Lebanon financial crisis, we make
assumptions commensurate to (relatively) good and
bad case scenarios. In the good case scenario, we
assume: (i) real GDP per capita trough occurs in 2021 Annex C for each of these episodes. To the extent
at the World Bank projected growth rate (–9.5 per- data is available, we plot each macroeconomic indi-
cent), inducing an overall –35.1 percent peak (2017) cator for the G8 plus Lebanon over the years leading
to trough (2021) change in real GDP per capita; and to the crisis point and observe dynamics in years that
that (ii) it takes 12 years for Lebanon to recover to follow.37
2017 real income per capita levels, the average crisis
duration of those cases ranking 16–25 in Table 3. The
good case CSI would be 47.1, ranking the Lebanon Per Capita Output
episode 6th after Australia (1893) and aligned with
Mexico (1929) in Table 3. In the bad case scenario, While Lebanon’s real GDP per capita has faltered
we assume: (i) real GDP trough occurs in 2022 at an since 2011, 2018 marks the beginning of a much
additional 5 percent annual contraction in real GDP, sharper decline. Only Argentina (01) and Uruguay
resulting in an overall –38.6 percent peak (2017) to (02) share Lebanon’s consistent negative growth in
trough (2022) change in real GDP per capita; and
that (ii) it takes 19 years for Lebanon to recover to
2017 real GDP levels, the average crisis duration of
35
Annual real GDP growth fell from a pre-Syria war average
the top 10 cases in R&R. Bad case CSI would be 57.6, of 4–5 percent to 1.8 percent over the 2011–2017 period.
36
Total population expanded by 27.3 percent from 2010
ranking the Lebanon financial crisis third after Chile
to 2017.
(1926) and the Spanish civil war. 37
In these charts, the indicator is plotted from 3 years prior
We also cross-compare key macroeco- to crisis year (t–3), to 5 (or 6) years post-crisis (t+5 or t+6),
nomic indicators for Lebanon with those for R&R’s of course going through crisis year (t). In such a way,
relatively more recent episodes. Specifically, we even when crisis years differ (say 2009 for Greece and
compare Lebanon to the following R&R episodes, 2001 for Argentina), plotting in reference to a crisis point
rather than the calendar year superimposes the same
henceforth referred to as G8: Chile (1980), Argentina
indicator for Lebanon with global crises comparators
(1980), Philippines (1981), Mexico (1981), Venezuela on one chart. This allows us to cross-compare how the
(1994), Argentina (2001), Uruguay (2002) and Greece macro indicator developed as the crisis is approached,
(2009). A summary of crisis events is presented in and how it evolved afterwards.
Global Crises Comparators: Looking for the Minimum 21
FIGURE 15 • Real GDP Per Capita for G8 Plus FIGURE 16 • Real GDP Per Capita for G8 Plus
Lebanon Lebanon
Real GDP per Capita Growth Real GDP per Capita Growth
10 10
5 5
0 0
Percent (%)
Percent (%)
–5 –5
–10 –10
–15 –15
–20 –20
t–3 t–2 t–1 t t+1 t+2 t+3 t+4 t+5 t–3 t–2 t–1 t t+1 t+2 t+3 t+4 t+5
Chile '80 Argentina '80 Philippines '81 Venezuela '94 Argentina '01 Uruguay '02
Mexico '81 Lebanon '19 Greece '09 Lebanon '19
Sources: WDI, CAS and WB staff calculations. Sources: WDI, CAS and WB staff calculations.
per capita real GDP in the period leading to the crisis by the Lebanon financial crisis. To the extent
point (Figures 15 and 16). While the contraction in data allows, we examine the trajectories of both the
real GDP commenced in 2018, it accelerated sharply exchange rate and inflation in G8 plus Lebanon.
in 2020. However, Lebanon stands out in magnitude Crisis-related currency depreciations were most
of the contraction at time t and t+1. Further, the severe in Argentina (80) (Figure 17), and to a lesser
protracted nature of the crises in the G8 countries extent in Mexico (81), Venezuela (94) and Argentina
is reflected in the difficulty for these economies to (01) (Figure 18). Exchange rate depreciation in
regain growth in their per capita GDP. Only Argentina Lebanon has so far been comparable to the latter
(01) and Uruguay (02) amongst the G8 were able to three. Loss in value for local currencies was relatively
regain consistent per capita growth by t+2. moderate in Chile (80) and the Philippines (81),
The contraction in Lebanon’s real GDP and mild in Uruguay (02). Meanwhile, Greece (09)
per capita is already worse than any of the had successfully retained membership of the Euro
G8’s Peak to Trough changes. The contraction monetary area.
in Lebanon’s real GDP per capita from its level in Inflation in Lebanon, long subdued since
peak year 2017 had already reached an estimated the early nineties in reflection of exchange rate
27.9 percent by 2020, and is projected to be 35.1 stability,38 has come to resemble the worst of
percent by 2021. This is significantly larger than G8 G8 episodes. Exchange rate pass-through effects
Peak to Trough changes in per capita GDP: –24.2 on prices have implied a correlation between cur-
percent for Venezuela (94); –24 percent for Greece rency depreciations and inflation rates. Due to data
(09); –21.8 percent for Argentina (80); –20.9 percent limitations, we examine inflation in a subsample of
for Argentina (01); –18.9 percent for each of Chile the G8 episodes, namely, Chile (80), Philippines (81),
(81) and Uruguay (02); –18.8 percent for Philippines Mexico (81), Uruguay (02) and Greece (09). Among
(81); and –14.1 percent for Mexico (81) (Table 3). this subsample, inflation was most severe in Mexico
(81), relatively moderate in Chile (80) and Philippines
(81), and mild in Uruguay (02) and Greece (09). This
Depreciation-Inflation
A potent depreciation-inflation dynamic is a key 38
This is the case since the last exchange rate collapse,
driver of macroeconomic instability for half of which saw the currency depreciate by over 400 percent
the G8 episodes, a characteristic also shared from January 1990 until Fall of 1992.
22 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
FIGURE 17 • CPI Indices for G8 Plus Lebanon FIGURE 18 • Overall Fiscal Balance for G8 Plus
Lebanon
Exchange Rates
150 900 Exchange Rates
130 800 250
110 700 200
600
Percent (%)
Percent (%)
90 150
Percent (%)
500
70
400 100
50 300
30 50
200
10 100 0
–10 0 –50
t–3 t–2 t–1 t t+1 t+2 t+3 t+4 t+5 t–3 t–2 t–1 t t+1 t+2 t+3 t+4 t+5
Chile '80 Philippines '81 Mexico '81 Venezuela '94 Argentina '01 Uruguay '02
Lebanon '19 Argentina '80 (rhs) Greece '09 Lebanon '19
Sources: WDI, CAS and WB staff calculations. Sources: WDI, CAS and WB staff calculations.
correlates to the extent of exchange rate volatility in FIGURE 19 • CPI Indices for G8 Plus Lebanon
each of these episodes. Notably, at time t, Lebanon’s CPI Indices
inflation rate is even starker than that of Mexico 300 1,800
(81) (Figure 19); specifically, Lebanon’s inflation rate 1,600
250
1,400
in t+1 (2020) was 84.3 percent, while in Mexico (81) it 200 1,200
Percent (%)
Percent (%)
was 58.9 percent.39 Even though official inflation data 1,000
150
for Argentina (01) is lacking, the failure of the cur- 800
100 600
rency board rendered inflation a main vulnerability in 400
50
the post-crisis period. 200
0 0
t–3 t–2 t–1 t t+1 t+2 t+3 t+4 t+5
Chile '80 Philippines '81 Uruguay '02
Fiscal Greece '09 Lebanon '19 Mexico '81 (rhs)
Sources: WDI, CAS and WB staff calculations.
Lebanon’s overall fiscal balance stands out as
a main vulnerability in the lead-up to the crisis,
comparable to Greece’s fiscal position on the emerged from their crises. The overall fiscal balance
eve of its crisis. Due to data limitations, we examine for each of Chile (80) and Mexico (81) deteriorated
the overall fiscal balance for a subsample of the G8 by around 10 pp of GDP (Figure 20). For the former,
episodes, namely, Chile (80), Mexico (81), Argentina
(01), Uruguay (02) and Greece (09). From this 39
In the period leading up to the crisis, Mexico’s inflation
subsample, only Chile (80) enjoyed fiscal surpluses
rate fluctuated between 17.5 to 29.1 percent annually,
in the lead to the crisis point. Meanwhile, Mexico (81), while Lebanon’s hovered between a minor deflation and
Argentina (01) and Uruguay (02) entered the crisis a moderate 6.1 percent.
with a modest deficit position, whereas Greece (09)’s 40
A noteworthy event in the Greek episode is that data
fiscal deficit was in double digits (as a percentage of revision by the authorities on October 2009, which
GDP), as was Lebanon’s.40 entailed a sizeable increase in the projected fiscal deficit
from 4 to 12.5 percent of GDP, constituted a trigger to the
An ostensible improvement in Lebanon’s
Greek crisis. The data revision came amidst concerns
fiscal indicators (as a percentage of GDP) masks raised by Eurostat—the statistical office of the European
an actual deterioration. There were variations in Commission—regarding the quality of Greece’s fiscal
the fiscal performances of the G8 subsample as they data on five occasions over the period 2005–2009.
Global Crises Comparators: Looking for the Minimum 23
FIGURE 20 • Overall Fiscal Balance for G8 Plus FIGURE 21 • Gross Debt for G8 Plus Lebanon
Lebanon
Gross Debt
Overall Fiscal Balance 250
10
200
5
0 150
% of GDP
% of GDP
–5
100
–10
50
–15
–20 0
t–3 t–2 t–1 t t+1 t+2 t+3 t+4 t+5 t–3 t–2 t–1 t t+1 t+2 t+3 t+4 t+5
Chile '80 Mexico '81 Argentina '01 Chile '80 Philippines '81 Mexico '81
Uruguay '02 Greece '09 Lebanon '19 Argentina '01 Greece '09 Lebanon '19
Sources: WDI, CAS and WB staff calculations. Sources: WEO and WB staff calculations.
this reflected a similar deterioration in the primary GDP) for a subsample of the G8 episodes, namely,
balance, driven by larger primary expenditures (from Chile (80), Philippines (81), Mexico (81), Argentina
23.5 percent of GDP in t–1 to 31.1 percent in t+2). In (01) and Greece (09). Leading to the crisis year, the
the Mexico (81) case, however, the primary balance public debt-to-GDP ratio was below 60 percent for
first deteriorated significantly (by about 6 pp of GDP the G8 subsample (Figure 21), with the exception of
from t to t+1) but eventually improved into a surplus. Greece.41 However, in all cases, the debt-to-GDP ratio
Interest payments, on the other hand, surged from deteriorated in varying degrees, even for Greece,
1.6 percent of GDP at t–1, to 8.5 percent at t+2. which benefitted from a large haircut on its public debt.
Meanwhile, the overall fiscal positions for Argentina
(01) and Uruguay (02) each improved by around 3
pp of GDP from t to t+3. In both cases, this reflected External Position
commensurate improvements in the primary balance.
While an improvement in Lebanon’s overall fiscal bal- A deteriorating current account balance in the
ance seems to resemble that of Argentina (01) and lead to crises is a recurring feature in emerging
Uruguay (02), it masks an actual deterioration. This economies, and we see evidence of this in
deterioration is partially driven by crashing revenues the G8 countries (Figure 22). A worsening of the
and despite significantly lower interest payments as external position raises risk exposures to external
a result of the Eurobond default and non-payments shocks and increases dependence on more volatile
on coupon rates for Eurobonds and TBs held by the portfolio financing. The current account balance
central bank. for large commodity exporters, in this case Chile
and Venezuela, would be highly dependent on
international commodity prices and are more exposed
Debt to commodity price shocks. While commodities are
also important exports for Argentina, Uruguay and
Public debt in Lebanon has consistently been Mexico, it is so to a much lesser extent; for these
one of the highest globally (as a ratio of GDP), countries, consumption and investment are significant
and as a result, a key source of macroeconomic
instability, comparable to only Greece among the 41
The 60 percent of GDP is the debt ceiling identified by
G8 episodes. Due to data limitations, we examine the Maastricht Treaty, which set macroeconomic targets
the central Government gross debt (as a ratio of for countries to qualify for the Euro membership.
24 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
FIGURE 22 • Current Account Balance for G8 the current account balance. This channel, however,
Plus Lebanon was not available for Greece, as its strategic objective
Current Account Balance was to remain in the Eurozone. Instead, the correction
20 in the Greek external position occurred as a result
15
10 of an internal adjustment mechanism, typically more
5 painful and prolonged.
0
% of GDP
–5
–10
–15
–20 Overall
–25
–30
t–4 t–3 t–2 t–1 t t+1 t+2 t+3 t+4 t+5 t+6 Lebanon’s financial crisis stands out as a
Chile '80 Argentina '80 Philippines '81 particularly arduous episode even when compared
Mexico '81 Venezuela '94 Argentina '01 to some of the most severe crises observed since
Uruguay '02 Greece '09 Lebanon '19
1900. In estimating the R&R CSI for the Lebanon
Sources: WDI and WB staff calculations. financial crisis, we make reasonable assumptions on
its depth and duration. The results suggest that the
Lebanon crisis is likely to be of the 10, possibly three,
most severe global crises episodes, as observed
drivers for current account balance through imports. It
and examined by Reinhart and Rogoff (2014) over a
is noteworthy that of the G8 countries, only Argentina
period surpassing a century and half. This is further
(01) and Uruguay (02) did not experience a worsening
confirmed when we compare select macroeconomic
of the current account balance up to the crisis point.
indicators for Lebanon with those for R&R’s relatively
A clear implication of the crises in most
more recent crises.
G8 countries is a sharp correction in the external
As such, we expect the adjustment process
position. Only Chile (81) deviates as prices of min-
to be more painful and to take longer, even with
erals, oils and metal were falling in 4 of the 5 post
optimal policy measures in place. As it currently
crisis years.42 For Mexico (81), the Philippines (81),
stands, however, the absence of a comprehensive
Venezuela (94) and Argentina (01), the correction in
and consistent adjustment strategy can only make
the current account balance occurred as a result of
this more difficult.
depreciations in the exchange rates, which increased
the competitiveness of their economies. Mild
depreciation in Uruguay (02) corresponds to mild 42
According to UNCTAD, Free Market Commodity Price
post-crisis correction in the current account balance. Indices, the annual percentage change in the price for
Noteworthy, Argentina (80) currency collapse cor- minerals, oils and metal, was –15.6, –12.2, 7.9, –8.1,
responded to only a minor post-crisis improvement in –4.6 percent from 1981 to 1985, respectively.
Global Crises Comparators: Looking for the Minimum 25
4
OUTLOOK
AND RISKS
S
ubject to extraordinarily high uncertainty, Lirafication and monetization of the fiscal deficit, will
we project real GDP to contract by a continue to be critical to the inflationary environment.
further 9.5 percent in 2021. Our projections Assuming that in 2021, the Lebanese pound suffers a
(Table 4) assume that COVID-19 conditions carry comparable depreciation in the US$ banknote market
through 2021, while macro policy responses remain as it did in 2020, we expect inflation to remain highly
inadequate. We also assume a minimum level of elevated in 2021, possibly surpassing 2020 rates.
stability on the political and security scenes, but On the fiscal front, a continued collapse in
refrain from assuming runaway inflation-depreciation, revenues is expected to lead a deterioration in
which is a realistic scenario. For more empirical Lebanon’s fiscal position compared to 2020. On
discussion on our projection, please refer to Annex A. revenues, we assume marginal nominal increases
As mentioned earlier, monetary and compared to 2020, as the economic contraction is
financial turmoil are driving crisis conditions, expected to be less severe and high inflation boosts
more acutely through interactions between the collection in nominal (but not real) terms. Very high
exchange rate, narrow money and inflation. The inflation rates will drive a denominator-led shrinking
unusual centrality of this dynamic on the macro of the revenues-to-GDP ratio by a projected 4.6 pps,
framework is an important caveat regarding our mac- to reach almost 7 percent, compared to a medium-
roeconomic outlook for 2021. In an attempt to pin term average of 20.7 percent. On expenditures, and
down the framework and help narrow the confidence despite proposed 2021 Budget numbers, we factor
intervals for our outlook, we undertake crisis-specific a 30 percent increase in nominal primary spending,
econometric estimations; specifically, we assess due again to inflationary pressures. As a ratio of
the response of inflation to changes in currency in GDP, however, primary current expenditures are also
circulation using multivariate time series models. The projected to decline due to the same denominator-
results, which are presented in Annex B, suggest that led effect. Based on this, the overall and primary
a 1 percent shock to currency in circulation gener- balances are expected to deteriorate in 2021.43
ates a response of 0.80 pp increase in the inflation
rate over a twelve-month horizon. Hence, policy 43
That suggests that BdL will have to keep financing the
with implications on narrow money supply, such as deficit which will accelerate its FX reserve losses. Hence,
27
Lebanon’s recession is likely to be arduous Therefore, the removal of the subsidy will have significant
and prolonged given the lack of policymaking lead- social implications due to direct and indirect effects
ership and reforms. Lebanon’s GDP/capita has fallen on residents’ purchasing power. Further, Lebanon’s
by around 40 percent over the 2018–2020 period and political economy suggests that, much like solutions (or
is expected to decline further. Hence, Lebanon’s World lack thereof) to other burdens and deficiencies, there is
Bank income classification is likely to be downgraded a high likelihood of suboptimality. Possible scenarios
from an upper-middle income economy, which should and associated social implications include:
enjoy a GNI per capita of between U$4,046 and
US$12,535, to a lower-middle income status.44 a. An orderly FX subsidy removal. This would involve
Lebanon faces realistic threats to its already effectual political and security coordination,
fragile social peace. Angry demonstrations have along with its replacement with a more effective
been erupting in cities across Lebanon, protesting the and efficient pro-poor (targeted) program. In this
depreciation of the local currency and the associated case, social implications can be mitigated.
very high inflation rates, as well as general economic b. Non-coordinated FX subsidy removal. This can
and political conditions. While numbers have not been involve BdL unilaterally halting the subsidy, with
large, these demonstrations have been spread across minimal political and security coordination. In this
the country, and have cut off important roads and case, BdL will carry the political cost, and can
highways, thereby causing significant disruptions. A involve increased demonstrations targeting BdL.
more recent phenomena has involved scuffles in super- The political class can pass on the blame and
markets and angry protests outside over access to feel somewhat relieved, but to many, the obvious
subsidized products. This is even occurring in high end responsibility of the political class will remain.
supermarkets. Further, the volatile situation allows for c. Disorderly FX subsidy removal due to depletion
parasitic groups that can be of a more sinister nature of FX reserves at BdL. This traverses subsidy
to usurp legitimate popular discontent creating grave management to constitute a balance of
security implications. This is in addition to Lebanon’s payments crisis within a balance of payments
long-term sectarian fragmentation. Hence, there is crisis—the complete depletion of FX resources in
growing wariness of potential triggers to social unrest. In an economy that has been trying to ration these
this LEM, we highlight two potential economic triggers. resources to pay for critical needs. It can trigger an
inflationary-depreciation spiral and cause import
shortages. Naturally, there are definite significant
The FX Subsidy45 social and possible security implications.
The first Special Focus of the LEM examines
Lebanon’s FX subsidy for critical and essential Large Scale Interruptions to Vital
imports, which offers a serious political and social Public Services
challenge. On the one hand, the current FX subsidy
is both distortionary, expensive, and regressive. Its The second Special Focus of the LEM discusses
elimination and possible replacement with a more the impact that the crises are having on four basic
effective and efficient pro-poor (targeted) program
would improve Lebanon’s balance of payments,
meaningfully extend the time-till-exhaustion of the fiscal trajectory is contingent on BdL having sufficient
remaining BdL reserves, and help cushion the impact reserves to continue to finance the government.
44
Latest available Gross National Income per capita for
on Lebanon’s poor and middle class. On the other
Lebanon is for 2019 US$7,380.
hand, the subsidy prevents, in the very short-term, 45
For a more detailed analysis on the FX subsidy, please
the prices of these products from increasing, which refer to: World Bank (2020), Lebanon Subsidy Reform
would exacerbate inflationary-depreciation pressures. Note, December 2020.
28 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
TABLE 4 • Selected Macroeconomic Indicators for Lebanon; 2016–2021
2013 2014 2015 2016 2017 2018 2019 Est. 2020 2021
Proj.
Real sector (annual percentage change, unless otherwise specified)
Real GDP 3.8 2.5 0.2 1.5 0.9 –1.9– 6.7 –20.3 –9.5
Real GDP per capita 1
–2.8 –3.2 –3.9 –1.2 –0.6 –2.5 –6.8 –20.7 –10.0
Agriculture (share of GDP) 3.9 4.4 3.8 4.0 4.5 4.4 5.0 5.0 6.0
Industry (share of GDP) 14.2 13.4 12.7 12.8 12.3 12.0 10.6 13.5 13.5
Services (share of GDP) 70.9 71.3 72.0 71.5 71.6 72.2 74.3 77.6 78.3
Net indirect taxes (share of GDP) 11.0 10.9 11.5 11.7 11.6 11.4 10.1 3.9 2.2
Money and prices
CPI Inflation (p.a) 2.7 1.2 –3.7 –0.8 4.5 6.1 2.9 84.3 100.0
Money2 9.0 6.0 5.1 7.3 4.2 3.0 –6.7 198.0 125.0
Investment & saving (percent of GDP, unless otherwise specified)
Gross capital formation 27.6 24.9 22.2 22.7 21.4 20.8 18.5 10.0 8.1
o/w private 25.8 23.4 20.8 21.3 19.9 19.1 17.2 8.8 6.8
Gross national savings 2.1 –1.3 5.1 2.2 –1.5 –3.5 –2.7 –1.0 –1.1
o/w private –1.8 –3.9 1.0 –1.0 –4.8 –5.3 7.9 2.7 2.9
Central government finance (percent of GDP, unless otherwise specified)
Revenue (including grants) 20.1 22.6 19.2 19.4 21.9 21.0 20.6 11.5 6.9
o/w. tax revenues 14.3 14.3 13.7 13.7 15.5 15.4 15.5 8.1 4.9
Total expenditure and net lending 29.0 28.9 26.9 28.6 28.6 32.0 31.2 16.4 12.2
Current 27.3 27.3 25.5 27.3 27.1 30.3 29.9 15.2 10.9
o/w interest payment 8.1 8.7 8.9 9.3 9.4 9.8 10.0 2.1 1.4
Capital & net lending (excluding 1.8 1.5 1.4 1.4 1.5 1.7 1.3 1.2 1.3
foreign financed)
Overall balance (deficit (-)) –9.0 –6.3 –7.7 –9.3– 6.7 –11.0 –10.5 –4.9 –5.3
Primary balance (deficit (-)) –0.9 2.4 1.2 0.0 2.7 –1.2 –0.5 –2.8– 3.9
External sector (percent of GDP, unless otherwise specified)
Current account balance –25.6 –26.2 –17.0 –20.5 –22.9 –24.4 –21.2 –11.0 –9.2
Trade balance –28.4 –29.9 –22.9– 23.6 –24.7 –24.8 –24.9 –17.9 –21.9
o/w export (GNFS) 44.5 40.0 39.7 37.3 36.0 35.7 35.4 26.9 41.9
Exports of goods 11.0 9.5 8.0 7.7 7.6 7.0 9.3 12.1 18.7
Exports of services 33.5 30.6 31.7 29.6 28.4 28.7 26.1 14.8 23.3
o/w import (GNFS) 73.0 69.9 62.6 60.9 60.8 60.5 60.3 44.8 63.8
Imports of goods 45.3 42.5 35.2 35.0 34.7 34.4 35.0 27.5 39.2
Imports of services 27.7 27.4 27.4 25.9 26.1 26.1 25.2 17.3 24.6
(continued on next page)
Outlook and Risks 29
TABLE 4 • Selected Macroeconomic Indicators for Lebanon; 2016–2021 (continued)
2013 2014 2015 2016 2017 2018 2019 Est. 2020 2021
Proj.
Net private current transfers: 3.4 4.9 6.8 4.8 2.3 2.5 5.6 9.6 16.7
Net remittances 5.0 5.8 7.2 6.6 5.2 4.2 6.1 9.7 16.5
Net income receipts –0.6 –1.2 –0.9 –1.7 –0.5 –2.1 –1.9 –2.7 –4.1
Capital accounts 0 0 0 0 0 0 0 0 0
Gross reserves (months of imports 11.7 13.1 13.8 15.2 15.6 14.3 14.3 19.8 10.6
GNFS)c,d
Total public debt
Total debt stock (in million US$) 63,490 66,564 70,325 74,900 79,530 85,139 88,900 58,082 43,958
Debt-to-GDP ratio (percent) 135.3 138.3 140.8 146.3 149.7 154.9 171.0 174.0 197.2
Memorandum items:
GDP (in million US$) 46,909 48,134 49,939 51,205 53,141 54,961 51,992 33,383 22,297
Source: Government data, and World Bank staff estimates and projections.
a
Population figures, which include Syrian refugees registered with the UNHCR, are taken from the United Nations Population Division.
b
Prior to 2020 this is M3, including non-resident deposits; 2020 and after, this is M0 (currency in circulation).
c
Gross Reserves (months of imports GNFS) = (Imports of Goods & Services / Gross Res. excl. Gold)*12.
d
Total Imports using the BOP data from the Quarterly Bulletin of BDL.
public services: electricity, water supply, sanitation further undermined this set up via two effects: (i) it has
and education. While Lebanon’s public service significantly increased poverty rates, expanding the
delivery has long been notoriously deficient relative demographic that is not able to afford these private
to its upper middle-income status, long established substitutables, and are thus more dependent on
mitigation measures and private substitutables have public services; and (ii) it threatens financial viability
traditionally filled the gap, particularly for those with and basic operability of the sector by raising its costs
economic means. The Deliberate Depression has and lowering its revenues.
30 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
SPECIAL FOCUS I:
FX SUBSIDY REFORM
IN THE DELIBERATE
DEPRESSION
T
he real question regarding BdL’s FX subsidy Background
for imports of critical and essential goods, is
when and how to remove it, not whether. This Policymakers in Lebanon are considering a shift
note shows that the sooner the subsidy is replaced away from the FX subsidy for imports of critical
with a cheaper and more effective compensation and essential goods, towards direct transfers to
scheme, the better for the economy and for people’s households. The decision to remove the FX subsidy
welfare. This is because the current FX subsidy is for imports of critical goods and essential food and
both distortionary, expensive (costing an estimated other items (henceforth referred to as essential
US$287m/month), and regressive (i.e., benefits items) should be based on the most efficient and
predominantly wealthier consumers). Its elimination cost-effective use of pro-poor public spending. The
and possible replacement with a more effective and subsidy concerns import of critical goods—energy,
efficient pro-poor (targeted) program would improve medicine, wheat—and essential items (as identified
Lebanon’s balance of payments, meaningfully extend by the MoET. The estimated cost of the subsidy
the time-till-exhaustion of remaining BdL reserves, amounts to a monthly average of US$287 million
and help cushion the impact on Lebanon’s poor and (Table 5).46 Maintaining the current subsidy scheme
middle class. This note suggests a broad-based cash accelerates the steady depletion of FX reserves at
transfer program as one such option, which should BdL and reduces the time available to undertake
be in place prior to subsidy removal. However, while reforms to avoid a forced and disorderly adjustment
the removal of the FX subsidy would be welcome, it of the exchange rate. On the other hand, the subsidy
is a mere short-term patch as only a comprehensive prevents, in the very short-term, the prices of these
reform package that is consistent with a credible
macroeconomic framework can prevent the country
from running out of reserves and being forced into 46
This is distinguishable from the total value of the subsidy,
a disorderly and highly disruptive exchange rate which amounts to an estimated US$437 million per
adjustment. month.
31
products from increasing, which would exacerbate and health externalities—namely, local air pollution,
inflationary-depreciation pressures. The question congestion, accidents and roadway wear and tear.
policymakers ought to consider is: can a cheaper and Hence, there is a socio-economic benefit in replacing
more effective compensation scheme be immediately the regressive portion of the subsidy with a targeted,
implemented to both protect poor and vulnerable and therefore cheaper, transfer program.
households, and gain some time to protect the Inflation is a highly regressive tax, dispro-
official exchange rate until a comprehensive set portionally affecting the poor and vulnerable,
of policy reforms that are consistent with a stable and more generally, people living on fixed
macroeconomic framework can be introduced? incomes, such as pensioners. This is especially
so in Lebanon’s case where key basic items of the
consumption basket are primary drivers of overall
Macroeconomic Considerations inflation, which has reached 145.8 percent in 2020.
In fact, the average yoy inflation rate over 2020
As it stands, half of the cost of the FX subsidy for food and non-alcoholic beverages was 254.3
is directed toward energy items, which are percent, while that for clothing and footwear was
regressive in nature. Higher-income earners are 289.8 percent, and 386.7 percent for furnishings and
more likely to consume more fuel, and thus profit household equipment.
more from the subsidy. The gains associated Lebanon’s sudden stop in capital inflows
by implicitly subsidizing road transport are not has implied a steady depletion of FX reserves at
distributed fairly; by income, the poorest 20 percent BdL, notwithstanding the introduction of informal
of the population receives only 6 percent of the capital controls (Figure 23). As of end-February
subsidy, while the richest 20 percent receives 55
percent.47 Finally, low fuel prices encourage over- 47
UNDP, 2015, “Fossil Fuel Subsidies in Lebanon: Fiscal,
consumption, adding to negative environmental Equity, Economic and Environmental Impacts.”
TABLE 5 • Cost of FX Import Subsidy and Impact of its Removal
BoP Impact due to Estimated Reduction in Demandd
Estimated Low
2020 Subsidized Subsidy in Cost of elasticity-high
consumptiona exchange % subsidizeda valueb subsidyc High elasticity Low elasticity passthrough
Product ($, mlns) Rate1 (LL/S) (%) ($, mlns) ($, mlns) ($, mlns) ($, mlns) ($, mlns)
Fuel EdL 1,000 1,515 90% 900 661 NA NA NA
Gaz (LPG) 110 1,515 90% 99 73 80 39 66
Mazout 1,195 1,515 90% 1,076 790 867 418 717
Other fuel 1,070 1,515 90% 963 707 776 375 642
Medication/supplies 1,300 1,515 85% 1,105 811 1,067 455 780
Wheat 150 1,515 90% 135 99 57 57 96
Essential items 960 3,900 100% 960 303 367 367 612
Total 5,785 1,911 91% 5,238 3,444 3,213 1,710 2,913
a
Source: Ministry of Economy and Trade (MoET).
b
Calculated as: (estimated 2020 consumption) * (the subsidized exchange rate).
c
Calculated as: [(the average black market exchange rate—the subsidized exchange rate)/ (the average black market exchange rate)] * (subsidy in value).
d
Calculated as: (estimated 2020 consumption) * (percentage change in demand per selected scenario per product) Average black market exchange rate in 2020 is assumed at 5,700
LL/US$.
32 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
FIGURE 23 • A Steady Depletion in the Gross Foreign Exchange Position at BdL
Gross reserves at BdL
50,000
45,000
40,000
35,000
US$ mln
30,000
25,000
20,000
15,000
10,000
5,000
0
Jan-13
Apr-13
Jul-13
Oct-3
Jan-14
Apr-14
Jul-14
Oct-14
Jan-15
Apr-15
Jul-15
Oct-5
Jan-16
Apr-16
Jul-16
Oct-16
Jan-17
Apr-17
Jul-17
Oct-17
Jan-18
Apr-18
Jul-18
Oct-18
Jan-19
Apr-19
Jul-19
Oct-19
Jan-20
Apr-20
Jul-20
Oct-20
Jan-21
Foreign currencies Foreign securities Gross FX reserves Compulsory FX reserve
Sources: WB staff calculations based on WDI data.
2021, BdL’s gross foreign asset position reached Demand effects resulting from the removal
US$22.9 billion, declining by US$14.4 billion since of subsidies are a key factor for an impact on
end-2019, and by US$1.2 billion from end-2020. The the balance of payments (BoP). In principal, the
gross position, however, differs widely from its net absence of demand effects, dragged by the fall in
reserves (i.e., FX reserves at the central bank net of FX income and increase in prices, would leave the BoP
liabilities to others).48 BdL’s gross position includes unchanged by the removal of subsidy. To estimate
US$5 billion in Lebanese Eurobonds, an unpublished the impact on demand, we use country specific,
amount lent out to banks since October 2019, and product-based, income and prices elasticities of
required reserves on banks’ customer FX deposits demand as presented in Seale et al (2003).50 Notably,
(estimated at US$16.7 billion). the study finds that energy products and medical care
BdL officials have warned that falling FX are generally elastic for most countries, including in
reserves will force the central bank to halt its sup- Lebanon; that is, if consumers’ incomes decline by
port for imports of critical goods and essential half, or if product prices increase by 50 percent, then
items at lower exchange rates. The officials indi- demand for these products declines by half or more.
cated that in the next few months BdL’s FX reserves Meanwhile, demand for food items are relatively
will drop to the level of the Required Reserves on inelastic; that is, if consumers’ incomes decline by
banks’ customer FX deposits. Should BdL abandon half, or if product prices increase by 50 percent,
its support for imports of critical goods and essential then demand for these products declines by less
items (i.e., including energy products, medicine and
food essentials), importers will be forced to fully revert
to the US$ banknote market rate49 for the needed 48
BdL, contrary to other central banks, does not publish
hard currency. New imports would thus be priced at net reserves.
the going US$ banknote market rate, exacerbating 49
This refers to the street market access to dollar
inflationary pressures and potentially stoking an banknotes, which has depreciated the Lira by up to 90
percent.
inflationary-depreciation spiral. In fact, in anticipation 50
James Seale, Anita Regmi, and Jason Bernstein (2003),
of higher prices, importers, retailers and customers International Evidence on Food Consumption Patterns,
are already hoarding critical and essential goods, United States Department of Agriculture, Electronic
bringing forward price and volume implications. Report from the Economic Research Service.
Special Focus I: FX Subsidy Reform in the Deliberate Depression 33
than half.51 We shall preclude from our simulations and disorderly exchange rate adjustment. The
EdL due to the company’s obvious deficiencies, spe- sudden stop in capital inflows has induced a more
cifically the severe and chronic shortages in power direct trade-off between the stock of FX reserves at
supply and tariff regulations that distort the impact BdL and the import bill. As a result, the BoP relief
on demand. Further, there is uncertainty regarding shown will reflect on the foreign exchange reserves
the elasticities for these consumption goods in the at BdL. We stress that the improvement in the BoP
Lebanese economy. Specifically, other literature position and the extension of reserves are temporary
suggest that energy products and medical care are and come at the expense of worsening economic
inelastic goods.52 For robustness, we also re-run activity and declining standard of living.
simulations based on relative in-elasticities of energy As a result of subsidy removal, inflationary
products and medical care.53 We shall henceforth pressures would materialize via direct and itera-
refer to simulations based on Seale et al (2003) as tive effects. Through the direct effect, the inflation
High Elasticity Scenario, and those that assume the rate would increase as prices for critical goods and
inelasticity of energy products and medical care as the essential items reflect the higher US$ banknote
Low Elasticity Scenario. We proceed by (i) assuming exchange rate.55,56 Additionally, an iterative effect
that income contracts by 20 percent across all sce- arises from an increased demand for dollars in the
narios; and (ii) simulating two scenarios for product dollar-note market, which further depreciates the
price increases: a 50 percent increase in prices of currency, fueling inflation. With surging inflation,
energy products, medicine and food items, and a 100 demand for narrow money increases. Meeting this
percent increase (dubbed the High Passthrough).
Hence, we present three scenarios:
51
According to Seale et al. (2003), income elasticities of
1. High Elasticity Scenario: a high elasticity scenario demand for Lebanon are 0.632 for food, beverage and
that assumes a 20 percent contraction in income tobacco, 1.2 for gross rent, fuel and power, and 1.357 for
medical care. The authors also present price elasticities
and a 50 percent increase in prices of energy
of demand for Lebanon, which are –0.511 for food,
products, medicine and food items; beverage and tobacco, –0.971 for gross rent, fuel and
2. Low Elasticity Scenario: a low elasticity scenario power, and –1.098 for medical care.
that assumes a 20 percent contraction in income 52
Accurate elasticities can be derived from more recent
and a 50 percent increase in prices of energy and comprehensive Lebanese household surveys which
products, medicine and food items; are not available.
53
Toward that end, we assume income (price) elasticities
3. Low Elasticity-High Passthrough Scenario: a low
of demand for Lebanon to be 0.5 (–0.5) for gross rent,
elasticity scenario that assumes a 20 percent fuel and power and 0.5 (–0.5) for medical care.
contraction in income and a 100 percent increase 54
Here we omit a High Elasticity-High Passthrough
in prices of energy products, medicine and food scenario that assumes a 20 percent contraction in
items.54 income and a 100 percent increase in prices of energy
products, medicine and food items. The reason is that
when price elasticity of demand is at or lower than –1, a
Removal of subsidies on imports of critical
100 percent increase in prices would wipe out demand,
goods and essential items can lead to some BoP which is not a realistic outcome.
relief. The sharp declines in consumption of the 55
We note that energy products are final and intermediate
subsidized products translate into BoP relief; all three goods, and changes to their prices impose economy-
scenarios suggest improvements to the BoP, ranging wide effects.
from US$1.7 to US$3.2 billion depending on the
56
While this will naturally reflect on volumes imported—
since consumers will cut down on purchases, and there
scenario (Table 5).
will be some substitutional effects for food products that
Based on the above, the removal of subsidy can be manufactured locally (a process likely ongoing)—
can meaningfully extend the time-till-exhaustion being critical goods, they will still be demanded in
of remaining BdL reserves, thus delaying a forced substantial volumes.
34 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
FIGURE 24 • Upon Removal of FX Subsidy, Direct Effects on Inflation are Substantial
140
130
120
110
100
CPI index
90
80
70
60
50
40
Aug-19
Sep-19
Oct-19
Nov-19
Dec-19
Jan-20
Feb-20
Mar-20
Apr-20
May-20
Jun-20
Jul-20
Aug-20
Sep-20
Oct-20
Nov-20
Dec-20
Jan-21
Feb-21
Mar-21
Apr-21
May-21
Jun-21
Jul-21
Aug-21
Sep-21
Oct-21
Nov-21
Dec-21
CPI, status quo scenario (Dec 2020 = 100) CPI, sudden stop scenario (Dec 2020 = 100)
Sources: WB staff calculations based on WDI data.
demand with a corresponding increase in money
supply—BdL’s current response—would lead to exac-
57
There are also stock implications, whereby high inflation
imposes a wealth effect via transfers from creditors to
erbated pressures on the exchange rate. Moreover,
debtors; ceteris paribus borrowers would owe less—
under panic conditions, inflation and exchange rates and hence, creditors would receive less—in real terms
are key observables that drive sentiment, and their than what was determined at the time the contract was
deterioration re-enforces the inflationary-deprecia- signed. Consequently, public debt denominated in local
tion cycle.57 currency would be worth less in real terms, providing a
Removing the subsidy is estimated to fiscal benefit.
58
We note that this estimation does not take into
increase inflation through the direct effect by 24
consideration the iterative effects, which can potentially
percentage points (Figure 24).58 This inflationary drive an inflation-depreciation spiral.
impact is frontloaded, hitting hardest the first months 59
The AER is calculated using consumption-based
and diminishing over the course of the year. We weights applied on multiple exchange rates, specifically,
caveat that this does not account for the impact on the official exchange rate, the platform rate at BdL and
inflationary expectations and other related iterative the US$ banknote rate. For further details on the AER,
please refer to: The Lebanon Economic Monitor, Fall
effects, which are likely to be forceful dynamics. The
2020, The Deliberate Depression.
inflation results are generated from an estimated 60
The exchange rate pass-through rate on inflation is
AER59 and an exchange rate pass-through effect on calculated by first dividing the inflation rate by the AER
prices of about 50 percent60 applied to the following depreciation rate for the same month and multiplying by
two scenarios: (1) a hypothetical continuation of the 100. This generates a series of pass-through rates for
the time period August 2019 to September 2020, which
FX subsidy through 2021, which we will refer to as the
we averaged out. The estimated exchange rate pass-
Status Quo Scenario;61 and (2) a full termination of through rate will likely change as more actual data are
the FX subsidy on January 2021, dubbed the Sudden populated.
Stop Scenario.62 61
This scenario is used for illustrative purposes and is not
The Government of Lebanon (GOL) needs a likely option. Sharply falling FX reserves at the central
to prioritize a comprehensive, consistent and bank precludes the option of continued support for the
full list of critical goods and essential items.
credible macroeconomic stabilization plan, the 62
The Sudden Stop Scenario assumes (i) the subsidy is
fiscal part of which should include a social safety eliminated on January 2021; and (ii) the US$ banknote
net (SSN) component. Given Lebanon’s rapidly exchange rate deteriorates to LL 10,000 per US$ by
and continuously deteriorating macroeconomic January 2021, stabilizing thereafter.
Special Focus I: FX Subsidy Reform in the Deliberate Depression 35
conditions, a social safety net by itself will be compensation program proposed is of a ‘broad
insufficient, and indeed, as shown above, can coverage’ nature.67 In year 1, the broad-coverage
trigger additional macroeconomic risks. Lebanon subsidy reform compensation scheme (Broad
needs to incorporate a SSN into a macroeconomic Coverage-Cash Transfer [BC-CT]) could aim to
stabilization strategy that prioritizes the arrest of cover approximately 80 percent of the Lebanese
hyperinflation, currency depreciation and prolifera- population. To ensure progressive coverage, the
tion of multiple exchange rates. This strategy would BC-CT will seek to exclude the top 20 percent of
be based on: (i) a debt restructuring that would the population as they are better able to absorb the
achieve short-term fiscal space and medium-term price increases and can afford to pay for the cur-
debt sustainability; (ii) comprehensively restructuring rently subsidized items at market price. The broad
the financial sector in order to regain solvency of the coverage approach would involve identifying and
banking sector; (iii) creating a new monetary policy excluding the top 20 percent of the population (in
framework that would regain confidence and stability terms of income/economic wellbeing) and covering
in the exchange rate; (iv) a phased fiscal adjustment the remaining 80 percent of the population. Unlike
aimed at regaining confidence in fiscal policy; traditional poverty-targeted programs that seek to
(v) growth enhancing reforms; and (vi) enhanced identify the poor (bottom 10–40 percent, depending
social protection. on the country), in this case, the identification of the
richest 20 percent would be based on affluence
tests (i.e., markers of affluence to identify the ineli-
Design of a Broad-Coverage Subsidy gible) such as asset filters, dwelling ownership and
Reform Compensation Scheme63 characteristics, and formal incomes. Preliminary
analysis undertaken by the World Bank points to
Governments often use generalized subsidies as a the potential use of indicators such as number of
tool to lower the cost of living for poor households rooms/bathrooms per household member, dwelling
and to shield households from price fluctuations— area per household member, and thresholds for
Lebanon is not alone in this approach. However, wages and other formal income for such an afflu-
subsidies are a blunt and inefficient instrument, and ence test. Implementing such an approach would
some are regressive, benefiting the rich more than the require careful integration of data across multiple
poor. International experience shows that a shift from databases to ensure reliability and automation of
generalized subsidies to direct support to the poor eligibility decisions. In subsequent years, as the
could result in an improvement in public welfare.64
The subsidy reform being considered by
the GOL aims to replace implicit FX subsidies 63
The proposal in this section reflects the recommendations
for fuel, medicines, wheat and essential items of the World Bank and not necessarily what GOL will
with direct cash transfers to Lebanese resident adopt.
households.65 The direct cash transfers should offer
64
Ruslan Yemtsov and Amr Moubarak “Assessing the
readiness of Social Safety Nets to Mitigate the Impact of
some compensation to households to purchase items
Reform”; World Bank Good Practice Note 5.
at market prices (which will increase following the 65
This section refers to the GoL subsidy reform proposal
withdrawal of subsidies as discussed previously) and as presented by the Minister of Economy and Trade in
should not be restricted in their use (allowing greater April 2021 and published on the MoET website. As of
autonomy and a dignified freedom of choice on the the publication of this note, the GOL has not reached
best use of assistance while offering compensation a decision on what kind of scheme will replace the
generalized subsidies scheme.
for the anticipated increase in prices). 66
Especially as higher fuel prices will lead to increases in
As the withdrawal of FX subsidies will prices of several other goods and services.
result in a considerable price shock66 that will 67
Covering the majority of Lebanese households though
be felt by large sections of the population, the not attempting universal coverage.
36 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
population adjusts to the new (market) prices for FIGURE 25 • Proposed Monthly Benefit Amount
the currently subsidized items and as economic (US$), and Coverage (%), by Year
recovery begins, the coverage of this compensa-
60 90%
tion program may be progressively reduced, along 80%
50 80%
with the level of benefits (per adult and per child) to 50 70%
70%
ensure a smooth phasing out of the compensation 40 40
40 60% 60%
Benefit Amount
program that dovetails with more sustainable and
Coverage
30 30 50%
30 45%
assured financing for a targeted safety net program 25 40%
20 20
that covers around 30 percent of the population 20 30% 30%
15 15
over time. 20%
10
We specifically (and demonstratively) 10%
analyze one proposal of a benefits scheme under 0 0%
Year 1 Year 2 Year 3 Year 4 Year 5
which Lebanese adults (age 23 and above) receive
Amount Adult Amount Child % Population Covered
US$50 per month in year 1 of the compensation
program, US$40 in years 2 and 3, and US$30 in
years 4 and 5. Children get half this amount. The cov-
erage progressively declines from 85 percent of the FIGURE 26 • Compensation Program Outlay and
Lebanese population in the first six months, to 75 per- Net Savings from Phase I Subsidy
cent in the next six. Over the next 4 years, coverage Withdrawal, US$ mn
declines to 70 percent, 60 percent, 45 percent and
Adult $50 Adult $40 Adult $30
30 percent of the population, respectively (Figure 25.) 3,500 Child $25 Child $20 Child $15
2,906 3,060
The scheme described represents compen- 3,000 2,751 2,592
2,500 2,407
sation equivalent to 41 percent of the average 1,835
2,000
Lebanese household’s monthly consumption 1,500
1,536 964
expenditure in Year 1, 28 percent in years 2 and 1,000 779
620 465
3, and 21 percent in years 4 and 5, reflecting a 500 311
substantial level of support to recipient households 0
Year 1 Year 2 Year 3 Year 4 Year 5 Average 5 Years
(compared to most unconditional cash transfer pro-
Cash Transfer Program Outlay ($ mn) Savings ($ mn)
grams in upper-middle income countries that provide
22 percent of beneficiary welfare).68
One option we model is a phased
approach to subsidy removal: Phase 1 is removal
of FX subsidy for the MOET essential items and million. The removal of phase I subsidies and the
fuel (except for EdL) to take place in January 2021. introduction of the cash transfer compensation as
Phase 2 is removal of fuel subsidy for EdL along described will result in a net savings of an average
with a reform of EdL’s pricing reform to take place annual value of around US$2.6 billion over the
on Jan 2022. Phase 3 is removal of medication 5-year period.69
subject to broad medical coverage for Lebanese BC-CT financing options and associated
citizens to take place in January 2023. risks necessitate a credible medium-term mac-
Figure 26 shows the required outlay of roeconomic framework. In principle, financing
the program and the resulting savings against options for the BC-CT include: (i) money printing by
Phase 1 of the removal of subsidies. The bud-
getary outlays start at US$1.5 billion in year 1 (80 68
World Bank (2018) The State of Social Safety Nets 2018.
percent coverage) and decline to US$311 million Washington DC: The World Bank.
in year 5 (30 percent coverage); equivalent to an 69
These savings do not account for administrative costs of
average annual expenditure of around US$779 implementing the compensation program.
Special Focus I: FX Subsidy Reform in the Deliberate Depression 37
the central bank (i.e., monetization); (ii) budgetary communication is critical for successful subsidy
allocations; and (iii) international assistance.70 In light reforms. In Jordan, for example, a widely cast and
of BC-CT’s large financing needs, monetization risks well-designed communication strategy played a
significantly exacerbate current macroeconomic crucial role in addressing uncertainties and managing
conditions, further entrenching macroeconomic expectations during the 2008 and 2011/12 subsidy
instability. This option will lead to further deteriora- reforms. Making an effective use of available
tion in monetary conditions, potentially stoking the channels to project transparency, clarity on the role
afore described inflationary-depreciation effects. of the program, public information on objectives,
Alternatively, fiscal financing via additional (unfunded) operational rules, and results are necessary to tackle
budgetary allocations would not be qualitatively dif- information asymmetries and concerns of different
ferent from the monetization option; currently, the stakeholders. Bringing in different stakeholders in
fiscal deficit is de facto monetized. The previous the design of the program would enrich the scheme
two options are also likely to preclude effective and its acceptability. Reforms can succeed only
international assistance (option iii), which can only be if an informed public accepts and supports the
as part of a credible medium-term macroeconomic reform’s rationale. Clearly communicating who will be
framework. As mentioned above, this framework impacted, and how, is vital to generate public buy-in.
should include: (a) a debt restructuring program
that would achieve short-term fiscal space and Identification and Beneficiary Selection
medium-term debt sustainability; (b) comprehensivly
restructuring the financial sector in order to regain The BC-CT scheme would aim to cover the
solvency of the banking sector; (c) creating a new Lebanese population residing in Lebanon only
monetary policy framework that would regain confi- (which refers to individuals who spend at least
dence and stability in the exchange rate; (d) a phased 183 days in the country). The program is proposed
fiscal adjustment aimed at regaining confidence in to involve a simple mobile-based registration (self-
fiscal policy; (e) growth enhancing reforms; and (f) declared information) to be completed by the
enhanced social protection (i.e., the BC-CT program). applicant. No household visits will be undertaken. The
program will require ownership of a National ID71 (NID).
Individuals who do not have a NID will be required to
Implementing a Broad-Coverage obtain an NID before registering. The program should
Subsidy Reform Compensation provide two different registration options: (i) online
Scheme through citizen interface (for individuals who can
access the internet72) or (ii) physical visit to one of the
Delivering a broad-coverage cash transfer (BC- registration sites which could include the LibanPost,
CT) scheme requires careful design, transparent and the Social Development Centers (SDCs), or the
implementation, and adequate financing. It
involves the following broad steps: (i) consultations
and communications; (ii) identification and beneficiary 70
For illustrative purposes here, we assume each of
selection; (iii) payment of benefits; and (iv) monitoring these options are utilized exclusively, when in practice
and grievance redress mechanism (GRM). there are financing modalities that use combinations
thereof.
Consultations & communication 71
Initial assessment indicates that 78 percent of the
Lebanese population have NIDs, World Bank Technical
Mission for a Digital Transformation Project, ID4D,
A well-prepared communication and out-reach
Sept 2019.
campaign must precede and accompany the 72
Share of population who uses internet is 75.9 percent
subsidy reform program. International experience (penetration rate), Internet usage, Broadband and
demonstrates that well-planned and consistent Telecommunications report by Internet World Stats, 2016.
38 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
use of registration trucks. Each entry point will use to the financial institution(s) based on the applicant’s
the same “basic delivery system” accessible through preferred payment choice. Giving recipients the
a mobile compatible citizen interface. A selfie- choice of the financial service provider often improves
based registration will be utilized for onboarding to customer experience by incentivizing providers to
the program.73 During the registration, additional offer the best services, while allowing governments
information will be requested to exclude well-off to cut on costs and avoid lengthy procurement
segments of the population. processes.75 For recipients opting to receive physical
The exclusion criteria will be based on notes, identification information, including biometrics
a phased approach whereby in Phase I basic where possible, will be used for authentication at
demographics and information on income and disbursement time, hereby ensuring that the applicant
assets will be collected for immediate exclusion. and whomever is redeeming the funds are the same
Households who pass the eligibility for Phase I will be person.
onboarded and start receiving the benefits based on It is important to note that the payment
their preferred payment modality (see payment sec- mechanism depends on whether GOL opts for
tion). Phase II exclusion will be a recurrent process and a restricted or an unrestricted use of funds. The
will commence 3 months after the onboarding. Detailed above applies to an unrestricted use of funds, whereby
demographics, information on employment, the value recipients can spend the funds without any condition,
of assets and ownership, etc. will be collected at Phase be it digitally or in physical notes. It is not possible to
II for exclusion purposes. The program should con- use transfer to bank accounts as a payment modality
struct interlinkages with other governmental databases if funds are restricted in their use.
at Phase II depending on technical availability and IT
infrastructure. A call center and a back-office team Monitoring & GRM
need to be developed to perform random spot checks
(up to 3–5 percent of beneficiaries) to decrease poten- An effective GRM and an effective monitoring
tial fraud and/or under/misreporting (which would be system is critical for the success of any social
communicated via a set of outreach and communica- safety net program and for establishing channels
tion activities). The back-office team will also manage of trust with the communities. A back-office
the program’s GRM to ensure fairness, transparency, team needs to be put in place to manage appeals,
objectiveness, and accessibility. complaints, data updates, etc. through a GRM
module.
Payments Mechanism
Upon registering for the program, applicants Complementarities Between the BC-
will be able to choose their preferred payment CT Program and Other SSN Programs
modality among the options currently available
in the Lebanese market, receiving funds: (i) on an The main targeted SSN program in Lebanon is the
individually or jointly held bank account (58 percent National Poverty Targeting Program (NPTP) which
of Lebanese residents being banked);74 (ii) on a is being scaled up. The scale up will be financed
physical or virtual banking card, with funds redeemed
at automated teller machines (ATMs), or digitally
spent for purchases on point-of-sale (POS) devices; 73
For each HH member, a photo of the NID and a selfie will
or (iii) in LBP physical notes, with funds redeemed be taken to create a user profile in the system. Please note
that this option is only valid for polycarbonate type of IDs .
through money transfer operators (MTO) agents 74
Financial Inclusion Demand Side Survey Report, BDL,
and/or LibanPost offices. Other payment modalities, 2018.
including mobile wallets, can be added once available 75
The Future of G2P Payments: Expanding Customer
in the market. The GOL will send the payment orders Choice, CGAP, 2019.
Special Focus I: FX Subsidy Reform in the Deliberate Depression 39
through a US$246 million World Bank supported time-bound. The NPTP/ESSN is a long-term sustained
Emergency Crisis and COVID-19 Response Social SSN program aimed at protecting extreme poor house-
Safety Net Project (ESSN) and support from donors. holds facing multiple shocks. The two programs, which
The ESSN aims to provide cash transfers and access to can carefully be run in parallel, are complementary to
social services to 147,000 extreme poor and vulnerable each other with gradual dovetailing over the next 4–5
Lebanese households (approximately 20 percent years, especially as the necessary delivery systems are
of the population), as well as top-up cash to 87,000 developed. The BC-CT program is likely to be financed
children ages 13–18 who are at risk of dropping out by GOL and will entail a gradual reduction in coverage
from school. In addition, funding from donor partners over 5 years. By year 5, the level of coverage would be
will support an additional 50,000 households reaching in line with a sustainable GOL-financed social safety net
a total of 200,000 of the poorest households—i.e., 27 program as the current crisis subsides. The 4–5-year
percent of the Lebanese population. horizon will also enable a smooth (rather than abrupt)
While the BC-CT and the ESSN have simi- reduction of coverage, as well as the establishment of
larities and can both be considered types of SSN a fully functional SSN delivery system (including social
programs, they have different (albeit related) registry, GRM, and payments system). The proposed
objectives and could be run in parallel. The timeframe is also suitable for the transition to a new
BC-CT is a compensation for subsidy reform aimed macroeconomic framework, allowing it to settle in and
at the general public and likely will be temporary and support an adequate and robust social safety net.
40 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
SPECIAL FOCUS II:
PUBLIC SERVICE
DELIVERY IN
THE DELIBERATE
DEPRESSION 76
L
ebanon is enduring a severe and prolonged therefore more dependent on public services; and (ii) it
economic depression. These conditions threatens financial viability and basic operability of the
transverse the financial and economic spheres public services sector by raising its costs and lowering
to directly impact the wellbeing of residents. Critical to its revenues. In this Special Focus, we shed light on 4
this wellbeing is delivery of essential public services. basic public services: electricity, water supply, sanita-
A sharp deterioration in basic services would have tion and education.
long-term implications, incurring permanent damage
to human capital, which should be a matter of grave
concern. Mass migration, loss of learning, poor health The Electricity Sector
outcomes, and lack of adequate safety nets, among
other consequences, would lead to a human capital The electricity sector has been suffering for
catastrophe, from which recovery would be very decades from a financial deficit that required
difficult. Perhaps this dimension of the Lebanese constant annual budgetary transfers of US$1–2
crisis makes the Lebanon episode unique compared billion to EdL. Using simple back-of-the-envelope
to other crises comparators presented in the report. calculations, these cumulative deficits amount to
Lebanon’s public service delivery has long been around US$40 billion as of 2020, a considerable
notoriously deficient relative to its upper middle-income share of public debt.
status. Nonetheless, long-established mitigation mea-
sures and private substitutables have traditionally filled
the gap, particularly for those with economic means.
76
This Special Focus section is a product of the Lebanon
country team, led by Wissam Harake (Senior Economist),
The Deliberate Depression has further undermined this
Sameh Mobarak (Senior Energy Specialist), Amal Talbi
set up via two effects: (i) it has significantly increased (Lead Water Resources Management Specialist), Sally
poverty rates, expanding the demographic that is not Zgheib (Senior Water Supply and Sanitation Specialist),
able to afford these private substitutables, and are and Nathalie Lahire (Senior Economist).
41
While EdL’s generation production has A SECTOR IN CRISIS:
already fallen significantly by 19 percent (yoy)
Annual budget transfers US$1–2 billion
over the first 11 months of 2020, more rolling
Accumulated budget transfers US$40 billion, almost half of
blackouts are expected, starting in April, if extra-
as of 2020 the national debt
budgetary allocations of 1,500 billion LBP ($996
Generation decline 19% year-on-year over the first
million at the official rate) for EdL to purchase fuel 11 months of 2020
are not approved by parliament. This reinforces
Collection losses increase 20%
existing socio-economic inequalities. Meanwhile, over 2020
EdL revenues, which are in Lebanese pounds, are
EdL cashflow decline 50% compared to 2019
shrinking because of increasing technical, commer-
EdL accrued arrears to private $320 million at the end of 2020
cial and collection losses that exacerbate the impact sector power barges, operation
of the related non-cost recovery tariffs that have and maintenance contractors, and
remained unchanged since 1994. This financial drain distribution service providers
in 2020—collection losses of 20 percent and decline Private diesel generators Cost of up to US¢30/kWh, 13,200
labor force
of EdL cashflow of 50 percent compared to 2019—is
expected to get worse. EdL is likely to increase rolling
blackouts to manage its cashflow shortfalls.
The fragile condition of EdL is exacerbated of private generators. As such, consumers may start
by a severe shortage of foreign currency needed cutting back on using private generators because of
for maintenance of power plants and purchase economic pressures, which may affect this industry’s
of temporary generation. There are already estimated 13,200 labor force.
accrued arrears to the power barges, operations and Urgent action to address these issues is
maintenance contractors, and distribution service needed to avoid the sector’s complete collapse
providers, estimated to be US$320 million at the end in the immediate future. In the short-term, there
of 2020. Many of these private sector contractors is a need to address EdL’s cash shortfall to avoid
have threatened termination of their contracts unless increased blackouts.
paid. Suspension of operations of these contrac-
tors will dramatically affect electricity production; 15
percent is generated from the Karadeniz barges and Water and Sanitation
40 percent from Deir Amar and Zahrani power plants
operated and maintained by PrimeSouth. Moreover, The WEs have witnessed serious depletions in
contracts for the barges and distribution service supplies, revenues, and financial and human
providers end in September and December 2021, resources. At the same time, WEs are affected by an
respectively, without clarity on the plan going forward. upward spiral in costs in the aftermath of the economic
To manage decreasing supplies from EdL crisis (since October 2019), the ongoing COVID-19
over the past year, consumers have been forced pandemic (since March 2020), and the Beirut Port
to increase their dependence on expensive and explosion (August 4, 2020). Further, the water supply
highly polluting private diesel generators, which and sanitation (WSS) service delivery has suffered
have an estimated cost of up to US¢30/kWh. from weakened institutions, limited mobility (due to
This is expected to intensify as EdL’s increases its COVID-19 lockdown), lack of funds, limited human
dependence on rolling blackouts as a cash manage- resources, and reduced subsidies due to the economic
ment tool. To make matters worse, as the economic crisis. The WEs are struggling to sustain their level of
crisis continues to deteriorate and the BdL’s foreign services by increasingly depending on their own funds
currency reserves become increasingly depleted, and the short-term support available from humanitarian
the Government is seriously considering lifting diesel actors (UN and NGOs) that help maintain the pumping
subsidies, which would significantly increase the cost systems and purchases of chlorine for water treatment
42 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
plants. Moreover, due to ongoing water shortages and the delivery of WSS services through delays in
intermittent water supply from the WEs—with hours addressing breakages and repairing leaks and
of supply further reduced in 2020—people have had assets, possibly leading to the non-operation of
to rely more on other costlier water alternatives such treatment plants. WEs might start diverting funds
as water tankers and bottled water. The cost of these from resource monitoring to other immediate O&M
alternatives is increasing. For example, since 2020, needs. Consequently, there is a risk of intensified
the cost of supply from water tankers increased by spread of water-borne diseases, adversely impacting
about 35 percent, while prices of bottled water almost an already vulnerable public health system. Thus,
doubled. there is urgent need for the following:
More specifically, WEs are facing the fol-
lowing challenges: • Reliable electricity supply is one of the main
limiting factors in ensuring operational continuity
• Lack of coverage for operation and maintenance of WSS service delivery. WEs require electricity
(O&M) costs further exacerbated by poor and to produce, treat and distribute water supply
falling collection rates. Even prior to the economic and continue treating wastewater. Thus, the
crisis, the applicable yearly flat fee (approximately Government of Lebanon should prioritize power
LBP 295,000) already failed to cover the full supply to WEs during this crisis despite delayed
cost of O&M. The crisis further exacerbated the payments to avoid shutting-off of the pumping
situation, leading to a 20 percent fall in collection stations and treatment plants.
in 2020 compared to 2019.77 • Maintaining the WSS services at a minimum
• Increase in service delivery costs. Financial crisis operating capacity to safeguard public health.
conditions—including FX shortages and sharp The ongoing humanitarian support can include
depreciation of the local currency—along with an contracting out emergency rehabilitation of WSS
inefficient FX subsidy on fuel imports compounded systems to construction firms (where and when
by rising global oil prices, have led to fuel supply possible) and providing spare parts (e.g., pumps,
shortages in the country. All this has resulted in an motors, pipes, valves, fittings, repair clamps,
increase in the service delivery costs. transformers) and consumables (e.g., chemicals for
• Poor incentives for staff. Staff are paid in LBP. water treatment and disinfection, fuel, lubricants).
As such, very high inflation rates have resulted in • Continue sustaining the flow of funds to procure
sharp declines in their real wages. In addition, the the spare parts and consumables. Standby
WEs have struggled to pay salaries and statutory financing mechanisms from the Government
commitments to retirees resulting in an adverse are needed to sustain the flow of spare parts
impact on staff morale. and consumables to WEs, to maintain personnel
• Delays in implementing ongoing work paid in (e.g., operators and technicians) and must be
US$. The WEs are facing challenges and delays equipped and able to meet the capital costs of
in paying contractors in US$, which are needed system repairs and rehabilitation.
to import the necessary goods and equipment,
thus delaying or blocking the work.
• Limited availability of stock of materials. The Education
available stock of material and equipment at WEs
is limited and will not sustain beyond the next In 2021, Lebanon’s education sector has to
few months. The need to import replenishments respond to five major crises, rather than just
necessitates access to FX.
Should the situation continue to dete- 77
This is according to unaudited 2020 accounts for the
riorate, these challenges will seriously impair Beqaa Water Establishment (BWE).
SPECIAL FOCUS II: PUBLIC SERVICE DELIVERY IN THE DELIBERATE DEPRESSION 43
one—the Syrian crisis, economic collapse, political financing.84 Expected austerity measures implemented
instability, the COVID-19 pandemic, and the PoB by local universities, will lead to a further decline in
explosion—putting severe strain on an already higher education quality, likely exacerbating a brain
struggling education system. Pre-COVID-19 learning drain as youth graduates seek employability abroad,
levels were already comparatively low, with only particularly in critical sectors such as medicine.
6.3 years of learning taking place, once schooling is
adjusted for actual learning.78 The most recent school Impact on learning
closures due to the COVID-19 pandemic—with schools
closed over 75 percent of the school year between The recent Programme for International Student
January 2020 and February 202179—will likely lead to a Assessment (PISA) (OECD 2019) and Trends in
further and significant decrease in learning. Effectively, International Mathematics and Science Study
students are facing a “lost year” of learning. (TIMSS) (IEA 2020) show Lebanon as one of the
lowest ranked countries in the region in terms of
Impact on access to education student learning outcomes. Time trends show that
learning outcomes have consistently declined over
The many crises severely impacted demand the last decade, pointing at systematic issues with
for education and student retention, especially quality of education. While quality of education is low
as many parents can no longer afford private overall, learning outcomes are highly unequal across
education for their children. School completion the country. The differences in the quality of individual
rates in primary (78 percent) and lower secondary schools are very large—more than in other countries—
education (59 percent) have already been declining and disproportionately affect students in public schools
over the last years,80 with completion rates being and those from lower socio-economic backgrounds. The
significantly lower among poorer students.81 The prolonged school closures and interrupted education
burden of education in Lebanon, which enrolled about service delivery will have long lasting negative effects
1.2 million students in 2019/2020, falls on parents’ on learning for all children, exacerbating inequalities
shoulders, who pay a combined US$1.5 billion and impacting the most marginalized.
annually, with the government paying an additional
US$1.2 billion (World Bank 2017).82 Impact on equity in education
This school year alone, 54,000 students
(11 percent of public sector students) transi- The prolonged school closures have
tioned from private to the public schools, putting disproportionately affected the most marginalized
additional strain on a public education sector,
which already faced severe constraints in terms of
available school infrastructure, education quality 78
World Bank Human Capital Index (2020). Learning-
and service delivery.83 It is estimated that through adjusted years of schooling.
this shift, the private sector is losing at least 8 percent
79
This amounts to 154 days of full closure according to
World Bank & UNICEF 2021.
of its financing, either through per-student tuition or 80
Abdul-Hamid and Yassine 2020
direct subsidies that are linked to enrollment. 81
Only half of 18-year-olds from the lowest economic
The contraction in the economy and quantiles completed school prior to the economic crisis
increase in poverty rates will likely lead to more in the country (CAS 2020).
parents shifting their children to public schools 82
Households private expenses comprise fees for private
in the coming years, as well as higher student schools or out-of-pocket expenses in public schools,
such as transportation costs and textbooks.
drop-outs, especially from the most marginalized 83
Most public schools are located in the poorest areas,
households. These large-scale shifts will change the where demand is greatest.
setup of the Lebanese education sector fundamen- 84
World Bank. 2020. “Emergency Social Safety Net Project
tally, necessitating re-evaluation of education sector Appraisal Document.”
44 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
students, who had little access to continued access to broadband and computers needed for
learning. Since February 2020, schools were mostly online education, to resourceful home environments
closed and remote learning the default for most needed to support learning, up to the misalignment
months. Remote learning requires availability of IT and between resources provided by the education sector
other resources. About 60 percent of students either and education needs.
do not have a computer or have to share it with at
least 3 other family members. Recent estimates show 85
Gajderowicz, Tomasz, and Maciej Jakubowski.
that only about 50 percent of students are connected Forthcoming. “PISA In-Depth Analysis of Results
to online learning.85 Inequities in the education system for Jordan and Lebanon (draft).” Washington, DC:
have been further exposed with the pandemic—from World Bank.
SPECIAL FOCUS II: PUBLIC SERVICE DELIVERY IN THE DELIBERATE DEPRESSION 45
ANNEX
Forecasts of Lebanon’s Real GDP Forecasting Real GDP Growth for 2020
Growth using MIDAS Regressions:
An Update for 2020 and 2021 The high frequency indicators used to nowcast and
forecast Lebanon’s real GDP growth in 2020 are:
annual growth rates in claims of the commercial
Introduction banking sector on resident customers, outstanding
lines of credit for imports, non-resident deposits and
The forecasts of Lebanon’s real GDP growth for the resident deposits (Figure 27). That is, in the MIDAS
year 2020 and 2021 are updated based on the new setup, our vector of high frequency indicators is,
incoming data for the high frequency indicators. The x tH = (cl , lc , nr , r ) , where cl, lc, nr and r denote,
data on the high frequency indicators are available for respectively, annual growth rates in claims of the
the first eleven months of 2020 at the time of writing.86 commercial banking sector on resident customers,
The data for real GDP growth are available until 2019. outstanding lines of credit for imports, non-resident
In forecasting growth in 2020 and 2021, we deposits and resident deposits. We also aggregate
make a distinction between the utility of financial versus the information from the four high frequency
real economy indicators. Financial indicators are likely indicators using principal components analysis. More
to better first capture financial crisis dynamics, making specifically, we extract the first principal component
them more relevant leading indicators for 2020 than from the four indicators and use it to forecast real
real economy indicators. However, over the course of GDP growth for 2020. The MIDAS model, which uses
2020, the financial sector became increasingly inop- the first principal component of the four indicators,
erative and segmented from both the local and global
economy. Meanwhile, real indicators increasingly 86
In the previous update to the MIDAS forecasting exercise,
capture the extent of the economic crisis and become the data on the high frequency indicators were available
more relevant leading indicators for 2021. until May 2020.
47
FIGURE 27 • Evolution of High Frequency Indicators Used to Nowcast and Forecast Lebanon’s Real GDP
Growth in 2020
Growth in Claims on Resident Sector Growth in Outstanding Lines of Credit for Imports
30 80
20
40
10
0 0
–10 –40
–20
–80
–30
–40 –120
08 09 10 11 12 13 14 15 16 17 18 19 20 08 09 10 11 12 13 14 15 16 17 18 19 20
Growth in Nonresident Deposits Growth in Resident Deposits
60 30
40 20
20
10
0
0
–10
–20 –10
–40 –20
08 09 10 11 12 13 14 15 16 17 18 19 20 08 09 10 11 12 13 14 15 16 17 18 19 20
is referred to as the factor augmented MIDAS TABLE 6 • Real GDP Growth Forecasts for 2020
model. The low frequency variable of interest in the
Real GDP growth forecasts
nowcasting or forecasting exercises is y tL = (gdpg ) for 2020
where gdpg is the growth rate in GDP.
Baseline
The dynamic (i.e., multi-step-ahead) forecasts
Growth in non-resident deposits –12.6%
of real GDP growth rates are generated using an
ADL-MIDAS model, which is employed to introduce Growth in resident deposits –14.8%
dynamics. The high frequency indicators are available Growth in claims on the resident –13.5%
sector
until November 2020.
Forecasts of real GDP growth for 2020 are pro- Growth in lines of credit for imports –14.4%
duced from the ADL-MIDAS using each of the above Factor Augmented MIDAS –20.3%
high frequency indicators. The forecasts of real GDP
growth are provided in Table 6.
Forecasting Real GDP Growth for 2021 This poses an immediate challenge: The set
of candidate high frequency indicators of economic
Under the assumption that the constraints relating activity that can be used to forecast real GDP
to import demand are less binding in 2021 and that growth is sizeable. Tiffin (2016) employs a machine
the bulk (but not the entirety) of the adjustment in learning approach to identify the best predictors of
the banking sector occurs in 2020, the set of high economic activity. More specifically, Tiffin (2016)
frequency indicators is enlarged to encompass real resorts to random forests, Least Absolute Shrinkage
activity indicators. and Selection Operator ridge and elastic net
48 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
TABLE 7 • Candidate Predictor Variables for the Forecasts of Real GDP Growth for 2021
TABLE 8 •
Real High Frequency Indicators Using Real Activity Indicators
Observations Forecast for 2021
Candidate predictor variables available until
BDL Coincident Indicator (annual change, –19.6%
BDL Coincident Indicator (annual change, percent) (CI) 2020:09 percent) (CI)
World Bank Coincident Indicator (annual change, 2020:09 World Bank Coincident Indicator (annual change, –16.0%
percent) (WBCI) percent) (WBCI)
Cement Deliveries (annual change, percent) (CD) 2020:09 Cement Deliveries (annual change, percent) (CD) –9.9%
Customs Receipts in Real Terms (annual change, 2020:09 Customs Receipts in Real Terms (annual change, –24.5%
percent) (CR) percent) (CR)
Import of Petroleum Derivatives (annual change, 2020:09 Import of Petroleum Derivatives (annual change, –8.7%
percent) (PI) percent) (PI)
Incoming Freight at the Port of Beirut (annual change, 2020:09 Incoming Freight at the Port of Beirut (annual change, –8.7%
percent) (IF) percent) (IF)
Outgoing Freight at the Port of Beirut (annual change, 2020:09 Outgoing Freight at the Port of Beirut (annual change, –7.1%
percent) (OF) percent) (OF)
Primary Spending in Real Terms (annual change, 2020:08 Primary Spending in Real Terms (annual change, –6.7%
percent) (PRIM) percent) (PRIM)
regressions to select the best predictor of quarterly nowcast of 2020 (i.e., it is a dynamic forecast). This
GDP growth from a pool of noisy high frequency translates into more uncertainty. Further, the forecast
indicators.87 The MIDAS setup is different from that of real GDP growth for 2021 will not reflect any
used in Tiffin (2006). Tiffin (2016) predicts quarterly positive developments on the policy front given that it
GDP using quarterly indicators and, therefore, does builds on an extrapolation of time series dynamics.88
not mix low and high frequency data. Nonetheless, a The forecast of GDP growth for 2021 should therefore
subset of the candidate predictor variables that are be used with these caveats in mind. The advantage
entertained by Tiffin (2016) is considered in addition of using a large pool of predictor variables is the
to the four high frequency indicators considered ability to generate a large set of forecasts of real
previously. The candidate predictor variables are GDP growth, which can then be combined. This will
provided in Table 7 and illustrated in Figure 28. attenuate uncertainty related to the forecast.
The nominal series are deflated by the As noted in Timmermann (2006), combining
Consumer Price Index (CPI). The data for the CPI are forecasts is desirable for a number of reasons.89 First,
available starting in January 2008. The availability of identifying the best performing model is not a straight-
the CPI data dictates the starting date of the MIDAS forward endeavor. Therefore, combining forecasts
forecasting exercise to be January 2009. The same provides diversification gains. Second, the combined
starting date is employed for all of the models to place forecast is more robust to structural breaks in the
them on an equal footing. individual forecasting models. Third, given that every
Forecasting Lebanon’s real GDP growth for
2021 is more complicated and subject to considerably
more uncertainty than nowcasting real GDP growth 87
Tiffin (2016) notes that machine learning methods
for 2020. To start with, none of the high frequency are particularly adept at prediction and that the best
predictor is determined based on its out of sample
indicators are observed for 2021. Therefore, monthly
predictive accuracy.
forecasts of the four high frequency indicators for 88
This extrapolation embeds mean reversion, but this is not
the year 2021 should be generated. In addition, the sufficient to reflect the positive effects of policy action.
forecast of real GDP growth for 2021 builds on the 89
This discussion is based on Jamali and Yamani (2019).
Annex 49
model is likely to be misspecified, combining forecasts 2016). Fourth, Timmermann (2006)’s synthesis of the
will alleviate the effects of misspecification in indi- empirical literature suggests that combining forecast
vidual forecasting models (Elliott and Timmermann, yields gains in predictive accuracy relative even to
FIGURE 28 • Growth of High Frequency Real Economy Indicators Used to Nowcast and Forecast
Lebanon’s Real GDP Growth in 2021
CIG WBCIG
40 40
20 20
0
0
–20
–20
–40
–40 –60
–60 –80
2008 2010 2012 2014 2016 2018 2020 2008 2010 2012 2014 2016 2018 2020
CDG CRG
80 120
80
40
40
0
0
–40
–40
–80 –80
2008 2010 2012 2014 2016 2018 2020 2008 2010 2012 2014 2016 2018 2020
PIG IFG
100 60
75 40
50 20
25 0
0 –20
–25 –40
–50 –60
2008 2010 2012 2014 2016 2018 2020 2008 2010 2012 2014 2016 2018 2020
OFG PRIMG
80 120
60
80
40
20 40
0 0
–20
–40
–40
–60 –80
2008 2010 2012 2014 2016 2018 2020 2008 2010 2012 2014 2016 2018 2020
50 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
the best performing individual forecasting model. The Forecasts of Real GDP Growth for
TABLE 9 •
simple mean, the trimmed mean and the median are 2021 Using Real Activity and Financial
Indicators
three simple forecast combination methods that can
be applied in this setup. Forecast for
Dynamic forecasts of the growth in the high Indicator 2021
frequency indicators are generated from a well- BDL Coincident Indicator (annual change, percent) (CI) –19.6%
specified Autoregressive Moving Average model. The World Bank Coincident Indicator (annual change, –16.0%
forecast sample begins on the month following the percent) (WBCI)
last for which an observation on the high frequency Cement Deliveries (annual change, percent) (CD) –9.9%
indicator is available. The forecast sample for the Customs Receipts in Real Terms (annual change, –24.5%
high frequency indicators ends in December 2021. percent) (CR)
The set of high frequency candidate predictors is Import of Petroleum Derivatives (annual change, –8.7%
x tH = (ci , wbci , cd , cr , pi , if , of , prim ). percent) (PI)
The time series dynamics of the high frequency Incoming Freight at the Port of Beirut (annual change, –8.7%
indicators of economic activity are provided. percent) (IF)
The simple average of the forecasts for 2021 is Outgoing Freight at the Port of Beirut (annual change, –7.1%
–12.65 percent whereas the median is –9.32 percent. percent) (OF)
The growth forecasts using the main high Primary Spending in Real Terms (annual change, percent) –6.7%
frequency indicators are provided in Table 9. (PRIM)
Again, given that the import constraint is likely Growth in non-resident deposits (NR) –10.5%
not binding in 2021, the GDP growth forecast for 2021 Growth in resident deposits (R) –11.3%
obtained from the growth in lines of credit for imports as
Growth in claims on the resident sector (CL) –8.9%
a high frequency indicator is dropped from the forecast
Growth in lines of credit for imports (LC) –12.6%
combination. Combining the forecasts from Tables 8
and 9 yields an average growth rate of –11.6 percent in
2021 and a median growth rate of –9.90 percent. If the
forecast of real GDP growth obtained with lines of credit Dynamic Response of Inflation to
for imports is included in the forecast averaging, the Currency in Circulation in Lebanon
average growth rate for 2021 becomes –12.05 percent
while the median growth rate is –10.22 percent. This section offers an assessment of the response of
inflation to changes in currency in circulation using
References multivariate time series models. The variables of
interest are the percentage change (growth) in the
Elliott, G., & Timmermann, A 2016. Economic currency in circulation, inflation and the percentage
Forecasting. Princeton University Press. change in the World Bank coincident indicator. The
Jamali, I., & Yamani, E. 2019. Out-of-sample exchange coincident indicator is employed to control for real
rate predictability in emerging markets: economic activity.
Fundamentals versus technical analysis. Journal
of International Financial Markets, Institutions Unit Root Tests
and Money, 61, 241–263.
Tiffin, A. 2016. Seeing in the dark: a machine-learning We begin the exercise by testing for stationarity for each
approach to nowcasting in Lebanon. IMF Working of the series using the augmented Dickey–Fuller (ADF)
paper WP/16/56, International Monetary Fund, (1979) and the Phillips and Perron (PP)(1988) tests. The
Washington DC. null hypothesis for both the ADF and PP tests is that the
Timmermann, A. 2006. Forecast combinations. Hand- series contains a unit root. The ADF test is known to
book of Economic Forecasting, 1, 135–196. exhibit low power when the alternative is near unit root
Annex 51
behavior (Elliot et al., 1996). Therefore, the ADF test with TABLE 10 • Unit Root Tests
GLS detrending of Elliott et al. (1996) is also employed.
ADF PP ADF-GLS ZA
The existing literature shows that the ADF–GLS test has
Growth of currency in –3.29 –11.29*** –2.62 –5.49***
good power properties against near unit root behavior. circulation
The crisis dynamics pervading the post
Inflation –4.33*** –7.46*** –4.18*** –6.27***
October 2019 period imply that the series might be
subject to a structural break. Therefore, the Zivot and Percentage change in the –12.96*** –13,08*** –4.01*** –5.36***
World Bank coincident
Andrews (ZA) (1992) test, which allows for a break in indicator
the intercept and the trend, is also employed. Notes: This table provides the results from the Augmented Dickey Fuller (ADF), Phillips
The results from the unit root tests suggest that and Perron (PP), Elliott, Rothenberg and Stock (1996) ADF test with GLS detrending
(ADF-GLS) as well as the Zivot and Andrews (ZA) (1992) tests.
all three series are stationary. The tests also indicate
that the growth in currency in circulation can be char-
acterized as a stationary variable that is subject to a in circulation is significant. More specifically, the
structural break. The Perron (2006) test designates response of inflation peaks five months following the
August 2019 as the break point.90 shock and remains positive and significant. The growth
in the WBCI exhibits a negative response to the shock
VAR Variables in Difference to currency in circulation. This response is significant
in the two to five months following the shock.
Let M0t, Pt and WBCIt denote, respectively, the The cumulative effect of the shock to the cur-
level of the currency in circulation, the CPI and the rency in circulation on inflation for the entire forecast
World Bank Coincident Indicator index. The natural horizon (12 months) can be computed by accumu-
logarithms of the currency in circulation and the lating the responses in inflation. The effect is provided
price level are denoted, respectively, mot and pt. The in Table 11. The results suggest that a 100 percent
percentage change in the WBCIt is computed as increase in the currency in circulation increase prices
WBCIt −WBCIt −1 . by 129 percent annually, which averages about 10.75
wbgt = 100 x
WBCIt percent per month.
The vector of variables in the VAR is thus
y t = [ Δm0t , Δpt ,wbgt ]’. Namely, the vector of variables
VAR Variables in (Log) Levels
yt includes the percentage change in the currency in
circulation, inflation as measured by the percentage
The robustness of the results is assessed by estimating
change in the CPI, and lastly, the percentage change
the VAR model in log levels. More specifically, the VAR
in the World Bank Coincident Indicator index. The
in log levels comprises in the following order: mot, pt
VAR model is estimated using percent changes in
and wbcit, where wbcit is the natural logarithm of the
the variables to circumvent non-stationarity.
A recursive ordering (Cholesky decomposi-
tion) is employed to identify the VAR. An optimal lag Cumulative Effect of a 1 Percent
TABLE 11 •
Increase in Currency in Circulation on
length of five is used based on the Akaike Information Inflation
Criterion (AIC). The analysis is carried out at the
monthly frequency over the period January 2008 to Percentage increase in currency
in circulation Cumulative effect on CPI
September 2020. The times series of the variables are
displayed in Figure 29. 1% 1.29%
The effects of a shock to currency in circula- 100% 129%
tion are assessed using impulse response analysis.
Figure 30 provides the responses of each of the vari-
ables to a 1 percent shock to currency in circulation. 90
Other unit root tests that account for a break include Lee
The response of inflation and the World Bank and Strazicich (2003) as well as Lumsdaine and Papell
coincident indicator to a 1 percent shock to currency (1997).
52 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
Time Series Dynamics of the Variables in Changes
FIGURE 29 •
Growth of Currency in Circulation
30
25
20
15
10
5
0
–5
–10
2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
Inflation
25
20
15
10
5
0
–5
2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
Change in the World Bank Coincident Indicator
15
10
5
0
–5
–10
–15
–20
–25
–30
2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
Notes: This figure provides the time series of the variable used in the estimation of the VAR. The shaded area is the post October 2019 period, which corresponds to the period of the
financial crisis.
World Bank coincident indicator.91 The variables in The effect of a 1 percent increase in the
log levels are displayed in Figure 31. currency in circulation on inflation is gauged by exam-
Again, the effects of a shock to currency ining the response of the logarithm of the CPI. The
in circulation are gauged using impulse response
analysis. Figure 32 provides the responses of
91
The variables in log levels are non-stationary. However,
each of the variables to a 1 percent shock to cur-
empirical analyses of the macroeconomic effects of
rency in circulation. The sample period is January monetary policy shocks commonly use a specification in
2008 to September 2020 and the optimal lag log levels. See, for example, Faust (1998) and Christiano,
length of the VAR is determined using the AIC. Eichembaum and Evans (1999). Such a specification
Annex 53
Responses to a 1 Percent Increase in the Growth of Currency in Circulation
FIGURE 30 •
Response of MOG to MOG
1.00
0.75
0.50
0.25
0.00
–0.25
0 1 2 3 4 5 6 7 8 9 10 11
Response of INF to MOG
0.4
0.3
0.2
0.1
0.0
–0.1
–0.2
0 1 2 3 4 5 6 7 8 9 10 11
Response of WBCIG to MOG
0.2
0.1
–0.0
–0.1
–0.2
–0.3
–0.4
–0.0
0 1 2 3 4 5 6 7 8 9 10 11
Notes: This figure provides the responses of each of the variables to a 1 percent shock to currency in circulation. The 95 percent confidence intervals are constructed using the
bootstrap method.
results are summarized in Table 12. A 100 percent Cumulative Effect of a 1 Percent
TABLE 12 •
increase in the currency in circulation increase prices Increase in Currency in Circulation on
by 76 percent, annually averaging 6.33 percent per Inflation
month. Percentage Increase in Currency
in Circulation Effect on CPI
The Cointegrated VAR: Vector Error 1% 0.76%
Correction Model 100% 76%
The response of inflation to a shock to currency in
circulation is examined next via a cointegrated VAR is not invalid, but care must be exercised to account for
or vector error correction model. Figure 31 suggests cointegrating relation if such long-run relations are present.
54 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
Time Series Dynamics of the Variables in Log Levels
FIGURE 31 •
log of Currency in Circulation
10.5
10.0
9.5
9.0
8.5
8.0
7.5
7.0
2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
log of CPI
6.0
5.8
5.6
5.4
5.2
5.0
4.8
4.6
2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
log of WBCI
5.25
5.00
4.75
4.50
4.25
4.00
3.75
2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
that the logarithms of the currency in circulation and where r is the number of cointegrating vectors under
–Τ∑ i =r +1ln ˆi )is the estimated ith ordered
(1– λ
g
the CPI may exhibit a long-run cointegrating relation. λ trace (
the r)=
null hypothesis and
The presence of a cointegration relation in eigenvalue of the matrix Π. The trace statistic tests
the VAR in log levels is tested using the Johansen the null hypothesis that the number of cointegrating
(1988) approach. More specifically, the existence vector is r or less against the alternative hypothesis
of cointegrating vectors can be examined using the that there are more than r cointegrating vectors. The
trace statistic: trace statistic is reported in Table 13.
The null hypothesis of no cointegrating vector
ˆi ) ,(5)
λtrace (r ) = –Τ∑ i =r +1ln(1– λ
g
is marginally rejected (at the 10 percent level). The
null hypothesis of one or less cointegrating vector
Annex 55
Responses to a 1 Percent Increase in the (log of) Currency in Circulation
FIGURE 32 •
Response of LMO to log of MO
1.50
1.25
1.00
0.75
0.50
0.25
0.00
0 1 2 3 4 5 6 7 8 9 10 11
Response of LCPI to log of MO
1.0
0.8
0.6
0.4
0.2
0.0
–0.2
0 1 2 3 4 5 6 7 8 9 10 11
Response of LWBCI to long of MO
0.2
–0.0
–0.2
–0.4
–0.6
–0.8
–1.0
0 1 2 3 4 5 6 7 8 9 10 11
is not rejected. The null of two or less cointegrating A VECM is estimated and the response to a one
vectors is rejected at the 5 percent. standard deviation shock are provided in Figure 33.
The 95 percent confidence bands are generated
using Monte Carlo simulation.
The impulse response analysis suggests that
TABLE 13 • The Johansen (1988) Trace Statistic
a 1 percent shock to currency in circulation gener-
Trace statistic ated a response of 0.80 percent increase in CPI over
r =0 28.30* a twelve-month horizon. This result is summarized in
Table 14. Hence, an increase in currency in circulation
r ≤1 12.86
by a 100 percent results in an 80 percent increase in
r ≥2 4.66**
56 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
Responses to a 1 Percent Increase in (the Log of) Currency in Circulation from a VECM
FIGURE 33 •
Response of log of MO to log of MO
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0 5 10 15 20 25 30 35
Response of log of CPI to log of MO
0.30
0.25
0.20
0.15
0.10
0.05
0.0
–0.05
0 5 10 15 20 25 30 35
Response of log of WBCI to log of MO
0.05
0.00
–0.05
–0.10
–0.15
–0.20
–0.25
–0.30
–0.35
0 5 10 15 20 25 30 35
the price level, annually, which averages about 6.66 References
percent per month.
Christiano, L. J., Eichenbaum, M., & Evans, C. L.
1999. Monetary policy shocks: What have
TABLE 14 •
Cumulative Effect of a 1 Percent we learned and to what end?. Handbook of
Increase in Currency in Circulation on Macroeconomics, 1, 65–148.
Inflation Elliott, G., Rothenberg, T. J., & Stock, J. 1996.
Percentage increase in currency in circulation Effect on CPI Efficient tests for an autoregressive unit
root. Econometrica, 64(4), 813–36.
1% 0.80%
Faust, J. (1998). The robustness of identified
100% 80%
VAR conclusions about money. In
Annex 57
Carnegie-Rochester conference series central bank, however, was not properly equipped to
on public policy (Vol. 49, pp. 207–244). supervise a banking sector that grew precipitously.
North-Holland. At the monetary policy level, a crawling peg
Johansen, S. 1988. Statistical analysis of cointegration system became a main stabilization tool to rein in
vectors. Journal of Economic Dynamics and inflation; pre-announced devaluations of the Perso was
Control, 12(2–3), 231–254. hoped would reduce uncertainty and guide tradable
Perron, P. 2006. Dealing with structural breaks. prices along international prices (with non-tradables
Palgrave Handbook of Econometrics, 1(2), following). Exchange rate policies “had important
278–352. effects on the soundness of the financial system,
Lee, J., & Strazicich, M. 2003. Minimum Lagrange directly by influencing the capital flows and the value
multiplier unit root test with two structural of the foreign debt of firms, and indirectly by dramati-
breaks. Review of Economics and Statistics, cally changing many relative prices in the economy—in
85(4), 1082–1089. particular, asset prices” (Balino 1991, pg. 71).
Lumsdaine, R. L., & Papell, D. H. 1997. Multiple trend Enterprise indebtedness increased signifi-
breaks and the unit-root hypothesis. Review of cantly, leading to business failures and soaring levels
Economics and Statistics, 79(2), 212–218. of NPLs, which increased from 1.7 percent in 1975
Phillips, P. C., & Perron, P. 1988. Testing for a unit root to 11.7 percent in 1980 in the primary sector and
in time series regression. Biometrika, 75(2), from 3 percent in 1975 to 12.8 percent in 1980 in the
335–346. manufacturing sector (Balino 1991). The rapid dete-
Sims, C. A. 1980. Macroeconomics and reality. rioration in the bank-lending portfolio in 1980 spelled
Econometrica 48(1), 1–48. trouble for the banking sector.
Zivot, E., & Andrews, D. W. K. 2002. Further evidence The crisis broke on March 18, 1980 with the
on the great crash, the oil-price shock, and the failure of Banco Intercambio Regional, one of the
unit-root hypothesis. Journal of Business & largest private banks in the country. This failure
Economic Statistics, 20(1), 25–44. struck at confidence in the domestic private banking
model, prompting a reallocation of deposits to State
banks and foreign affiliates. Further, the interbank
Global Financial Crises Episodes lending market froze. Despite a retroactive increase
in deposit insurance, uninsured Peso deposits
The Argentinian Banking Crisis of 1980 exceeding 100 million Pesos were lost. Dollar deposi-
tors in failed institutions suffered a complete loss. The
Prior to 1976, the Argentinian economy contended central bank was forced to intervene to rescue three
with high inflation, balance of payments pressures major banks, two of which were ultimately liquidated
and fiscal deficits. Since 1976, and in a bid to address (Balino, 1991).
these chronic imbalances, policymakers undertook The adverse developments in the banking
short-term and structural reforms that predominantly sector strained the crawling peg regime. The
affected the financial sector. exchange rate devalued by about 23 percent on
A rapid liberalization of the financial sector June 2 owing to losses in foreign exchange reserves
was in full swing in 1977. The central bank gradually at the central bank. A two-tiered exchange rate regime
loosened prudential regulations relating to capital emerged: a commercial rate set by the central bank
requirements, asset immobilization and limits on bank and a market determined rate for financial markets.
lending. It also eased its heavy regulatory oversight Confused and inconsistent exchange rate policies
on the establishment of new banks as well as new ensued by subsequent governments; first liberaliza-
branches of existing banks. The changes effectively tion and unification in December 1981, and then the
transformed the Argentinian banking sector from reintroduction of a dual exchange rate system in July
a 100 percent to a fractional reserve system. The 1982, before unification again in November 1982.
58 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
In response to the crisis, authorities pursued a The large capital inflows stoked inflationary
three-pronged solution. “First, they had to take emer- pressures and pushed real lending rates below zero
gency measures to avoid a bank panic; second, they in 1980, further enticing credit growth and a misal-
had to search for longer-term solutions to the private location of resources. An overreliance by enterprises
debt problem and its effects on financial institutions; and the nongovernment sector on debt to finance
and third, they had to find ways to restructure the finan- expansion doubled the outstanding debt between
cial system.” (Balino 1991, pg. 92). The emergency 1972 and 1980 and quadrupled foreign currency
measures included liquidating Banco Intercambio denominated debt. The external imbalances and the
Regional, establishing a new credit facility as well sharp rise in foreign savings as a percent of GDP
as retroactively increasing the deposit insurance. made the Philippines economy vulnerable to shocks.
The central bank also intervened with three troubled In January 1981, confidence in the financial
banks on April 28, 1980. The authorities addressed system of the Philippines was shaken by fraud in the
the private debt problems via maturity extensions commercial paper market—an episode that became
and refinancing schemes. Despite the central bank’s known as the Dewey Dee affair. The confidence crisis
success in stemming the run on deposits, the crisis rapidly spread from the periphery of the financial
had far-reaching consequences and resulted in the sector to its core. Rippling through the system, it first
liquidation of 71 financial institutions between 1980 caused a collapse of the commercial paper market,
and 1982. failures of nonbank money market institutions and
the crash of the two largest investment houses
The Philippines Financial Crisis of 1981 belonging to two major holding companies, which
also subsequently failed. Wealth holders reacted by
The Philippines crisis followed a boom-bust shifting savings to the commercial banking sector.
phenomenon typical of those analyzed in crisis The loss of confidence consequently afflicted the
literature (Nascimento 1991). Over the 1970’s, The thrift banking system and precipitated failures in
Philippines experienced high growth rates in its real rural banks. As the year progressed, it became
Gross National Product (GNP). It was a demand- readily apparent that the Dewey Dee affair had
driven economic boom that reflected an investment- lasting effects on public trust in the financial system
intensive development strategy by the Government. as it triggered bank runs, capital flight and a “flight to
In order to finance this boom and attract quality” towards the (perceived) relative safety of the
foreign capital, The Philippines authorities and the commercial banking sector.
Central Bank of the Philippines (CBP) embarked In this first phase of the crisis, between 1981
over the 1972–81 period on a liberalization pro- and 1983, authorities’ priority focused on containing
gram in the financial sector and the capital controls the spread of financial distress. The CBP extended
regime. The authorities and the CBP relaxed emergency lending to quasi banks through a special
controls on foreign capital, facilitated the entry of rediscount facility. An Industrial Fund, co-financed by
foreign banks and encouraged the establishment Government budget and the CBP, was also estab-
of “universal” banks—commercial banks whose lished and, in early 1982, replaced with a special
activities encompass, in addition to retail banking, rediscount window. The CBP used the rediscount
securities transaction and investment banking. The window to extend medium and long-term loans
CBP introduced important changes to prudential to universal banks and to the Government-owned
regulations by lowering banks’ capital requirement Philippines National Bank and Development Bank
from 15 percent in 1972 to 6 percent in 1980, of the Philippines. These two banks accounted for a
thereby encouraging loose lending practices. significant share of credit to the private sector. The
An investment (and credit) boom ensued and the Government also provided emergency lending and
external debt burden increased from 31.3 percent equity contributions to nonfinancial public corpora-
of GDP to 48.9 percent in 1980 (Nascimento 1991). tions from 1981 to mid-1983 (Nascimento 1991).
Annex 59
However, increasing uncertainty in the political Bank of International Settlement (on behalf of main
climate and unfavorable external conditions caused central banks), and commercial banks, both large
a balance of payments crisis in the second half of and small. Significant linkages to the global financial
1983. Faced with these adverse developments, the system implied serious implications of a Mexican
Government declared a moratorium on external debt default to its commercial creditors; indeed, this fact
repayments on October 1983. This renewed the run ironically bestowed some leveraged on the Mexican
on banks and deprived the Philippines from external negotiating position. The role of commercial banks in
financing. this mechanism can be considered as a progenitor of
The CBP injected reserve money to meet the later participations in crisis resolutions (i.e., Greece,
increase in currency demand and provided extensive bail ins etc.).
monetization of the fiscal deficit. This policy stance In 1970s, on prospects of expanded volumes
caused inflation to jump from 12 percent in 1983 to of oil exports, the Mexican Government borrowed
50 percent in 1984, and led to a crowding out of the heavily in foreign currency from commercial markets to
private sector. Inflationary pressures and accelerated finance a rapid rise in public expenditures (Dornbusch
depreciation in the Peso in 1983 and 1984 exac- and Werner 1994; van Wijnbergen 1991). The rise in
erbated financial difficulties of the corporates and expenditures stoked inflationary pressures, caused
increased distress among banking institutions. an appreciation in the real exchange rate and led to
The Government extended emergency lending a rapid accumulation of debt; inflation averaged 20.6
and equity contributions to Government financial percent over the 1972–1981 period (Dornbusch
institutions from 1983 to 1985. The assistance aimed and Werner 1994). Meanwhile, the external debt
at financing acquisitions of distressed institutions and of the public sector, a significant part of which was
facilitating the conversion of debt owed to Government short-term, increased at an annual rate exceeding 30
financial institutions into equity (Nascimento 1991). percent from US$4 billion in 1973 to US$43 billion in
The volume of CBP assistance soared from 1981 (Boughton 2001).
3,054 million Pesos in 1981 to 13,549 million Pesos The Mexican economy faced a significant external
in 1985 (Nascimento, 1991). The financial crisis, shock as the price of oil, its major export, declined by
which erupted in 1981 and lasted until mid-1987, 65 percent between 1981 and 1986 (Cantu, Park and
resulted in a major contraction in the financial system. Tornell 2015) due to a softening in the demand for oil
In total, the authorities intervened with 128 rural (Boughton 2001). The tightening of monetary policy in
banks, 32 thrift institutions and 2 private banks. The the US and Europe—in a bid to combat inflation—in the
Government-owned commercial banks, PNB and late 1970s pushed world interest rates higher than 15
DBP, became practically insolvent in 1986 and their percent (Cantu, Park and Tornell 2015) exacerbating
non-performing assets were transferred to the Asset Mexico’s financing difficulties.
Privatization Fund, a special purpose entity created On February 17, 1982, the authorities
to manage impaired assets. Between 1985 and 1986, announced that the central bank will temporarily
CBP finally brought inflation under control by main- withdraw from the foreign exchange market. The
taining a tight policy. Peso immediately depreciated by more than 40 per-
cent (Boughton 1991), worsening Mexico’s external
The Mexican Debt Crisis of 1981–82 debt servicing prospects. By April 1982, capital flight
accelerated and Mexico’s largest conglomerate,
The intrigue and distinction of the Mexican debt crisis Alfa Industrial Group, defaulted on debt payments of
rest with its globally systemic nature that necessitated US$2.3 billion (Boughton 1991).
a complicated multi-stakeholder coordination and co- On April 20, Mexico’s Minister of Finance,
financing mechanism. This mechanism involved, in Silva Herzog, announced a stabilization program that
addition to Mexican authorities, the US Government, lowered the fiscal deficit by 3 percent by year-end.
multilateral organizations, especially the IMF and the This announcement, however, was insufficient to
60 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
stem the crisis. Faced with dwindling foreign cur- Bank of International Settlements was announced on
rency reserves, the Bank of Mexico drew US$800 August 28. The IMF conditioned a Mexico program
million on a swap line with the Federal Reserve. A on a clear contribution by private creditors, which
large debt repayment was looming in August 1982 involved rolling over the debt and extending further
and prospects for meeting this obligation appeared credit to close a US$ 7 billion funding gap. Once this
impossible (Oks and van Wijnbergen 1994). In July, was secured, an IMF Extended Fund Facility (EFF) of
commercial banks expressed increasing reluctance US$3.75 billion was announced on December 23.
to roll over their maturing debt or to extend more
funding to Mexico. This caused a widening in spreads The Chilean Banking Crisis of 1981
relative to the LIBOR.
Mexico’s political transition, from outgoing Following the overthrow of the left-leaning,
President Lopez Portillo to President-elect Miguel de democratically elected President Salvador Allende by
La Madrid following presidential elections on July 4, General Augusto Pinochet on September 11, 1973,
1982, complicated and increased vagueness around Chile embarked on an intensive market-oriented
political consensus on a crisis management strategy. liberalization of the economy; indeed, “the dictatorship
It was not clear the extent to which Herzog had a that supplanted Allende used its powers to open the
political mandate from his superiors to structure a economy and to give market economics as free a rein
strategy and negotiate with the IMF, US and other as Latin America had ever seen” (Boughton 2001, pg.
counterparts; decisions made in Mexico City contra- 346). This developed to be one of the more prominent
dicted those made by the Mexico negotiating team macroeconomic liberalization case studies, especially
(Boughton 1991). espoused by free-market proponents.
By August, the crisis was in full swing. On Prevailing conditions at the time of the coup
August 4, 1982, the Bank of Mexico drew a three- included a stagflationary economy with rising
month loan of US$700 million on a swap line with the government expenditures, import controls and an
Federal Reserve. A dual exchange rate system was overvaluation of the currency, with black-market rates
announced on August 5 with the aim of restricting that were over 10 times the official rate. Pinochet
speculative capital flows. At that point, it became handed economic management to the so-called
apparent that Mexico’s debt problems could not be “Chicago boys”—a group of Chilean economists
resolved solely via negotiations with commercial educated at the Department of Economics of the
creditors. Instead, on August 12, authorities closed University of Chicago. The economic team, led by
the foreign exchange market, restricted banks’ for- finance minister Sergio de Castro, undertook fiscal
eign exchange operations to the Bank of Mexico at consolidation and structural reforms that succeeded
a rate that was less favorable than the market rate, in raising growth to 7.5 percent. The reform program
and paid out dollar-denominated deposits in Pesos. was supported by the IMF in January 1974 via a
In addition, commercial creditors, the US Treasury, one-year stand-by arrangement (SBA) of about $95
and the Federal Reserve were notified that Mexico is million as well as a Compensatory Financing Facility
unable to honor the principal debt payment. (CFF) that allowed for weathering the effects of export
The implication of a Mexican default on its shortfalls and the 1973 oil supply shock.
international commercial creditors directly threatened The economic team proceeded with the liber-
the global financial system. Faced with that fact, a alization of domestic markets, including the following:
concerted, multilateral funding effort was required removal of controls on interest rates and credit
to stem the crisis. Such an effort was underway by growth; lowering of reserve requirements; reduction
August 15. The US Treasury extended a line of credit of barriers to entry into the banking sector, including
of US$1 billion to Mexico, while the US Department of for foreign banks; the privatization of previously
Agriculture offered credit guarantees of US$1 billion nationalized banks; open access to foreign borrowing
on August 15. A US$925 million bridge loan by the by private banks and businesses. Chile, however,
Annex 61
did not adequately adapt and update its supervisory confidence. Despite official reaffirmations of the peg,
capabilities and prudential regulations over the authorities could not honor this commitment; on June
financial sector; “supervision continued to focus on 13, 1982, the Minister of Economy, Luis Francisco
reviewing compliance with accounting rules and Danus, announced a devaluation in the Peso of about
related regulations, but did not concentrate on the 18 percent vis-à-vis the dollar. Danus also announced
overall risks affecting the operations of each bank” that the Peso would continue to be depreciated at
(Larrain 1989, page 10). a fixed rate vis-à-vis a basket of the currencies. The
Starting in 1978, these changes resulted in a crawling peg regime subsequently collapsed on
substantial accumulation of foreign currency debt August 5, 1982 causing a run on bank deposits and
by the private sector. Buoyed by the high economic forcing the authorities to float the Peso.
growth, on June 29, 1979, the crawling peg regime The depreciation of the Peso severely ham-
was replaced with a firm peg. The move to a firm peg pered firms’ ability to repay or service their foreign
aimed at establishing a nominal anchor to counter currency debts to the banking sector. Non-performing
stubbornly high inflation rates. Nonetheless, inflation loans soared from 2.3 percent in 1981 to 18.4 percent
and inflation expectations continued to be elevated in 1983. Domestic banks borrowed more in foreign
and the firm peg resulted in a further loss of com- currencies to honor their obligations to foreign banks.
petitiveness and a widening current account deficit. On January 10, 1983, Chile secured a 24-month
Indeed, the current account deficit stood at 18.9 SBA from the IMF in the amount of US$550 million,
percent of GDP in 1981 (Larrain 1989). and drew an additional US$325 million CFF. This,
Toward end-1981, domestic and external factors however, was insufficient to reverse the spiral. On
helped plunge the economy into a recession and made January 14, eight financial institutions (seven banks
the debt burden more formidable. Difficulties included and one finance company) became insolvent and
a decline in copper prices, Chile’s main export. Further, required intervention, 3 of which were subsequently
an increase in global interest rates stalled the domestic liquidated. A bank holiday was declared. At that stage,
credit boom and led to soaring domestic real interest the IMF program was off track.
rates in a bid to stanch capital flight. By year-end, the On March 22, 1983, the new finance minister
current account deficit stood at 15 percent of GDP and announced an emergency economic program, which
the annual lending rate reached 63 percent (Boughton consisted of accelerating the Peso’s devaluation, fiscal
2001). This made the servicing of bank loans more measures and the tightening of foreign exchange con-
challenging for debtors and forced the central bank to trols. In addition, a US$1.3 billion funding and public
supply liquidity to the banking sector. debt rescheduling package, supplanted with a bridge
A systemic banking crisis was in full swing by loan from the Bank of International Settlements, was
November 1981 (Laeven and Valencia 2014); eleven agreed upon with the banks. In light of the agreement
financial institutions were intervened between 1981 with the banks, the IMF disbursed based on a new
and 1982 and eventually liquidated (Larrain 1989). shadow program on July 27.
These institutions accounted for about 14.5 percent Efficient and sensible restructuring of the
of deposits in the banking sector. The two largest banking sector and the credit portfolio were critical
banks were put into receivership and their assets for an accelerated recovery. This included bank
were transferred to other institutions. The govern- re-capitalization and incentives for recouping bad
ment and shareholders absorbed the losses of the loans. Further, debt relief was offered to borrowers
failed institutions and depositors were compensated. who were assessed to be “productive.” These inter-
The pressures on a weakened banking sector ventions were expensive; the cost of foreclosure of
were exacerbated by adverse macroeconomic insolvent institutions was 10.6 percent of the GDP
developments in 1982 and 1983. Rising fears of and the cost of portfolio purchase under conditions
an impending devaluation were met with a cabinet of repurchase reached 6.7 percent of the GDP
shuffle on April 11, 1982, aiming at shoring up (Sanhueza 2001).
62 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
The Venezuelan Banking Crisis of 1994 adopted a crawling peg. The rising interest rates
precipitated a recession, reduced bank lending and
In the period preceding Venezuela’s financial crisis, led to an increase in non-performing loans from 4 per-
the country’s banking system was oligopolistic cent in 1991 to 10 percent in 1993 (Garcia-Herrero
in nature, composed of a large number of 1997). Higher interest rates did not curtail the outflow
specialized banks belonging to financial groups of deposits to the offshore banking sector. Indeed,
that were ultimately owned by a few individuals. This deposits decreased in real terms by 11 percent in
concentration encouraged lax supervision, resulting 1993 relative to 1991 (Garcia-Herrero 1997) and
in low capitalization and incentives to divert losses banks siphoned their liquidity to their offshore facili-
and problem loans to offshore branches, which were ties. Despite liquidity injections by the CBV, rumors
subject to even less supervision. This created large off- concerning the health of the banking sector continued
balance sheet items for Venezuela’s financial sector, unabated at end-1993.
which was not monitored by authorities. Moreover, The crisis broke with the collapse of Banco
the share of state banks was relatively low compared Latino in mid-January of 1994, the second largest
to other Latin American economies; as of June 1993, bank in terms of deposits, prompting a run on two
private banks held about 90 percent of total assets, other banks that belonged to the same financial group.
with the largest six banks holding 52 percent of the This affected over 10 percent of commercial banks’
total (Garcia-Herrero 1997). deposits. Hence, panic soon spread to other banks.
The state of the macroeconomy in the 1980s The Deposit Guarantee Fund reacted by assisting
reflected anemic growth and rising inflation. This the distressed banks and the CBV continued to inject
resulted in negative real interest rates causing liquidity into the banking sector. The resources of the
disintermediation and capital outflows; in the late Deposit Guarantee Fund were depleted by February
part of the decade, large amounts of deposits were 1994 and the foreign currency reserves of the CBV
transferred to banks’ off-shore facilities. In 1989, the diminished considerably amid an unrelenting run on
Venezuelan government, in coordination with the IMF, deposits and capital flight.
launched a macroeconomic adjustment program in a The authorities responded by passing a law
bid to stimulate the economy. The program consisted on March 1994 to protect depositors, nationalizing
of abandoning interest rate controls and shifting to Banco Latino and reopening it with withdrawals
indirect monetary policy tools, as well as the unifica- limits on deposits that were frozen, even those in
tion and floating of the Bolivar. off-balance sheet accounts. The Deposit Guarantee
Nonetheless, the resurgence of inflationary Fund injected the equivalent of 3.6 percent of GDP
pressures led the Venezuelan authorities to rein- to cover the losses of Banco Latino (Garcia-Herrero
state the peg in 1990. Two coup attempts in 1991 1997).
and political instability throughout 1992 aggravated These actions proved insufficient to shore
capital flight and made the economy more vulner- up the public’s feeble confidence in the banking
able. An exchange rate crisis occurred in October sector. Deposit runs continued in light of fears of
1992 following the loss of about US$1 billion in partial deposit freezes, devaluation and exchange
reserves (Garcia-Herrero 1997). The Venezuelan rate controls. When the CBV ceased supplying the
economy depended heavily on oil exports and the Deposit Guarantee Fund with liquidity in June 1994,
weakening demand for oil in 1993, political uncer- eight distressed financial institutions, accounting for
tainty and loose fiscal policy weighed negatively on 21 percent of deposits, had to be intervened despite a
the economic outlook and exerted pressure on the massive liquidity injection of 6 percent of GDP by that
Bolivar. date (Garcia-Herrero 1997). In an attempt to restore
Amid these adverse conditions and due to confidence in the banking sector, the authorities
a significant loss in reserves, the Central Bank of established a Financial Emergency Board. However,
Venezuela (CBV) tightened monetary policy and the financial distress of the Deposit Guarantee Fund
Annex 63
continued to weigh heavily on confidence in the liquidity held by banks and international reserves. As
public sector. the system lacked a lender of last resort in dollars, the
In the following weeks, capital flight led to financial system was inherently unstable, subject to
a large loss of reserves at the CBV. The authorities bank runs (Kiguel 2016).
responded by fixing the exchange rate, and insti- In the period just prior to the abandonment
tuting price and exchange rate controls. Rumors of the currency board (1998–2001), the economy
surrounding two large banks—Banco Consolidado witnessed a deep contraction, exposing mounting
and Banco Venezuela—renewed the deposit run and vulnerabilities in the economy. The hard peg and
the authorities decided to nationalize both banks. In a lack of fiscal space precluded countercyclical
December, two additional banks—Banco Progreso macroeconomic measures. In response, in January
and Republica—were in distress. Authorities closed 2001, the IMF approved an augmentation of
the former and nationalized the latter. Further, financing, boosting an existing SBA program, to
the liabilities of Banco Progreso were migrated to an equivalent of US$14 billion, centered on fiscal
the nationalized banks (Banco Latino, Venezuela, adjustment and accelerated structural reforms
Consolidado and Republica) without a corresponding (IMF 2003). However, this failed to achieve stability.
transfer of assets (due to large one-sided off-balance So did various attempts by the Government for
sheet items in the form of deposits that suddenly voluntary debt arrangements. The IMF approved a
appeared), thus placing enormous pressure on the new program, disbursing US$5 billion immediately
four banks which continued to experience deposit and pledging another US$3 billion in support of
withdrawals. prospective debt restructuring.
The passage of a Financial Emergency Law The crisis broke with a run on private deposits,
in July 1995, which gave broader powers to the which fell by more than US$3.6 billion (6 percent of the
Financial Emergency Board and eased restrictions deposit base) over November 28–30, 2001. The authori-
on the liquidation of impaired assets at the Deposit ties responded with a wide range of controls on banking
Guarantee Fund, succeeded in mitigating, but not and foreign exchange transactions. As the economy
completely resolving, the banking crisis. Deposit flight faltered, social and political unrest ensued, forcing the
continued, albeit at a slower rate, and a small bank resignation of President de La Rua on December 20,
was intervened in August 1995. followed by 4 other (Congress-appointed) Presidents
within 3 weeks. On December 23, President Sáa
The Argentinian Financial Crisis of 2001–02 declared a default on Government debt; on January
3, 2002, President Duhalde announced the end of the
The Argentine Currency Board—via the so-called convertibility regime (IMF 2003).
Convertibility Law—pegged the Argentine peso to the
U.S. dollar at a parity from March 1991, until its failure The Uruguayan Banking Crisis of 2002
in January 2002 (Spiegel 2002). Argentina adopted
the hard peg in an attempt to eliminate hyperinflation On the eve of the crisis, Uruguay’s banking sector,
and stimulate economic growth following a which consisted of two large public banks—Banco
tumultuous economic performance in the 1980s. de la Republica Oriental del Uruguay and Banco
While successful initially, it became a main constraint Hipotecario del Uruguay—and approximately thirty
on the economy’s competitiveness in general, and private banks, was widely regarded as sound and
on countercyclical macroeconomic management healthy (de la Plaza and Sirtaine 2005). The banking
in particular. Moreover, there was a large degree sector, however, was highly dollarized on the asset
of financial dollarization in the economy with the and liability sides and, although not as highly exposed
banking system functioning mainly in US dollars. The to the sovereign as its neighbors, it was prone to
banking system’s dollar-denominated, short-term external shocks and to cross-border bank runs due
liabilities exceeded its stock of dollar assets—namely to its exposure to Argentina. On the liabilities side,
64 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
foreign currency deposits constituted 90 percent of May of 2002. The deposit withdrawals, which were
total deposits in the system, with the latter amounting no longer confined to specific banks, accelerated
to US$ 15.4 billion (representing 83 percent of in May and June of 2002. On June 21, 2002, Banco
Uruguay’s 2001 GDP). Further, 47 percent of foreign de Monteviedo-Caja Obrera, the third largest private
currency deposits were held by non-residents. The bank experienced severe liquidity shortages causing
dollarization of assets reached 75 percent of the total the authorities to intervene. As sentiment deteriorated
loan book of the system, with the latter amounting to further in July 2002 and government spreads widened,
US$ 11.5 billion. The banking sector also suffered the bank run extended to local currency deposits. By
from currency mismatches as 71 percent of foreign the end of July 2002, a cumulative 38 percent of total
currency loans were extended to residents, the vast deposits had been withdrawn from the system. The
majority of who earned in Pesos (de la Plaza and majority of banks had become technically insolvent
Sirtaine 2005, pg. 4–5). (de la Plaza and Sirtaine 2005, pg. 11).
A weakening macroeconomic environment Faced with declining foreign currency reserves
weighed on banks’ profitability. Uruguay had due to liquidity support to the banking sector, the
experienced a prolonged recession since 1999 authorities were forced to abandon the crawling peg.
and recurrent fiscal deficits were financed by As a result, the Peso immediately depreciated by 27
issuing mostly foreign currency denominated debt. percent forcing the authorities to declare a five-day
Government debt increased from 38 percent of banking holiday on July 30, 2002; by then, the Peso
GDP 1998 to 58 percent of GDP in 2001. The fixed had lost 57 percent of its value.
exchange rate regime (Roubini and Sester 2004), The bank run had turned into a system-wide
which weakened external competitiveness, became credit crunch; credit to the non-financial sector
untenable with devaluations in the currencies of shrunk by 37 percent in 2002, greatly contributing to
Uruguay’s two largest neighbors and trading partners, a GDP contraction of 10.7 percent for that same year
Brazil and Argentina, in 1999 and 2001, respectively. (de la Plaza and Sirtaine 2005, pg. 11). The two public
In the lead up to the banking crisis, the Uruguayan banks were in a perilous position. The Banco Galicia
economy was characterized by a high level of foreign Uruguay had been suspended and the government
currency indebtedness and a significantly overvalued took control of the Banco Commercial and Banco de
exchange rate vis-à-vis its major trading partners and Montevideo-Caja Obrera.
the rest of the world. In the first stage, the authorities’ response to
The crisis broke in December 2001 when the crisis consisted of: provision of ample liquidity
Argentina imposed deposit freezes (“el corralito”). support via already established lender of last resort
Two of Uruguay’s largest private banks—Banco Galicia facilities; the restructuring and/or liquidation of
Uruguay (a subsidiary of an Argentinian bank) and troubled institutions; and the expansion of the
Banco Commercial—which combined accounted for crawling exchange rate band from 6 to 12 percent.
around 20 percent of the deposit base, came under While this response appeared to stem cross-border
intense pressure due to their exposure to Argentinian bank run, the deepening of the crisis in June and July
depositors and debt. Deposit withdrawals continued 2002 compelled the authorities to scale up and better
unabated as the crisis in Argentina deepened. target their response.
By March 2002, deposit withdrawals, mostly The central bank of Uruguay subsequently
by non-residents (i.e., Argentinians), amounted to 12 prioritized liquidity support to core banks, which
percent of the deposit base (de la Plaza and Sirtaine played an instrumental role in the payment system;
2005). Argentina’s tightening of its deposit freeze in non-core banks, which were mostly Uruguayan
February 2002 (“El Corralon”) as well as Uruguay’s branches of foreign banks, were to rely on liquidity
downgrade from investment grade prompted a support from their headquarters abroad. In light of the
second wave of withdrawals by residents and non- central bank’s diminishing scope to act as a lender of
residents totaling 18 percent of deposits in April and last resort, the authorities established in June 2002 a
Annex 65
US$2.5 billion facility, the Fondo para la Fortificacion crisis, Greece was gripped with deep twin structural
del Sistema Bancario (FFSB), to provide equity and deficits, lack of competitiveness as wage growth
liquidity support to the core banks. The FFSB was outpaced productivity growth and a real exchange
funded by an augmentation of the SBA with the IMF, rate overvaluation (IMF 2012a).
other multinational institutions and the government. The onset of the global financial crisis exacer-
The FFSB ultimately proved to be insufficient bated the mounting imbalances of the Greek economy.
and was suspended. Following the five-day bank After the failure of Lehman brothers in September
holiday that was declared on July 30, the authorities 2008, the spread between the Greek Government
announced on August 5, 2002 the creation of US$1.4 bonds and the German bunds soared to 100 basis
billion stabilization fund, the Fund for the Stability of the points (IMF 2013) and led to downgrades by Standard
Banking System (FSBS), that was funded by the IMF, and Poor’s. A main trigger to the Greek episode was
the World Bank and the Inter-American Development data revision by the authorities in October 2009, which
Bank (IADB). The FSBS was sufficient to fully back the entailed a sizeable increase in the projected fiscal
remaining US$ deposits at core banks (Seeling 2007; deficit from 4 to 12.5 percent of GDP.92 This roiled
de la Plaza and Sirtaine 2005). The establishment of markets further, weakened confidence in the Greek
the FSBS, coupled with maturity extensions of dollar economy and prompted a downgrade by Fitch (IMF
deposits in the public banks and changes to macropru- 2013). The loss of confidence in the Greek economy
dential regulations, succeeded in stopping the bank prompted capital outflows from the banking sector
run. The IMF’s exposure (of US$1.3 billion) to Uruguay, estimated at 30 percent of the deposit base. It also
relative to GDP, was its largest to date (Seeling 2007). suspended Greece’s access to financial markets by
significantly widening yields on Greek bonds to unaf-
The Greek Financial Crisis 2009 fordable rates.
Given that Greece is an EMU member, a
Greece’s accession to the European Monetary Union nominal currency devaluation that corrects the built-
(EMU) in 2001 did not correct structural imbalances up imbalances was not possible. Instead, under the
that included large internal and external deficits Troika of the IMF, the European Commission and the
coupled with a low growth environment. Public sector European Central Bank (ECB), Greece underwent a
net borrowing averaged around 7 percent of GDP very sharp internal devaluation, including a reduction
annually in the Euro-but-pre-crisis period (2002– in the wage bill and pension benefits. A deep reces-
2008), compared to 6.7 percent in the pre-Euro sion ensued over the next decade.
period (1994–2000). Public debt remained relatively In return, the Greek authorities secured a €30
stable over the former period fluctuating between billion SBA from the IMF in May 2010, complemented
100 and 110 percent of GDP. Externally, the current with a cooperative package of financing from the
account deficit rose from a pre-Euro average of 3.3 European Union amounting to €110 billion (IMF,
percent of GDP to Euro-pre-crisis average of 10.5 2012b). The SBA was underpinned by a stringent
percent. Moreover, external debt also rose decidedly fiscal consolidation program that aimed at putting
from about 100 percent of GDP end-2003, to 133 Greek debt on sustainable footing. However, key
percent by end-2008. Clearly, the EMU facilitated SBA macro-fiscal targets, namely, fiscal and current
easier access to foreign financing for both the public account indicators, failed to be met. The SBA was
as well as the private sectors; by 2009, private sector subsequently cancelled in 2012 and was replaced
external debt stood at 175 percent of GDP (IMF with an Extended Fund Facility (EFF) arrangement
2017a). The adoption of the Euro also adversely
affected the Greek economy’s competitiveness 92
The data revision came amidst concerns raised
relative to its trading partners, given that wages in by Eurostat—the statistical office of the European
the European periphery countries rose relative to the Commission—regarding the quality of Greece’s fiscal
core countries (Hale 2013). Hence, on the eve of the data on five occasions over the period 2005–2009.
66 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
on March 15, 2012, also under the Troika. The EFF 1982 Mexican Debt Crisis for Greece | VOX,
arrangement included financing of about €173 billion CEPR Policy Portal (voxeu.org).
over four years. de la Plaza, L. and Sirtaine, S. 2005. An Analysis of the
Critically, and as a prior action for the EFF, a 2002 Uruguayan Banking Crisis, World Bank
Private Sector Initiative (PSI) was announced on June Policy Research Working Paper 3780, World
2011 for a voluntary debt swap of Greek sovereign Bank, Washington DC.
bonds involving a haircut on private creditors who Dornbusch, R., Werner, A. 1994. Mexico: stabilization,
were represented by the Institute of International reform, and no growth. Brookings Papers on
Finance. This haircut, which was subsequently imple- Economic Activity, 253–315.
mented on March 2012, was equivalent to a 53.5 Garcia-Herrero, A. 1997. Banking crises in Latin
percent cut in the face value (principal) of the bonds, America in the 1990s: Lessons from Argentina,
corresponding to an approximately €107 billion Paraguay, and Venezuela. IMF working paper
reduction in Greece’s debt stock.93 WP/97/140, International Monetary Fund,
The internal adjustment proved harsh and Washington DC.
counter-productive, as macro-fiscal targets remained Hale, G. 2013. Balance of Payments in the European
elusive due to unaccounted for economic contrac- Periphery, Federal Reserve Bank of San
tions and an unsustainable public debt that persisted Francisco Economic Letter.
despite the PSI. This translated into social pain and IMF 2003. Lessons from Argentina, Washington DC,
political instability. A banking sector crisis ensued October 2003.
in 2015, which required the introduction of capital IMF 2012a. Greece: Request for Extended
controls. Once again, the EFF program faltered Arrangement Under the Extended Fund
and was eventually cancelled in January 2016 (IMF Facility, Washington DC.
2017a). The prolonged economic contraction helped IMF 2012b. Press Release: IMF Executive Board
drive the debt-to-GDP ratio to a peak of 180 percent Approves €30 Billion Stand-By Arrangement
of GDP in 2016. In July 2017, the IMF approved a for Greece.
precautionary €1.6 billion SBA for Greece (IMF IMF 2013. Greece: Ex-Post Evaluation of Exceptional
2017b). This time the program explicitly noted that, Access Under the 2010 Stand-By Arrangement,
without debt relief, Greece’s debt would continue to Washington DC.
be unsustainable. IMF 2017a. Greece: Ex-Post Evaluation of Exceptional
Access Under the 2012 Extended Arrangement,
References Washington DC.
IMF 2017b. Greece: Request for Stand-By
Balino, T. J., T. 1991. The Argentine Banking Crisis Arrangement, Washington DC.
of 1980. In Banking Crises: Cases and Issues, Kiguel, Miguel 2016. Argentina’s 2001 Economic
edited by V. Sundararajan and Tomâs J. T. Balino, and Financial Crisis: Lessons for Europe,
International Monetary Fund, Washington DC. Brookings, June 2016.
Boughton, J. 2001. The Crisis Erupts. In Silent Laeven, L., and Valencia, F. 2014. Systemic
Revolution: The International Monetary Fund, Banking Crises. In Financial Crises: Causes,
International Monetary Fund, Washington DC. Consequences and Policy Responses, edited
Boughton, J. 2001. The Mexican Crisis: No by Stijn Claessens, M. Ayhan Kose, Luc Laeven
Mountain Too High?. In Silent Revolution: The
International Monetary Fund, International 93
The European Stability Mechanism: https://www.
Monetary Fund, Washington DC.
esm.europa.eu/content/what-was-private-sector-debt-
Cantu, C., Park, K .Y., and Tornell, A. 2015. Lessons restructuring-march-2012#:~:text=Also%20known%20
from the 1982 Mexican Debt Crisis for Greece. as%20the%20PSI,lighten%20Greece’s%20overall%20
VoxEU, CEPR, retrieved from Lessons from the debt%20burden.
Annex 67
and Fabian Valencia, International Monetary Sanhueza, G. I. 2001. Chilean Banking Crisis of the
Fund, Washington DC. 1980s: Solutions and Estimations of the Costs.
Larrain, M. 1989. How the 1981–1983 Chilean Central Bank of Chile Working Paper No 104.
Banking Crisis was Handled, Working Paper Seeling, S. 2007. Resolving the Banking Crisis in
WPS 300, World Bank, Washington DC. Uruguay. In Building Monetary and Financial
Nascimento, J. C. 1991. Crisis in the Financial Sector Systems: Case Studies in Technical Assistance
and the Authorities’ Reaction: The Philippines. edited by Charles Enoch, Karl Habermeier and
In Banking Crises: Cases and Issues, edited Marta Castello-Branco, International Monetary
by V. Sundararajan and Tomâs J. T. Balino, Fund, Washington DC.
International Monetary Fund, Washington DC. Spiegel, Mark 2002. Argentina’s currency crisis:
Oks, D., and S van Wijnbergen 1994. Mexico after the lessons for Asia, Federal Reserve Bank of San
debt crisis: is growth sustainable?, World Bank Francisco, December 2002.
Policy Research Working Paper, WPS 1378. Van Wijnbergen, S. 1991. Mexico and the Brady plan
Roubini, N., & Setser, B. 2004. Bailouts or bail-ins?: Economic Policy, 6 (12), 14–56.
responding to financial crises in emerging
economies. Peterson Institute.
68 LEBANON ECONOMIC MONITOR: LE NAUFRAGE DU LIBAN (AU NUMÉRO 3)/LEBANON SINKING (TO THE TOP 3)
1818 H Street, NW
Washington, DC 20433