77779 Reversal and Reduction, Resolution and Reform Lessons from the Financial Crisis in Europe and Central Asia to Improve Outcomes from Mandatory Private Pensions Private and Financial Sector Development Department Europe and Central Asia Region Vice Presidency Non-Bank Financial Institutions Unit; Capital Markets Department Financial and Private Sector Development Vice Presidency The World Bank May 2013         Acknowledgments     This   report   was   written   by   Will   Price   (Senior   Financial   Sector   Specialist),   and   Heinz   P.   Rudolph   (Lead   Financial   Economist),   both   at   the   Global   Capital   Markets   and   Non-­�Bank   Financial   Institutions   Group,   Financial   and   Private   Sector   Development   Vice   Presidency   at   the   World   Bank.     The   ECA   Region   task   manager   for   the   report   was   John   Pollner   (Lead   Financial   Officer).   This   report   is   a   joint   publication   of   the   ECSPF   and   FCMNB   groups.   The   findings,   interpretations,   and   conclusions   expressed   in   this   report   are   entirely   those   of   the   authors,   and   they   do   not   necessarily   represent   the   views   of   the   World   Bank   Group.   The   authors   are   grateful   for   the   comments   received   from   John   Pollner,   Sebastian   Eckhard,   Asta   Zvinienne,   Tomas   Prouza,   Zeljko   Bogetic,   Robert   Palacios,   and   the   participants   at   the   January   2013   workshop   at   the   World   Bank,   where   a   preliminary   version   of   this   report   was   presented.  The  authors  would  also  like  to  thank  Otari  Dzidzikashvili  for  valuable  research  assistance.     Contents   2   Contents  ...................................................................................................................   1   Abbreviations  ...........................................................................................................   2   Executive  Summary  ..................................................................................................   4   Chapter  1.  Pre-­�crisis  Pensions  and  the  Impact  of  the  Crisis  .....................................   1.1  Introduction  .......................................................................................................................................  4   1.2  The  original  pension  reforms  .............................................................................................................  4   Table  1.1  Structure  of  Pension  Systems  in  Central  and  Eastern  Europe  ........................................  5   1.3  The  evolution  of  pension  systems  after  the  initial  reforms  ...............................................................  5   Figure  1.1  Increases  in  Pension  Generosity  in  Poland  after  the  Initial  Reform  ..............................  6   Figure  1.2  Increases  in  Pension  Generosity  in  Hungary  after  the  Initial  Reform  ............................  7   1.4  The  financial  crisis  ..............................................................................................................................  7   Figure  1.3  Real  GDP  Growth  in  ECA  Countries,  2002–12  ................................................................  8   Table  1.2  Real  GDP  Growth  in  ECA  Countries,  2007–12  .................................................................  9   1.4.1  Government  finances  ............................................................................................................  9   Figure  1.4  Fiscal  Deficits  as  a  Share  of  GDP  in  ECA  Countries,  2000–12  .......................................  10   Figure  1.5  Sovereign  Debt  as  a  Share  of  GDP  in  ECA  Countries,  2000–12  ....................................  11   Figure  1.6  Ratio  of  Government  Debt  to  GDP  in  ECA  Countries  and  Select  EU-­�15  Countries,  2008   ….  ..................................................................................................................................................  11   1.4.2  Financial  market  conditions  .................................................................................................  12   Table  1.3  Sovereign  Spreads  in  ECA  Countries,  2007–11  .............................................................  12   Table  1.4  Long-­�Term  Sovereign  Bond  Ratings  in  ECA  Countries,  2007–12  ...................................  13   Figure  1.7  Nominal  Returns  of  Mandatory  Pension  Funds  Relative  to  MSCI  Europe,  2004–12  ...  14   Table  1.5  Average  Allocation  of  Pension  Fund  Portfolios  in  ECA  Countries,  2004  and  2010  .......  15   1.5  Changes  caused  by  the  crisis  ...........................................................................................................  15       Table  1.6  Second-­�Pillar  Policy  Reactions  in  ECA  Countries:  From  Reversal  to  No  Change  ...........  16   ..........................................  17   Table  1.7  Impact  on  Final  Pensions  of  Different  Types  of  Reductions   Table  1.8  Changes  in  Revenue  and  Expenditure  to  Restore  Fiscal  Stability  in  Three  ECA  Countries,     2011–12  ........................................................................................................................................  18   1.6  Conclusion  .......................................................................................................................................  19   20   Chapter  2.  Diversified  Solutions  to  the  Retirement  Challenge  ...............................   2.1  Introduction  .....................................................................................................................................  20   2.2  The  nature  of  the  retirement  problem  ............................................................................................  20   Figure  2.1  Diversified  Sources  of  Retirement  Income  ..................................................................  21   2.2.1  Claims,  assets,  income,  and  consumption  ...........................................................................  21   Table  2.1  Sources  of  Retirement  Income  and  Risks  to  Good  Outcomes  ......................................  22   2.3  Scale  of  the  retirement  income  challenge  .......................................................................................  23   Figure  2.2  Actual  and  Estimated  Population  Dependency  Rates  in  ECA  Countries,  2010–60  ......  24   Table  2.2  Actual  and  Estimated  Changes  in  Gross  Pension  Expenditures  as  a  Share  of  GDP  in  ECA     Countries,  2010–60  ......................................................................................................................  25   Figure  2.3  Actual  and  Estimated  Old-­�Age  Economic  Dependency  Ratio  in  the  EU,  by  Country,   2010–60  ........................................................................................................................................  26   Figure  2.4  Size  of  Retirement  Challenge  and  Role  of  Offsetting  Policies,  2010–30  ......................  27   2.4  Insights  from  other  countries  and  pension  pillars  ...........................................................................  27   Figure  2.5  Impact  of  Announced  Changes  on  Replacement  Rates  and  Old-­�Age  Poverty  in  Select     Countries,  2010–50  ......................................................................................................................  28   2.4.1  Voluntary  pensions  as  an  alternative  to  mandatory  pensions  ............................................  28   Figure  2.6  A  Spectrum  of  Coverage  from  Mandatory  to  Voluntary  ..............................................  29   Figure  2.7  Mandatory  and  Voluntary  Pension  Coverage  in  Select  Countries  ...............................  30   Figure  2.8  Second-­�  and  Third-­�Pillar  Pensions  as  a  Share  of  GDP  in  ECA  Countries,  2012  .............  32   2.4.2  Comparing  similar  second-­�pillar  systems  in  different  regions  .............................................  32   2.4.3  Fourth  pillar  .........................................................................................................................  33   2.5  Conclusion  .......................................................................................................................................  33   35   Chapter  3.  Issues  Identified  with  Mandatory  Private  Pensions  ..............................   3.1  Introduction  .....................................................................................................................................  35   3.2  Tension  between  the  short  and  long  term  ......................................................................................  35   3.2.1  Financing  of  the  transitional  deficit  .....................................................................................  36   Figure  3.1  Pension  Balances  with  Multi-­�pillar  Pension  Reform  as  a  Share  of  GDP,  1997–2069  ...  37       ...........................  37   Figure  3.2  Old-­�Age  Fiscal  Expenditure  as  a  Share  of  GDP  in  Hungary,  2000–08   Figure  3.3  Privatization  Revenues  as  a  Share  of  GDP  in  Poland,  1999–2010  ...............................  38   Figure  3.4  Projected  Average  Old-­�Age  Pension  Benefits  as  a  Share  of  the  Average  Wage  in     Poland,  2011–71  ...........................................................................................................................  39   3.2.2  Implicit  and  explicit  government  debt  and  the  Stability  and  Growth  Pact  ..........................  40   Figure  3.5  Implicit  and  Explicit  Debt  Liabilities  as  a  Share  of  GDP  in  European  Countries,  2006     and  2011  .......................................................................................................................................  41   3.2.3  Weak  political  consensus  on  the  reforms  ............................................................................  45   Figure  3.6  Average  Second-­�Pillar  Contributions  as  a  Share  of  GDP  in  ECA  Countries,  2002–10  ..  46   3.3  Incomplete  reforms  .........................................................................................................................  46   3.3.1  Fiscal  discipline  in  the  post-­�reform  period  ..........................................................................  47   Figure  3.7  Fiscal  Balance  of  Hungary,  2001–10  ............................................................................  47   Figure  3.8  Fiscal  Balance  of  Poland,  2001–10  ...............................................................................  48   Figure  3.9  National  Savings  Rates  as  a  Share  of  GDP  in  ECA  Countries,  2000–12  ........................  49   3.3.2  The  capital  market  agenda  and  the  abundance  of  bank  financing  ......................................  49   Figure  3.10  Credit  and  Deposit  Growth  Rates  in  the  ECA  Region,  2006–12  .................................  50   3.4  Institutional  and  market  design  issues  affecting  costs  and  investment  ..........................................  51   3.4.1  High  administrative  fees  ......................................................................................................  51   Figure  3.11  Fees  as  a  Share  of  Assets  from  Inception  of  the  New  System,  by  Region  .................  52   Figure  3.12  Fees  as  a  Share  of  Total  Assets  in  Mandatory  Funded  Systems  in  ECA  Countries,  2006     and  2011  .......................................................................................................................................  52   3.4.2  Inefficient  asset  allocation  ...................................................................................................  54   Figure  3.13  Portfolio  Diversification  of  Pension  Funds  as  a  Share  of  Total  Assets  in  ECA  Countries,     2012  ..............................................................................................................................................  55   Figure  3.14  Yield  Curves  in  Poland,  2000–05  (End  of  Year)  ..........................................................  56   Box  3.1  Government  Bond  Market  and  Policy  Reversals:  The  Case  of  Government  Securities  in     Hungarian  Pension  Portfolios  .......................................................................................................  57   3.4.3  Investment  strategy  and  benchmarks  .................................................................................  58   Box  3.2  Is  the  Short-­�Term  Rate  of  Return  the  Right  Metric?  ........................................................  59   3.5  Diversification,  correlations,  and  the  accumulation  and  payout  phase  ..........................................  60   3.5.1  Relative  rates  of  return  and  correlations  .............................................................................  61   Figure  3.15  Wage  Growth  and  Real  Returns  in  Select  ECA  Countries,  Inception  (or  2002)  to  2007   ….  ..................................................................................................................................................  62       Table  3.1  Annual  Correlation  of  Real  Wages  and  Real  Returns  in  ECA  Countries,  Various  Years,     2000–12  ........................................................................................................................................  63   3.5.2  Retirement  ages  and  the  payout  phase  ...............................................................................  63   Table  3.2  Risk  Characteristics  of  Retirement  Products  .................................................................  64   3.6  Conclusion  .......................................................................................................................................  64   65   Chapter  4.  Lessons,  Actions,  and  Further  Work  .....................................................   4.1  Introduction  .....................................................................................................................................  65   4.2  Lessons  and  Recommended  Actions  ...............................................................................................  65   4.2.1  Tax  not  debt  financing  of  the  transitional  pension  deficit  ...................................................  65   4.2.2  The  Stability  and  Growth  Pact  and  funded  pensions  ...........................................................  65   4.2.3  Estimates  of  implicit  debt  ....................................................................................................  66   4.2.4  Cost  reductions  ....................................................................................................................  66   ....................................................................................................................  67   4.2.5  Price  regulation   4.2.6  Portfolio  benchmarks  ..........................................................................................................  67   4.2.7  Guarantees  ..........................................................................................................................  68   4.2.8  Quasi-­�mandatory  coverage  .................................................................................................  69   4.2.9  Retirement  age  and  payout  phase  ......................................................................................  69   4.3  Next  Steps  ........................................................................................................................................  69   70   References  ..............................................................................................................               List  of  Tables,  Figures,  and  Boxes     Table  1.1  Structure  of  Pension  Systems  in  Central  and  Eastern  Europe  ..........................................  5   Table  1.2  Real  GDP  Growth  in  ECA  Countries,  2007–12  .................................................................  9   Table  1.3  Sovereign  Spreads  in  ECA  Countries,  2007–11  .............................................................  12   Table  1.4  Long-­�Term  Sovereign  Bond  Ratings  in  ECA  Countries,  2007–12  ...................................  13   Table  1.5  Average  Allocation  of  Pension  Fund  Portfolios  in  ECA  Countries,  2004  and  2010  .......  15   Table  1.6  Second-­�Pillar  Policy  Reactions  in  ECA  Countries:  From  Reversal  to  No  Change  ...........  16   ..........................................  17   Table  1.7  Impact  on  Final  Pensions  of  Different  Types  of  Reductions   Table  1.8  Changes  in  Revenue  and  Expenditure  to  Restore  Fiscal  Stability  in  Three  ECA  Countries,   2011–12  .....................................................................................................................................................  18   Table  2.1  Sources  of  Retirement  Income  and  Risks  to  Good  Outcomes  ......................................  22   Table  2.2  Actual  and  Estimated  Changes  in  Gross  Pension  Expenditures  as  a  Share  of  GDP  in  ECA   Countries,  2010–60  ...................................................................................................................................  25   Table   3.1   Annual   Correlation   of   Real   Wages   and   Real   Returns   in   ECA   Countries,   Various   Years,   2000–12  .....................................................................................................................................................  63   Table  3.2  Risk  Characteristics  of  Retirement  Products  .................................................................  64     Figure  1.1  Increases  in  Pension  Generosity  in  Poland  after  the  Initial  Reform  ..............................  6   Figure  1.2  Increases  in  Pension  Generosity  in  Hungary  after  the  Initial  Reform  ............................  7   Figure  1.3  Real  GDP  Growth  in  ECA  Countries,  2002–12  ................................................................  8   Figure  1.4  Fiscal  Deficits  as  a  Share  of  GDP  in  ECA  Countries,  2000–12  .......................................  10   Figure  1.5  Sovereign  Debt  as  a  Share  of  GDP  in  ECA  Countries,  2000–12  ....................................  11   Figure  1.6  Ratio  of  Debt  to  GDP  in  ECA  Countries  and  Select  EU-­�15  Countries,  2008  ..................  11   Figure  1.7  Nominal  Returns  of  Mandatory  Pension  Funds  Relative  to  MSCI  Europe,  2004–12  ...  14   Figure  2.1  Diversified  Sources  of  Retirement  Income  ..................................................................  21   Figure  2.2  Actual  and  Estimated  Population  Dependency  Rates  in  ECA  Countries,  2010–60  ......  24   Figure   2.3   Actual   and   Estimated   Old-­�Age   Economic   Dependency   Ratio   in   the   EU,   by   Country,   2010–60  .....................................................................................................................................................  26   Figure  2.4  Size  of  Retirement  Challenge  and  Role  of  Offsetting  Policies,  2010–30  ......................  27   Figure  2.5  Impact  of  Announced  Changes  on  Replacement  Rates  and  Old-­�Age  Poverty  in  Select   Countries,  2010–50  ...................................................................................................................................  28   Figure  2.6  A  Spectrum  of  Coverage  from  Mandatory  to  Voluntary  ..............................................  29   Figure  2.7  Mandatory  and  Voluntary  Pension  Coverage  in  Select  Countries  ...............................  30   Figure  2.8  Second-­�  and  Third-­�Pillar  Pensions  as  a  Share  of  GDP  in  ECA  Countries,  2012  .............  32       Figure  3.1  Pension  Balances  with  Multi-­�pillar  Pension  Reform  as  a  Share  of  GDP,  1997–2069  ...  37   ...........................  37   Figure  3.2  Old-­�Age  Fiscal  Expenditure  as  a  Share  of  GDP  in  Hungary,  2000–08   Figure  3.3  Privatization  Revenues  as  a  Share  of  GDP  in  Poland,  1999–2010  ...............................  38   Figure   3.4   Projected   Average   Old-­�Age   Pension   Benefits   as   a   Share   of   the   Average   Wage   in   Poland,  2011–71  ........................................................................................................................................  39   Figure   3.5   Implicit   and   Explicit   Debt   Liabilities   as   a   Share   of   GDP   in   European   Countries,   2006   and  2011  .........................................................................................................................................................    ...................................................................................................................................................................  41   Figure  3.6  Average  Second-­�Pillar  Contributions  as  a  Share  of  GDP  in  ECA  Countries,  2002–10  ..  46   Figure  3.7  Fiscal  Balance  of  Hungary,  2001–10  ............................................................................  47   Figure  3.8  Fiscal  Balance  of  Poland,  2001–10  ...............................................................................  48   Figure  3.9  National  Savings  Rates  as  a  Share  of  GDP  in  ECA  Countries,  2000–12  ........................  49   Figure  3.10  Credit  and  Deposit  Growth  Rates  in  the  ECA  Region,  2006–12  .................................  50   Figure  3.11  Fees  as  a  Share  of  Assets  from  Inception  of  the  New  System,  by  Region  .................  52   Figure  3.12  Fees  as  a  Share  of  Total  Assets  in  Mandatory  Funded  Systems  in  ECA  Countries,  2006   and  2011  ....................................................................................................................................................  52   Figure  3.13  Portfolio  Diversification  of  Pension  Funds  as  a  Share  of  Total  Assets  in  ECA  Countries,   2012  ...........................................................................................................................................................  55   Figure  3.14  Yield  Curves  in  Poland,  2000–05  (End  of  Year)  ..........................................................  56   Figure   3.15   Wage   Growth   and   Real   Returns   in   Select   ECA   Countries,   Inception   (or   2002)   to   2007  ...................................................................................................................................................................  62     Box   3.1   Government   Bond   Market   and   Policy   Reversals:   The   Case   of   Government   Securities   in   Hungarian  Pension  Portfolios  ....................................................................................................................  57   Box  3.2  Is  the  Short-­�Term  Rate  of  Return  the  Right  Metric?  ........................................................  59         Abbreviations     ATP   Danish  Labor  Market  Supplementary  Pension  (Denmark)   ECA   Europe  and  Central  Asia   EFC   Economic  and  Financial  Committee   EPF   Employees  Provident  Fund  (Malaysia)   EU   European  Union   GDP   Gross  domestic  product   IMF   International  Monetary  Fund   MTO   Medium-­�term  budgetary  objective   NEST   National  Employment  Savings  Trust  (United  Kingdom)   OECD   Organisation  for  Economic  Co-­�operation  and  Development   PAYG   Pay-­�as-­�you-­�go   PFMC   Pension  fund  management  company   PPM   Premium  Pensions  Authority  (Sweden)   SGP   Stability  and  Growth  Pact     SODRA   Social  security  (Lithuania)   TSP   Thrift  Savings  Plan  (United  States)   UCITS     Undertakings  for  collective  investment  in  transferable  securities   VAT   Value  added  tax       1     Executive  Summary     This  report  examines  the  impact  of  the  financial  crisis  on  privately  funded  pensions  in  the  Europe  and   Central  Asia  region.  It  describes  their  pension  systems  and  the  impact  of  the  crisis.  It  explores  whether   funded  private  systems  have  a  continuing  role  to  play  in  meeting  the  pension  challenge.  It  then  focuses   on  why  the  crisis  led  to  the  changes  that  took  place  and  the  policy  implications  of  those  changes.   Chapter  1  sets  out  the  key  facts.  It  describes  the  original  pension  reforms  and  what  the  pension  systems   looked  like  before  the  recent  financial  crisis.  It  then  briefly  describes  the  global  financial  crisis  and  the   changes   that   were   made   to   the   private   pension   system   as   a   result.   A   companion   paper   sets   out   the   longer  and  deeper  historical  context  in  which  these  events  occurred  (Schwarz  and  others  2013).   Chapter  2  asks  what  these  facts  tell  us  about  the  role,  if  any,  that  privately  funded  pension  pillars  have  in   a  country’s  overall  pension  system.  It  argues  that  a  funded  second  pillar—mandatory  private  pensions— has   a   role   to   play   in   a   diversified   approach   to   meeting   the   pension   challenge.1   A   diversified   approach   entails   a   mix   of   state   and   privately   funded   pensions   as   well   as   pension   and   non-­�pension   sources   of   income  and  consumption  in  retirement.   Chapters  3  and  4  build  on  the  premise  that  funded  systems  have  a  critical  role  to  play  as  one  part  of  a   diversified   system.   Chapter   3   highlights   the   weaknesses   in   the   mandatory   funded   schemes   that   were   exposed  by  the  crisis.  Chapter  4  describes  policy  responses  to  these  weaknesses.  It  sets  out  an  agenda   for   change.   While   much   of   the   original   foundations   are   sound,   all   countries   with   mandatory   and   voluntary  pensions  need  to  focus  on  several  critical  areas.  This  is  true  whether  their  pension  systems  are   mature,   relatively   new,   or   newly   implemented.   The   region   is   very   diverse.   Not   all   proposals   will   apply   with   equal   weight   to   each   country.   But   given   the   very   long   time   periods   over   which   pension   systems   develop   and   pay   out   benefits,   it   is   likely   that   all   countries   will   eventually   face   some   or   all   of   these   challenges.   Some   were   well   known   but   not   tackled   before   the   financial   crisis.   But   the   crisis   exposed   them   more   starkly   and   caused   greater   problems   than   if   the   challenges   had   been   addressed   earlier.   Taking  action  makes  sense—whether  to  remedy  current  problems  or  to  prevent  future  ones.   The   report   distinguishes   between   three   main   responses   to   the   crisis:   reversals,   which   end   mandatory   funded  private  plans  by  collapsing  them  back  into  the  state  pension  system;  reductions  (both  permanent   and   temporary),   which   reduce   contributions   to   the   privately   funded   systems   but   keep   the   second   pillar;   and   resolution,   where   countries   have   made   no   adjustments   to   the   privately   funded   pension   system   despite  the  challenges  of  the  crisis.     There   are   two   overall   messages.   The   first   is   that   reversals—collapsing   funded   private   pensions   back   into   the  state  system—are  the  wrong  answer.  Long-­�term  damage  is  done  for  no  long-­�term  benefit.  If  the  first   message  is  “do  not  reverse,�  the  second  is  “pursue  reforms  actively.�  The  scale  of  the  pension  challenge   remains.   The   original   drivers   of   reform   were   not   wrong.   If   anything,   they   have   become   more   acute.   Funded  pensions,  particularly  mandatory  funded  pensions,  still  have  a  role  to  play  as  part  of  a  diversified   system   for   ensuring   a   decent   and   sustainable   retirement   income   for   all.   But   the   robustness   to   shocks   and  the  outcomes  for  members  can  and  must  be  improved.  So  the  reform  messages  certainly  apply  to                                                                                                                           1   The   first   pillar   consists   of   state   income-­�related   pensions   (often   on   a   pay-­�as-­�you-­�go   basis),   the   second   pillar   consists  of  mandatory,  funded,  defined  contribution  schemes,  the  third  pillar  consists  of  voluntary,  private  pension   schemes,  and  the  fourth  pillar  includes  alternative  sources  of  retirement  income  and  consumption,  including  other   financial  assets,  family  support,  and  housing.   2     those  countries  that  have  lowered  contributions  to  the  system.  But  most  of  the  messages  apply  more   generally   to   other   countries   in   the   region,   including   those   that   have   resolved   to   stick   to   the   reform   path   in   the   face   of   significant   shocks.   The   recommendations   offered   here   are   intended   to   improve   the   outcomes   for   a   given   level   of   contributions.   But   for   countries   that   have   lowered   the   contributions,   reform   can   only   achieve   so   much.   If   the   contributions   are   not   increased,   and   the   reductions   are   not   made  up,  workers  will  retire  on  lower  pensions  in  the  future.  For  most  countries,  there  is  time  to  ensure   that  the  impact  of  the  crisis  is  not  still  being  felt  in  30  years,  when  current  members  retire  with  lower   pensions.   Importantly,   higher   contributions   should   not   always   be   linked   to   wages.   Other   sources   of   contributions,   such   as   consumption   taxes,   are   also   needed   to   reduce   the   labor   market   impact   of   high   labor  taxes.   What  does  reform  mean?     • Use   taxes   instead   of   debt   financing   for   financing   the   transitional   pension   deficit   created   by   pension  reforms.     • Make  the  Stability  and  Growth  Pact  truly  reward,  not  penalize,  governments  for  trying  to  ensure   pension  promises  are  funded  and  not  hidden  in  implicit  (and  risky)  unfunded  liabilities.   • Reduce   costs   to   improve   outcomes.   Reducing   management   costs   by   0.5   percent   (50   basis   points)  could  increase  future  pension  plans  by  some  10–15  percent  over  current  levels.   • Diversify   the   asset   allocation   of   pension   portfolios.   Key   solutions   include   using   portfolio   benchmarks   set   externally   or   ensuring   that   the   demand   side   of   the   market   has   the   combination   of   scale,   expertise,   and   member-­�focused   governance   to   ensure   asset   allocation   for   the   long   term.   • Realize  that  improving  performance  with  regard  to  costs  (and  investment)  may  require  changing   the   “industrial   organization�   or   the   demand,   supply,   and   distribution   dynamics   of   the   pension   industry  to  tackle  issues  created  by  vertically  integrated  pension  management  companies.   • Develop   a   proactive   agenda   of   capital   market   reforms;   do   not   expect   development   simply   because  a  private  pension  system  is  in  place.   • Ensure   labor   market   policies   that   make   working   lives   long   enough   to   support   adequate   replacement  rates  and  make  retirement  ages  rise  with  longevity.   • Secure  pensions  until  death  so  that  old-­�age  poverty  is  defeated,  not  postponed.     • Create,   develop,   and   sustain   political   and   public   acceptance   of   the   need   for   long-­�term   contributions  to  tackle  demographic  challenges.   3     Chapter  1.  Pre-­�crisis  Pensions  and  the  Impact  of  the  Crisis     1.1  Introduction   This   report   looks   at   the   impact   of   the   financial   crisis   on   privately   funded   pensions   in   the   Europe   and   Central  Asia  (ECA)  region.  This  chapter  sets  out  the  key  facts.  It  describes  the  original  pension  reforms  to   establish   what   pension   systems   in   the   region   looked   like   pre-­�crisis.   It   then   reviews   the   impact   of   the   global  financial  crisis  and  the  changes  that  have  been  made  to  private  pensions  as  a  result.     The   next   chapter   takes   these   core   facts   and   asks   what   they   tell   us   about   the   role,   if   any,   of   privately   funded  pension  pillars  in  a  country’s  overall  pension  system.  It  argues  that  funded  pillars  have  a  role  to   play   as   part   of   a   diversified   approach   to   meeting   the   pension   challenge.   Addressing   this   challenge   requires   a   mix   of   state   and   privately   funded   provision   as   well   as   pension   and   non-­�pension   sources   of   income  and  consumption  in  retirement.   The   final   two   chapters   build   on   this   premise.   Chapter   3   highlights   the   weaknesses   that   have   been   exposed  by  the  crisis.  The  final  chapter  sets  out  policy  responses  to  these  weaknesses  and  proposes  an   agenda  for  change.  It  argues  that  much  of  the  original  foundations  are  sound,  but  that  all  countries  with   mandatory  pensions  need  to  focus  on  certain  critical  areas.       1.2  The  original  pension  reforms   A   number   of   countries   in   the   region   introduced   mandatory,   privately   funded   pension   funds   (second   pillar)  in  the  late  1990s  and  early  2000s.  Hungary  (1998)  and  Poland  (1999)  led  the  way.  Five  countries   introduced  funded  pension  pillars  in  2001  and  2002  (Latvia,  Estonia,  Bulgaria,  Croatia,  Kosovo).   A  steady   flow   of   countries   have   introduced   funded   pillars.   They   include   Lithuania   (2004),   the   Slovak   Republic   (2005),  the  former  Yugoslav  Republic  of  Macedonia  (2006),  and  Romania  (2008).   The  introduction  of  mandatory  second  pillars  was  part  of  a  wider  strategy  consisting  of  reforms  to  the   public  sector  pay-­�as-­�you-­�go  (PAYG)  schemes  (first  pillar).  Among  other  objectives,  they  aimed  to  address   growing  and  unsustainable  deficits  in  these  schemes  (Schwarz  and  others  2013).     These  reforms  are  summarized  in  table  1.1.  This  sets  out  key  elements  of  the  changes  in  the  first  pillar  as   well   as   the   headline   features   of   the   (new)   second   pillars   and   their   date   of   enactment.   As   the   final   column  shows,  by  the  start  of  the  crisis,  most  countries  had  a  reformed  first  pillar  and  a  relatively  new   second   pillar.   They   also   typically   had   voluntary   funded   private   pensions   (third   pillar).   It   is   crucial   to   understand   what   each   of   these   pillars   can   and   cannot   deliver   with   regard   to   ensuring   that   a   country   delivers   pension   outcomes   for   its   people   that   achieve   adequacy   and   coverage   alongside   efficiency,   sustainability,  and  security.  A  central  argument  of  this  report  is  that  no  single  part  of  a  pension  system   can  deliver  all  of  the  objectives  required.       4     Table  1.1  Structure  of  Pension  Systems  in  Central  and  Eastern  Europe   Pension'System'Overview'in'Eastern'and'Central'Europe Public3 Second3Pillar3 Enactment3 Pillars3as3of32008 Public3(PAYG)31st3Pillar3Reform Who3Participates Country Scheme3Type Contributions Date 1st 2nd 3rd Ret$Age:$60/55→63/63 Estonia DB Benefits:$55%$$best$5$+1%→flat$ 6% 2002 Mandatory$only$for$new$entrants X X X component+earnings$related$points Notional$ Mandatory$for$new$and$young$ Ret$Age:$60/55→62/62 accounts workers,$<30,$Voluntary$30I50 Latvia 2%↗8% 2001 X X X Benefits:$55%$$highest$5$+1%→full$ career$based$on$notional$accumulation Voluntary$for$current$and$new$ Ret$Age:60/55→62.5/60 DB workers Lithuania 2.5%↗5.5% 2004 X X X Benefits:$55%$highest$5$plus$1%→$ earnings$unrelated$and$related$ Notional$ component Ret$Age:$65/60$(many$exemptions)$ Mandatory$for$new$and$young$ accounts →65/60$(fewer$exemptions) workers,$<30,$Voluntary$30I50 Poland 7.30% 1999 X X X Benefits:$Flat$+1.3%$→full$career$ notional$accumulation Ret$Age:$60/53I57→62/62 Slovak$R. Points$ Benefits:$2%$first$25,$1%$ 9% 2005 Mandatory$only$post$1983$born X X X thereafter→1.19%$full$career Ret$Age:$60/55→62/62 Hungary DB Benefits:$33%$for$first$10$ 6%↗8% 1998 Mandatory$only$for$new$entrants X X X +2.5%→1.65%$full$career Mandatory$for$<35,$voluntary$ Ret$Age:$62/57→65/60 36I45 Romania DB 2%↗3% 2008 X X X Benefits:$75%$first$30/20$ year+1%→full$career$points$based Ret$Age:$60/55→63/60 Bulgaria DB 2%↗5% 2002 Mandatory$for$workers$<42. X X X Benefits:$55%$highest$3→full$career$1% Mandatory$for$workers$<40,$ Ret$Age:$65/60 Croatia Points$ 5% 2002 voluntary$40I50 X X X Benefits:$0.825%$full$career Ret$Age:$64/62 Macedonia DB Benefits:$defined$by$minimum$wage$ 7.42% 2006 Mandatory$for$new$entrants X X ratio$and$total$years$of$participation Universal$flat$ Mandatory$for$all$working$ Ret$Age:$65/65 Kosovo for$all$at$$>65 10% 2002 habitual$residents,$<55 X X X Benefits:$defined$annually$by$the$value$ of$minimum$consumption$basket Notional$ Ret$Age:$60/55 Russia accounts Benefits:$defined$by$individual$lifetime$ 3%↗6% 2002 Mandatory$only$post$1966$born X X X accumulations Ret$Age:$63/58 Kazakhstan $DB$ Benefits:$60%$of$monthly$average$of$ 10% 1998 Mandatory$for$all X X X best$consecutive$three$years Sources:  Pension  fund  supervision  agencies  of  respective  countries.  Kąsek,  Laursen,  and  Skrok  2008;  Schwarz  2011;   OECD  2012a.     Note:  Newer  reforms  such  as  those  in  the  Kyrgyz  Republic  and  Armenia  could  be  included  in  future  work.  For  the   relative  importance  of  the  second  pillar  in  each  country,  see  the  modeling  presented  in  OECD  (2011b)  and  Pallares-­� Miralles,  Romero,  and  Whitehouse  (2012),  among  other  publications.     1.3  The  evolution  of  pension  systems  after  the  initial  reforms   A  central  problem  in  pension  policy  is  that  systems  need  to  be  designed  to  deliver  over  60–80  years,  but   they  operate  as  an  important  part  of  the  (daily)  political  debate.  When  constructing  any  reform  package,   the  balance  between  adequacy  and  sustainability  (for  government  and  for  workers  and  employers)  has   to   be   based   on   long-­�term   projections,   which   have   to   be   based   on   assumptions.   These   assumptions   relate,   for   example,   to   life   expectancy,   investment   returns,   GDP   and   wage   growth,   labor   force   participation,   contribution   rates,   and   contribution   density   (how   many   years   a   person   actually   contributes).   5     Even  if  the  assumptions  are  correct,  volatility  in  the  economic  cycle  means  that   nearly  every  important   variable   will   show   regular   and   persistent   deviations   from   its   long-­�term   path.   In   the   pre-­�crisis   period,   the   variables   performed   above   expectations,   while   in   the   crisis   period,   they   often   performed   below   expectations.   Unfortunately   before   the   crisis,   when   many   ECA   countries   experienced   strong   economic   growth,   governments  expanded  pension  benefits  above  the  initial  reform  package.  In  other  words,  a  period  of   temporary   above-­�trend   performance   led   to   permanent   increases   in   the   generosity   of   the   system.   The   experiences  of  Poland  and  Hungary  provide  a  good  case  study  of  this  impact  (see  figures  1.1  and  1.2).     Figure  1.1  Increases  in  Pension  Generosity  in  Poland  after  the  Initial  Reform     Source:  National  authorities.     6     Figure  1.2  Increases  in  Pension  Generosity  in  Hungary  after  the  Initial  Reform     Source:  National  authorities.     1.4  The  financial  crisis   The  countries  of  the  ECA  region  entered  the  financial  crisis  with  pension  systems  that  had  often  changed   profoundly  in  the  past  decade.  These  initial  reforms  had  been  followed  by  a  very  supportive  economic   environment,   as   experienced,   on   average,   globally   during   the   period   of   the   “great   moderation.�   The   reforms   to   the   second   pillars   were   part   of   an   effort   to   reduce   the   political   risks   to   future   pensioners   from  unsustainable  promises  “funded�  by  implicit  government  debt.  Crucially  they  did  this  by  investing   funded  contributions  in  financial  markets,  thereby  taking  on  market  risk.  But  as  the  reaction  to  the  crisis   shows,  when  the  pressures  are  great  enough,  political  risk  can  reappear—especially  when  government   seeks   to   divert   contributions   earmarked   for   funded   pensions   to   finance   a   (growing)   general   budget   deficit.   The   financial   crisis   has   already   been   studied   extensively   (see   Narain,   Ötker,   and   Pazarbasioglu   2012;   World   Bank   2009,   2010;   OECD   2009,   2010),   so   this   report   does   not   repeat   the   analysis   in   detail.   But   it   is   important   to   highlight   the   impact   of   the   crisis   on   broad   economic   aggregates   such   as   GDP,   the   fiscal   position,   and   the   capital   market   to   provide   the   backdrop   for   the   changes   that   occurred.   And   it   is   essential  to  highlight  the  diversity  of  experience.  There  is  no  simple  link  between  those  countries  having   the  worst  experience  of  the  crisis  being  those  deciding  on  reversals  or  reductions.     Figure   1.3   shows   GDP   over   the   past   10   years.   There   are   three   important   messages.   The   first   is   that   GDP   growth  rates  were  healthy  in  the  period  after  the  original  reforms  and  up  until  the  crisis.  This  experience   was  positive,  but  it  made  the  long-­�term  problem  seem  simpler  than  it  was.  It  led  to  some  improvements   in   the   generosity   of   the   systems,   which   exacerbated   the   pension   problem.   In   other   words,   it   led   to   changes  that  offset  some  of  the  tough  decisions  made  in  the  original  reforms.   7     The  second  message  is  that  the  crisis  had  a  very  clear  impact  on  real  GDP.  There  was  a  wide  range  of   experience—from   Poland   and   Kosovo,   which   avoided   any   negative   growth   on   an   annual   basis,   to   Estonia,   Latvia,   and   Lithuania,   which   had   negative   growth   rates   of   −14.3,   −17.7,   and   −14.8   percent,   respectively,  in  2009.  As  we  will  see,  the  decision  to  reverse  or  reduce  payments  into  the  second  pillar  is   not  well  correlated  with  the  GDP  story.   The  third  message  is  that  all  countries  rebounded  quickly,  followed  by  subdued  growth.  In  2011,  Estonia,   Latvia,   and   Lithuania   reported   annual   growth   above   5   percent.   The   size   of   the   decline   in   output   in   a   number   of   countries   means   that   reaching   pre-­�crisis   levels   of   output  will   take   many   years   of   growth.   But   the   performance   has   been   relatively   strong   considering   the   size   of   the   headwinds   from   weak   global   demand   and   the   continuing   sovereign   debt   crisis   in   the   European   Union  (EU).   For   more   information   see   World   Bank   (2011).   Whether   GDP   growth   can   attain   and   sustain   pre-­�crisis   levels   is   a   fundamental   factor   in  determining  outcomes  in  the  future.  Table  1.2  shows  the  figures.     Figure  1.3  Real  GDP  Growth  in  ECA  Countries,  2002–12   Real%GDP%Growth%Rates%(200222012)% 15$ 10$ 5$ 0$ %$ 2002$ 2004$ 2006$ 2008$ 2010$ 2012$ !5$ !10$ !15$ !20$ Estonia$ Latvia$ Lithuania$ Poland$ Slovak$R.$ Hungary$ Romania$ Bulgaria$ CroaDa$ Macedonia$ Kosovo$ Russian$ Kazakhstan$   Source:  IMF  database.     8     Table  1.2  Real  GDP  Growth  in  ECA  Countries,  2007–12     Source:  IMF  database.     Note:  2012  figures  are  based  on  estimates.     1.4.1  Government  finances     Figure  1.4  shows  fiscal  deficits  in  the  ECA  countries.  For  many  countries,  fiscal  deficits  were  already  large   before   the   crisis,   despite   generally   strong   economic   growth.   Therefore,   some   economies   entered   the   crisis   with   little   fiscal   space.   So   they   needed   to   find   short-­�term   savings   on   both   the   revenue   and   expenditure  sides  from  across  the  whole  government  budget  more  quickly  than  if  they  had  had  greater   fiscal   prudence   beforehand.   The   weak   fiscal   picture   in   the   pre-­�crisis   years   is   particularly   important   for   the   pension   story.   First,   the   planned   fiscal   consolidation   as   part   of   the   reforms   to   ensure   that   future   funded   liabilities   were   financed   with   taxes   rather   than   debt   in   general   did   not   happen.   Second,   when   the   crisis   started,   the   public   and   privately   funded   pension   system   was   at   risk   because   both   pillars   constituted  large  components  of  government  spending.   As   with   GDP,   there   is   a   considerable   range   in  the   fiscal   numbers,   although   again   the   size   of   deficit   is   not   necessarily  related  to  the  scale  of  reductions  in  the  contributions  to  pensions.     9     Figure  1.4  Fiscal  Deficits  as  a  Share  of  GDP  in  ECA  Countries,  2000–12   Government)Fiscal)De�cit)Dynamic)(200062012)) 10$ 8$ 6$ 4$ 2$ 0$ %$of$GDP$ !2$ !4$ !6$ !8$ !10$ !12$ !14$ 2000$ 2002$ 2004$ 2006$ 2008$ 2010$ 2012$ Estonia$ Latvia$ Lithuania$ Poland$ Slovak$R.$ Hungary$ Romania$ Bulgaria$ CroaFa$ Macedonia$ Kosovo$ Russia$ Kazakhstan$   Source:  IMF  database.       The   evolution   of   gross   sovereign   debt   to   GDP   closely   mirrors   the   rising   deficits   caused   by   the   crisis.   However,   as   shown   in   figure   1.5,   many   ECA   countries   entered   the   crisis   with   relatively   low   levels   of   government   debt   to   GDP.   Important   exceptions   were   Hungary   and   Poland.   Strong   GDP   growth   had   helped  to  keep  the  ratio  of  government  debt  to  GDP  relatively  low.  This  changed  quickly  as  fiscal  policy   aimed   to   support   growth   at   the   cost   of   rapidly   rising   debt   levels.   The   need   to   access   highly   stressed   global  capital  markets  limited  the  government’s  ability  to  use  fiscal  policy  to  support  demand  (see  figure   1.5).     10     Figure  1.5  Sovereign  Debt  as  a  Share  of  GDP  in  ECA  Countries,  2000–12     Gross%Sovereign%Debt%(200022012)% 90" 80" 70" 60" 50" %"of"GDP" 40" 30" 20" 10" 0" 2000" 2002" 2004" 2006" 2008" 2010" 2012" Estonia" Poland" Slovak"R." Hungary" Romania" Bulgaria" CroaGa" Macedonia" Russia" Kazakhstan" Latvia" Lithuania"   Source:  IMF  database.     Many   of   the   debt   levels   at   the   start   of   the   crisis   were   below   EU-­�15   comparators   (see   figure   1.6),   but   the   drivers  of  change  vary  by  country.  For  example,  in  Poland  debt  levels  before  the  crisis  were  allowed  to   approach   the   constitutional   ceiling   of   government   debt   to   GDP   of   60   percent.   Intended   to   encourage   fiscal   sustainability,   the   ceiling   meant   that   Poland   had   to   react   more   quickly   than   would   have   been   required   if   pre-­�crisis   fiscal   policy   had   been   tougher   and   if   debt   rather   than   taxes   had   financed   the   transition  to  a  funded  pension  system.       Figure  1.6  Ratio  of  Government  Debt  to  GDP  in  ECA  Countries  and  Select  EU-­�15  Countries,  2008   Gross%Debt%as%%%of%GDP%in%2008% 80" 73" 67" 70" 60" 52" 47" 44" 50" 39" 40" 28" 29" 30" 21" 17" 16" 15" 20" 14" 8" 10" 5" 0" a" " " ov " Hu ." Bu " Cr " " " " Ge d" " Ro y" " Sw " ia ia d ia ia M aEa ia ia en UK y "R ni ar an n an tv an an ar on ss ak ed la to ng Ru La o l lg rm hu m Po Ire ed Es Lit ac Sl   Source:  IMF  database.   11       1.4.2  Financial  market  conditions     Sovereign  bond  spreads   The   financial   markets   did,   to   some   extent,   differentiate   between   the   experiences   of   the   different   countries   in   the   ECA   region.   Table   1.3   shows   the   spreads   between   government   bonds   in   the   different   markets   and   German   10-­�year   bonds.   Spreads   were   very   low   in   2007,   but   widened   dramatically   as   the   crisis   took   hold.   Poland,   whose   fundamentals   were   relatively   strong,   saw   an   increase   in   spreads   of   2   percentage   points   between   2007   and   2009.   Even   strong   performers   faced   problems   with   the   flight   to   “quality.�   For   some   countries,   particularly   Latvia,   Lithuania,   Hungary,   and   Romania,   spreads   widened   dramatically,  ranging  between  6.7  and  11  percent  in  2009.     Table  1.3  Sovereign  Spreads  in  ECA  Countries,  2007–11   Sovereign3Bond3Spreads3 (German;10Y;Sov;as;a;benchmark,;June;1) Country 2007 2008 2009 2010 2011 Estonia n.a. n.a. n.a. n.a n.a Latvia 1.06 1.73 9.28 7.58 2.98 Lithuania 0.01 0.81 11.03 2.61 2.16 Poland 0.96 1.90 2.87 3.33 2.99 Slovak;R. 0.10 0.42 1.61 1.19 1.50 Hungary 2.15 3.98 6.68 5.06 4.33 Romania 2.43 2.63 7.79 4.56 4.53 Bulgaria 0.01 0.65 3.83 3.67 2.50 Russia 5.81 5.43 5.04 6.43 5.60 Croatia 0.12 1.07 4.26 1.93 1.07 German;Yields 4.56 4.52 3.47 2.54 2.89   Sources:  European  Central  Bank;  European  Commission.   Note:  Estonia  does  not  have  sovereign  long-­�term  debt  securities  on  the  market.     Credit  ratings  of  sovereign  bonds  and  explicit  versus  implicit  debt   As   is   discussed   in   chapter   3,   one   problem   is   that   levels   of   and   changes   in   implicit   rather   than   explicit   debt   appear   to   have   little   impact   on   a   country’s   credit   rating   (see   Cuevas   and   others   2008).2   This   produces  an  incentive  to  “hide�  future  pension  liabilities  in  implicit  debt  rather  than  to  pre-­�fund  them,   through   either   government   reserves   or   second   pillars.   However,   the   rating   agencies   should   not   necessarily  be  identified  as  misrepresenting  the  underlying  picture  of  a  country’s  total  debt  position.  It   does  matter  whether  there  are  explicit  contractual  creditors  or  not.  This  is  because  pension  promises  by                                                                                                                           2  Credit  ratings  used  by  Standard  and  Poor’s  are  defined  as  follows:  A  indicates  strong  capacity  to  meet  financial   commitments,   but   somewhat   susceptible   to   adverse   economic   conditions   and   changes   in   circumstances;   BBB   indicates   adequate   capacity   to   meet   financial   commitments,   but   more   subject   to   adverse   economic   conditions;   BBB−   indicates   lowest   investment   grade;   and   BB+   indicates   highest   speculative   grade.   Credit   ratings   above   and   below  BBB−  are  considered  investment  grade  and  junk  bonds,  respectively.   12     governments   are   subject   to   often   very   significant   “defaults.�   Both   the   original   reform   programs   and   changes  made  during  the  crisis  and  since  have  reduced  the  cost  of  the  pension  promises  governments   have   made—for   example,   by   increasing   the   retirement   age   so   that   pensions   are   paid   later   and   for   a   shorter   number   of   years.   As   we   highlight   in   the   next   chapter,   pension   risk   is   minimized   when   contributors  diversify  the  sources  of  retirement  income.     Table  1.4  shows  the  change  in  ratings  over  the  crisis  period.  Among  other  considerations,  Poland’s  credit   rating   remained   unchanged   because   the   package   of   responses   to   the   crisis   allowed   the   country   to   avoid   a  recession  and  maintain  stable  government  debt  levels.  Hungary,  for  example,  was  downgraded  from   investment  grade  to  junk  status.  Latvia  was  downgraded  and  then  partially  upgraded.     Table  1.4  Long-­�Term  Sovereign  Bond  Ratings  in  ECA  Countries,  2007–12     Long0Term0Sovereign0Ratings0by0Standard0and0Poors'0000000000000000000000000000000000000000000000000000000 (in0foreign0currency,0from020070to02012,0as0of0June01) Country 2007 2008 2009 2010 2011 2012 Estonia A A A A) AA) AA) Latvia BBB+ BBB+ BB+ BB+ BB+ BBB) Lithuania A A) BBB BBB BBB BBB Poland A) A) A) A) A) A) Slovak5R. A A A+ A+ A+ A Hungary BBB+ BBB+ BBB) BBB) BBB) BB+ Romania BBB) BBB) BB+ BB+ BB+ BB+ Bulgaria BBB+ BBB+ BBB BBB BBB BBB Croatia BBB BBB BBB BBB BBB) BBB) Russia BBB+ BBB+ BBB BBB BBB BBB Macedonia BB+ BB+ BB+ BB BB BB   Source:  Standard  and  Poor’s.     Investment  returns   It   is   also   important   to   highlight   what   happened   to   different   investment   markets.   One   of   the   most   important   policy   areas   in   a   second   pillar   is   asset   allocation.   Given   the   mandatory   nature   of   pension   systems,   governments   typically   regulate   the   investments   of   pension   funds.   Unfortunately,   these   regulations   may   be   driven   by   goals   other   than   optimizing   future   pensions.   These   include   the   need   to   fund   fiscal   deficits,   to   maintain   low   volatility   in   returns,   and   even   to   direct   investments   to   specific   sectors  of  the  economy,  among  others.3     As   argued   later,   inadequate   investment   regulations   have   undermined   an   essential   benefit   of   funded   pillars.   But   decisions   on   asset   allocation   are   also   influenced   by   what   are   considered   to   be  “appropriate�   or  “safe�  investments  for  pension  funds.  Figure  1.7  shows  how  returns  differed  between  2004  and  2012.   The   optimal   long-­�term   strategy   may   involve   high   return   volatility.   These   strategies   may   bring   high   volatility   in   the   short   term.   But   if   properly   implemented,   they   will   also   minimize   the   pension   risk   at   retirement  age,  which  is  the  final  objective  of  a  pension  system.  The  use  of  well-­�designed  and  targeted   schemes  of  guarantees—for  example,  guarantees  of  the  value  of  contributions  at  retirement  age—can                                                                                                                           3  The  Russian  Federation  has  used  the  default  pension  fund  to  satisfy  government’s  social  policy  objectives.   13     help   to   increase   the   tolerance   of   people   and   politicians   for   short-­�term   volatility   and   therefore   avoid   unnecessary   distortions   in   asset   allocation   (Antolin   and   others   2011).   These   guarantees   need   to   be   priced  at  market  value  and  provided  by  an  independent  agent.       Figure  1.7  Nominal  Returns  of  Mandatory  Pension  Funds  Relative  to  MSCI  Europe,  2004–12     Annual"Nominal"Returns"of"Mandatory"Pension"Funds" 40$ 30$ 20$ 10$ 0$ %" 2004$ 2005$ 2006$ 2007$ 2008$ 2009$ 2010$ 2011$ 2012$ !10$ !20$ !30$ !40$ !50$ Estonia$ Latvia$ Lithuania$ Poland$ Slovak$R.$ Romania$ Bulgaria$ CroaDa$ Macedonia$ Russia$ Kazakhstan$ MSCI$European$Equity$Index$   Sources:  National  regulatory  authorities;  MSCI.   Note:   Returns   for   Russia   refer   to   the   performance   of   Vnesheconombank’s   “Enhancement�   portfolio,   which   constitutes  more  than  70  percent  of  total  second-­�pillar  assets.     Table   1.5   shows   allocations,   by   major   asset   classes,   for   different   countries   between   2004   and   2010.   The   allocation  to  fixed-­�income,  particularly  government,  bonds  was  relatively  high  in  the  early  years  of  the   reform   process.   For   some   countries,   this   changed   significantly   over   time,   for   example,   in   Estonia.   In   Poland   the   high   allocation   to   bonds   did   not   produce   particularly   low   returns,   as   the   secular   fall   in   interest   rates   over   the   first   decade   of   the   system   led   to   large   capital   gains   for   the   portfolios.   This   highlights  that  the  (very)  long-­�run  recommendation  to  have  a  diversified  portfolio  in  equities  needs  to   be   seen   in   light   of   the   potential   for   a   country   to   reap   a   “stability�   bounce   as   interest   rates   fall   significantly  as  the  economy  develops.  But  such  a  bounce  is  not  guaranteed  and  will  not  deliver  the  best   long-­�term  result  for  pensioners  if  the  asset  allocation  does  not  subsequently  adjust.     14     Table  1.5  Average  Allocation  of  Pension  Fund  Portfolios  in  ECA  Countries,  2004  and  2010     Pension2Fund2Portfolio2Allocation2in220042and220102(as