twpsp a4nq POLICY RESEARCH WORKING PAPER 2749 Loan Loss Provisioning and Most banks around the world delay provisioning for bad Economic Slowdowns loans until it is too late- when cyclical downturns Too Much, Too Late? have already set in. The size and timing of loan loss provisions tend to improve Luc Laeven with the level of economic Giovanni Ma/noni development. The World Bank Financial Sector Strategy and Policy Department U December 2001 I POLICY RESEARCH WORKING PAPER 2749 Summary findings Only recently has the debate on bank capital regulation provisioning practices around the world. They find that devoted specific attention to the role that bank loan loss in the vast majority of cases banks tend to delay provisions can play as part of a minimum capital provisioning for bad loans until it is too late-When regulatory framework. Several national regulators have cyclical downturns have already set in-possibly adopted or are planning to introdure a cyclically magnifying the impact of the economic cycle on the adjustable requirement for loan loss provisions, and the income and capital of banks. Basel Committee on Banking Supervision is considering Notwithstanding the considerable variation in the how to provide adequate treatment to provisioning patterns followed by banks around the world, Laeven practices within a broad bank capital regulatory and Majnoni find that the size and timing of provisions framework. tend to improve with the level of economic development. Laeven and Majnoni contribute to the ongoing debate by exploring the available evidence about bank This paper-a product of the Financial Sector Strategy and Policy Department-is part of a larger effort in the department to study the impact of financial regulation on economic development. Copies of the paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Rose Vo, room MC9-624, telephone 202-473-3722, fax 202-522-2031, email address hvol @worldbank.org. Policy Research Working Papers are also posted on the Web at http:/ /econ.worldbank.org. The authors may be contacted at llaeven@worldbank.org or gmajnoni@worldbank.org. December 2001. (27 pages) The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the view of the World Bank, its Executive Directors, or the countries they represent. Produced by the Policy Research Dissemination Center Loan Loss Provisioning and Economic Slowdowns: Too Much, Too Late? Luc Laeven and Giovanni Majnoni** JEL classification numbers: G21, G28. Keywords: banks, bank regulation, loan loss provisions Luc Laeven: email: llaeven(0i,worldbank.org. Giovanni Majnoni: email: gmainoni(&-).worldbank.org; The World Bank, 1818 H Street, NW, Washington, DC, 20433. Tel: (202) 458 7542; Fax: (202) 522 2031. We are grateful to Franklin Allen, Jerry Caprio, Patrick Honohan, Rick Mishkin, Larry Promisel and Anthony Santomero for useful comments on earlier drafts of this paper. 1. Introduction. Risk-based bank minimum capital requirements tend to have a pro-cyclical effect on the economy (Basel Committee on Banking Supervision, 2000). The deterioration of the quality of bank loan portfolios during economic downturns inevitably increases banks' risk exposure - and therefore the level of capital requirements - exactly when capital becomes more expensive or simply unavailable to weaker institutions. The discussion on this topic has raged ever since the 1988 Capital Accord was originally enforced in GIO economies and subsequently, following the introduction of Basle-like approaches by most developed and emerging countries around the world. On one side it has become widely perceived that risk exposures need to be explicitly mirrored in the level of bank capital if regulatory arbitrage is to be avoided and bank stability pursued. On the other side, potential negative externalities of capital regulation have been stressed, pointing to the contraction of credit supply that higher capital requirements may generate during economic downturns. In general, critics of the solvency ratios discipline warn that controlling individual risk positions may not always minimize systemic risks and strict capital standards may, for instance, have aggregate undesirable liquidity effects. The discussion has become more animated in the last couple of years as a consequence of the ongoing revision of the old Basle Capital Accord. This paper contributes to the ongoing debate by focusing on a frequently ignored aspect of bank capital regulation: the role of bank loan loss reserves as a component of bank regulatory capital. The question addressed is twofold. First, are there good reasons - conceptual and empirical - for a specific regulation of loan loss reserves within the general regulation of 1 solvency ratios? Second, is it likely that a distinct treatment of loan loss reserves may affect the pro-cyclical features of capital regulation? Following what appears to be the consensus view among practitioners and analysts of risk management we relate the volume of bank capital to the size of unexpected credit losses and loan loss reserves to the size of expected losses. We also argue that, consistently with this view, loan loss reserves should be left free to fluctuate over the cycle and comply with a minimum requirement to be respected on average over a predefined period and not at every single moment in time. A constant minimum requirement would therefore apply to economic capital and an average minimum requirement would instead apply to loan loss reserves. This approach would clearly strike a balance between the supporters of the opposing views of a fixed versus adjustable solvency regulation, but its relevance can hardly be defined at a theoretical level. Only an empirical verification can show whether bank managers are already pursuing a desirable pro-cyclical cyclical management of loan loss provisions and reserves, making additional regulatory incentives useless. This paper therefore proceeds to analyze the cyclical patterns of bank loan loss provisions followed by large commercial banks in different geographical areas of the world. We anticipate some of the relevant results, noting that clearly different patterns prevail according to the location of the banks. Bankers on average create too little provisions in good times and are then forced to increase them during cyclical downturns magnifying losses and the size of negative capital shocks. These patterns are considerably diversified within the group of industrialized countries as well as within emerging 2 economies. Larger and more timely provisions, though, appear to be positively affected by the level of economic development. The paper is structured as follows. Section 2 draws from the current debate on the cyclical impact of banks' capital requirements. Section 3 discuss the role of bank loan loss provision in the current debate of banks' minimum solvency ratios. Section 4 describes the empirical analysis and the data. Section 5 reports the empirical results, and Section 6 concludes. 2. Bank capital requirements and the economic cycle. The cyclical effects of bank capital regulation have been thoroughly analyzed by a wide theoretical and empirical literature that has flourished in the 1990s following the introduction of the 1988 Capital Accord. The concern raised by academic and policymakers in the wake of the new regulation was that new higher capital ratios could lead to a reduced credit supply in periods of economic slowdown. Concerns were twofold. On one side there was the preoccupation that the shift to a new regulatory regime could impact negatively on the supply of credit with a once for all effect. A second and more generalized concern was that a risk-based capital regulation by increasing capital requirements might increase the likelihood of capital shortages during recessions potentially reducing the supply of credit to the economy. The expression "capital crunch" was coined in the early nineties to characterize the 3 simultaneous shortage of capital and the contraction in the supply of new loans that affected banks in New England during the early 1990s recession in the United States A capital crunch could result in the reduction of total bank assets or alternatively in a shift toward less risky assets such as government bonds. An extensive survey of the empirical evidence available for industrialized economies, has concluded that "there is some evidence that bank capital pressures during cyclical downturns in the US and in Japan may have limited lending in those periods and contributed to economic weakness in some macroeconomic sector" (Basel Committee on Bank Supervision, 1999). Recent empirical evidence shows that the introduction of more severe capital regulation may have reduced bank credit supply also across emerging economies (Chiuri et al., 2002). These concerns have recently been addressed by policy makers as well. The Financial Stability Forum, for instance, has raised the question whether several features of the new capital regulation currently discussed by the Basel Committee on Banking Supervision could increase the cyclical fluctuations of the economy. In response, the Basel Committee has confirmed that risk-based capital requirements are inevitably pro- cyclical (more capital is required during recessions exactly because credit risks in banks' portfolios increase in cyclical downturns) and suggested that the cyclicality question should be addressed by means of different instruments. For examnple, national supervisors (under Pillar II of the new accord) could request banks to comply with higher than minimum capital requirements and leave bank capital free to fluctuate above that level. See Bernanke and Lown (1991) and Peek and Rosengren (1995) for evidence in favor of the presence of a capital crunch during the 1990-91 recession in the US. A contrary view is taken by Berger and Udell (1994). 4 At a theoretical level, an explicit treatment of the impact of capital requirements on the level of economic activity is provided by Holmstrom and Tirole (1997) in a model that provides a rationale for applying lower solvency ratios in recessions. They find that, in a world where agents both in the real and in the financial sector may be capital constrained, market-determined solvency ratios are pro-cyclical, i.e., they are higher during expansions and lower during recessions. More precisely, they show that a negative shock to banks' capital negatively affects the level of economic activity and that the lower level of investment generated by the capital crunch requires a reduction of market determined solvency ratios. Tirole and Dewatripont (1994) also remark that the lack of discrimination between idiosyncratic and macroeconomic shocks may have undesirable effects negatively affecting bank managers risk taking incentives. Bank managers would in fact be punished both for idiosyncratic shocks, that are under their control, and for macroeconomic shocks, that are independent from their control. They conclude that Basle standards are "excessively tough on bank managers in recessions". How can concerns about the cyclical effects of a risk based capital regulation be reconciled with the Basel Committee assessment that risk based capital requirements are a necessary ingredient of financial stability? This paper suggests that a compromise between these opposing position may in fact exist. The suggested reconciliation is based on the recognition that bank capital and bank loan loss reserves perform different functions and that therefore their regulatory requirements could differ. For example, while capital may be re, dlated by a fixed minimum requirement, loan loss reserves may 5 be required to meet a minimum requirement on average over a predefined period, allowing them to fluctuate over the cycle. 3. Loan loss reserves and banks minimum capital requirements. Current minimum solvency regulations commonly refer to a particular notion of capital called "regulatory capital" which differs from "economic capital" and that results from the sum of Tier 1 and Tier 2 capital (Berger et al., 1995). The bulk of Tier 1 capital is represented by paid-in capital and retained earnings, while Tier 2 capital includes general loan loss reserves and a variety of bank liabilities characterized by a lower degree of seniority with respect to other non-capital bank liabilities. The sum of Tier 1 and Tier 2 capital represents the numerator of the solvency ratio and needs to meet minimum regulatory requirements. We suggest that a reconciliation of the different views about banks capital requirements could be envisioned by considering a partition of regulatory capital based not only on seniority considerations - as is the case for Tier 1 and Tier 2 capital - but also and foremost on risk management considerations. Following the general consensus among risk management analysts and practitioners, economic capital should be tailored to cope with unexpected losses, and loan loss reserves should instead buffer the expected component of the loss distribution. Coherently, loan loss provisions required to build up loan loss reserves should be considered as a cost - that may be delayed in time but eventually will realize - differently from earnings which affect the stock of capital. A more detailed description of the conceptual difference between loan loss reserves and provisions and capital and earnings is provided in Appendix 1. 6 We can show that a loan provision management coherent with an increase of loan loss reserves in good time and a decrease in bad times reduces bank profit volatility and the probability of a negative shock to economic capital. For simplicity, consider a bank with only loans as assets. Let L be the amount of bank loans, rL the lending rate and r the average cost of funding. In this case, net interest income equals L * (rL - r). Let bank profits (7r) be expressed by the difference between net interest income, operating costs (OC) and the amount of asset depreciation (AD) = L -(rL - r ) - OC -AD (1) Let the lending rate (rL) be defined as the sum of the risk-free interest rate (rf), the pro- rated (unconditional) expected loss ratio E(d), the level of unit operating costs (c), and the level of risk premium (k): rL =rf +E(d)+k+c (2) By substituting equation (2) in (1) and abstracting from operating costs (we assume that c*L=OC), we have that profits may be represented by the following equation: or= L-. (rf +k) -r + (L *E(d ) -AD) (3) It is clear that, if asset depreciation is kept equal to the value of unconditional expected losses (AD = L *E(d)), the volatility of bank profits is not affected by the fluctuations of credit losses over the cycle. To keep asset depreciation (AD) constant is sufficient for loan loss provisions (LLP) to compensate the difference between realized credit losses and average credit losses by taking positive values (LLP>O) during cyclical expansions 7 and negative values (LLP,r) and downturns would possibly generate losses (i