1. Ist die Wende erreicht ? Gläubigerbeteiligung bei Bankrestrukturierungen in Europa Ist die Wende erreicht ? Münchener Seminare November 2013 Hans-Joachim Dübel Finpolconsult, Berlin

2. A Corner Turned ? Creditor Participation in European Bank Restructurings A Corner Turned ? Münchener Seminare November 2013 Hans-Joachim Dübel Finpolconsult, Berlin

3. Discussion Topics Bank liquidity and capital policies during crisis Creditor participation before and during restructuring/resolution Lessons: policies and institutions for Europe 3

4. Bank Liquidity Policies Temporary Fix for Low Bank Capital Euro crisis High EUR demand by banks hit by real estate and sovereign crisis, pullout of banks / institutions, Note that country ranks have changed over time (Italy!). ECB has been slow on bank capital o It took ECB until 2013 to call for comprehensive recapitalizations ( ‘Asset Quality Review’), o Resistance against bail-in  Draghi letter of July 2013, o Lots of unintended consequences... 4 Source: Osnabrück University. Eurosystem Net Lending / Borrowing of Selected Countries, 2007-2009

5. Bank Liquidity Policies Your Central Bank as an ATM Machine.. 5 Source: Central Bank of Cyprus, Finpolconsult computations. ELA – Emergency Liquidity Assistance, IBU – International Business Unit. Laiki Bank Large ECB ELA borrower, at peak EUR 9.8 billion (33% of assets), Some 50% can be estimated to have used to pay out mostly large deposits (IBU), senior and subordinated bond holders, Asmussen: ‘did not reach 2/3 majority in Gov Council to block Laiki ELA’,  some tightening of ELA rules Oct 2013 ( reporting on borrower/collateral, GC must take decision if >EUR 2 billion ). Discredits Walter Bagehot’s Lender of Last Resort rule? - lend without limits, at penalty rate. Not quite: bank capital !! Laiki Bank Deposits by Source 2010-2012

6. Bank Liquidity Policies Temporary Fix for Low Bank Capital USD crisis (of European banks..) 2008 SIV/ABCP after pullout of US MMFs, Ongoing: European banks’ role in the global China/Petrodollar recycling, The Fed shoots fast !! o After 2008 Steinbrueck interview closed window for HRE/Depfa, o After the 2011 US MMF run on French banks, the Fed passed on credit risk to the ECB (EUR-USD swap agreement). 6 NY Fed Term Auction Facility Lending to German Banks, 2007-2010 Source: Federal Reserve Bank of New York, Finpolconsult computations. Cumulative lending. SIV – Structured Investment Vehicles, ABCP – Asset-backed Commercial Paper, MMF – Money Market Funds.

7. USD Target Debate 7

8. Bank Capital Policies Early Bailout Cases: Hypo Real Estate/Germany 8 Public recapitalization 2009 o Private deposit insurance fund de- facto insolvency by late 2008 (“17 billion in potential losses”), o Early public capital injection and squeeze-out of shareholders by June 2009, o Followed by second recap in same year, o And creation of ‘Bad Bank’ in 2010. Source: Bank reporting, Finpolconsult. HRE Total Liability Structure

9. Bank Capital Policies Bad Bank Is Often Stealth Public Recap Bad banks systematically untie the fate of junior (and senior unsecured) debt from the fate of assets, Transfer pricing chosen may be right or wrong, Why expose government to such risk? 9 Bad Bank Model (Asset Swap Model)

10. Bank Capital Policies FMS Wertmanagement 10 Bad Bank o By October 2010 with entire Greek exposure !! o and with full public ownership, no transfer of junior debt (despite law): = de-fact third – and most expensive - public recapitalization.

11. Bank Capital Policies Classic Bailout: Citibank 11 “Indeed, I was surprised when Tim started reaching out to me directly on the possibility of doing a good bank/bad bank structure for Citi. Initially, he raised the idea of the FDIC setting up and funding a bad bank, without imposing any loss absorption on shareholders and bondholders. I was flabbergasted. Why in the world would FDIC take all of the losses and let Citi’s private stakeholders take all the upside with the good bank? During the second meeting, we discussed a proposal to have the common equity and some of the preferred shareholders help absorb losses. Our view was that all of the private preferred shareholders should convert and that the bondholders should take some losses as well...That was a nonstarter for Tim. He wanted FDIC to take a hit. Sheila Bair ‘Bull by the Horns’

12. Bank Capital Policies Classic Bailout: Hypo Real Estate/Germany 12 Financial result o Shareholders received cash, o Junior creditors were almost entirely protected, except hybrid coupons (EU rules) and UT2 haircuts (held by retail), o Ca 18 billion in public capital investment with perhaps 10% recovery expectation, reduced by almost EUR 4 billion outstanding junior debt. Landesbanken o WestLB last minute junior bail-in did not use the bail-in legislation. Germany estimate: EUR 50 (hybrid) + 60 (sub) billion  investors lost a fraction only, Total loss for taxpayer could be ca EUR 100 billion. Source: bank reporting, Finpolconsult. HRE Junior Debt Structure

13. Bank Capital Policies Basel III Transition In the Middle of Crisis 13 Basel III (Dec 2010) o Fights failure of Basel II to bail in junior debt, o New core capital definition : no revaluation, all capital is permanent, hybrid is limited. o Strongly rising capital demands while second leg of financial crisis unfolds, o Incentivizes buybacks, early calls of Tier 2 and non-eligible Tier 1 instruments. The PROBLEM: o This is all fine for going concern banks.. o.. while it somehow ignores progress in bank restructuring / resolution law (UK, DE), which has widened the bail-in definition. o But what happens, when a bank in difficulties starts “buying” core capital by selling extended capital? Source: FitchRatings (2010)

14. Discussion Topics Bank capital and liquidity policies during crisis Creditor participation before and during restructuring/resolution Lessons: policies and institutions for Europe 14

15. When Gone Concerns Buy Core Tier 1 Liability Management Exercises in Greece Sources: Alpha Bank, Finpolconsult LME deal analysis. Alpha: ‘Rationale for the Offers’, May 2013: Eurogroup announcement of Nov 2012: “ liability management exercises should be conducted in respect of remaining subordinated debt holders so as to ensure a fair burden sharing ” “The Offers are made in order to provide investors with an opportunity to monetise their investments at the relevant Purchase Price on a voluntary basis.” Alpha Bank Liability Structure

16. When Gone Concerns Buy Core Tier 1 Liability Management Exercises in Greece Sources: bank reporting, Finpolconsult LME deal analysis. Is offering between 40 and 60% in cash on subs and hybrids in entirely voluntary LME ‘fair burden sharing’? Greek government recap effectively replaces some 70% of GGB losses (four large banks), Once government is invested in shares, subordinated debt investors can expect to be paid par  moderate acceptance, Financial result : Junior bank debt investors are largely exempt from GGB losses, ca EUR 2 out of 3.5 billion per Q IV 2011 = cash payment. Greek Bank Junior Debt Repurchases Cash offer conditions to investors 2012, 2013 1: National Bank of Greece, 2: Alpha Bank, 3: EFG Eurobank, 4: Piraeus Bank. EFG Eurobank DES – debt equity swap.

17. When Gone Concerns Buy Core Tier 1 Liability Management Exercises in Cyprus Laiki Bank June 2012, first restructuring Cypriot government invests EUR 1.8 billion (>10% GDP) into shares, before hybrids and subs are haircut/converted, Financial result: Hybrid capital investors in are offered a voluntary equity swap, which only by December 2012 becomes mandatory (still far too high number of shares). Subordinated bond investors enjoy a highly concessionary deal (EUR 450 million subordinated bond) Bond offer: 72.5% exchange into senior bond with twice the interest level of deposits (MTM~85), accepted by EUR 132 million, Cash offer: 55% cash offer, accepted by EUR 182 million.

18. Laiki Bank Subordinated Bond LME SLE Capital Loss EUR 150 mln = 1/3 18 Source: Laiki Bank Reporting, Finpolconsult. SLE – Subordinated Liability Exercise.

19. When Gone Concerns Buy Core Tier 1 Liability Management Exercises in Cyprus Laiki Bank (other) In September 2012, EUR 330 million in senior bonds matured days before the PIMCO due diligence exercise started, The June 2012 subordinated bond deal was partially ‘clawed back’ in March 2013 through the mandatory debt-equity swap of senior unsecured. Bank of Cyprus In May 2011 called a subordinated bond of EUR 200 million at par, at the first possible date, Note: Deutsche Bank had broken with this policy in 2008.

20. When Gone Concerns Buy Core Tier 1 Creditor Rotation Through LME Sources: bank reporting, Finpolconsult computations. Bankia Funding Structure Bankia Large volumes of senior unsecured and covered bonds sold to foreign investors, these investors rotated with the ECB, 2006/7 subordinated debt issued mainly to professional investors.

21. When Gone Concerns Buy Core Tier 1 Creditor Rotation Through LME Sources: bank reporting, Finpolconsult computations. Bankia Funding Structure, Expanded Bankia Since 2010, covered bonds were issued into ECB repo (i.e. disappear from balance sheet and ECB claims appear), Foreign bank investors recycle proceeds into ECB surplus reserves.

22. When Gone Concerns Buy Core Tier 1 Creditor Rotation Through LME Sources: bank reporting, Finpolconsult computations. Bankia Cash payments to (largely professional) subordinated investors in 2010 and 2011 of ca. EUR 2 billion, In parallel new subordinated debt and equity was issued, at this time mostly to retail investors/households, 2012 cash paid had to be reinvested in equity (LME). Bankia Cash Flow

23. Share Buyback as a Costly Policy to Support Equity LME Sources: http://ftalphaville.ft.com/2012/06/26/1056401/the-spanish-bank-buy-back-riddle/http://ftalphaville.ft.com/2012/06/26/1056401/the-spanish-bank-buy-back-riddle/ Onvista.de. Stabilizing LTRO-effect in Q I 2012 for share prices was used by Spanish banks for share buybacks. Bankia: -Offered hybrid investors shares at high exercise prices into collapsing share price trajectory, benefiting insiders who sold immediately. -Bank may have lost EUR three-digit millions in cash (Core Tier 1) -Reports for total year 2012 75 million loss from own share dealing operations.

24. When Gone Concerns Buy Core Tier 1 Creditor Rotation Through LME Sources: bank reporting, Finpolconsult computations. SNS Reaal 2008 public (senior) hybrid capital injection. In 2009 the profit situation was still seen as healthy enough for SNS to buy back EUR 250 million of the hybrids, most of which from government, and retire other junior debt, In 2010, SNS Bank placed a EUR 500 million subordinated bond with a 10-year maturity, 2011 large cash LME over EUR 420 million of old subordinated debt, 2012 first possible call of 2003 issued hybrid capital securities exercised. SNS Reaal Cash Flow

25. Restructuring – Debt Equity Swap or Haircuts? Source: Finpolconsult SLE – Subordinated Liability Exercise Coco – Contingent Convertibles CDS – Credit Default Swaps. Spain (Bankia) - Haircut & DES Group 1 – mandatory SLE, Group 2 – first voluntary, then mandatory SLE. - Pricing approach: “market price” (first law draft permitted 10% over) vs. liquidation value, Ultimately ‘negotiated’. Netherlands (SNS Reaal) – Expropriation 100% haircut, liquidation value. Both approaches lead to same desired Core Tier 1 effect. Bankia: Haircut & Debt Equity Swap Combined Debt equity swap leaves possible economic upside on the table for investors, tied to asset performance. Parallel to Coco debate: 0-1 (insurance) instruments carry significant legal risk, esp. if triggers are regulatory. Alternatives? CDS written by bond investors on initial bank portfolio, Example: KfW credit-linked notes program.

26. Restructuring – Comprehensive Debt Equity Swap Bank of Cyprus = extension of Spanish junior debt bail-in to senior unsecured - No initial haircuts of debt !!  swap. -Allocation of sub, senior unsec, large deposits to thin equity classes, - Full voting rights for these classes, - High interest rates for preferred shares if bank performs, -Current main shareholder is Laiki Bank unwinding vehicle (18%), - Issue: ad-hoc seniority given to Cyprus public sector and ECB. 26 Bank of Cyprus – Debt Equity Swap Source: Finpolconsult.

27. Restructuring – Good Bank Model (FDIC Standard) Cases Laiki Bank, Good Bank & P&A combined with super-seniority of insured deposits. o Denmark (Amagerbanken), Greece (ATE, Hellenic Postbank). Problem as with all P&A is determination of sales price during stress. The larger the bank, the greater valuation risk (e.g. U.S. Washington Mutual). Response: Iceland - bridge banks to be sold later. Would have required public funding in the case of Laiki. Did Greek subsidiary P&A to Piraeus cut losses or profit? 27 Laiki - Good Bank, Purchase and Assumption (P&A) Source: Finpolconsult. *

28. Fiscal Expenditure Private Sector Involvement, 3 Countries 28 Source: Central banks, Analistas Financieras Institucionales, Finpolconsult. Gap based on target CT1, ‘internal includes proceeds from sales, accounting issues (DTA)’ *pre-Sareb, **without impact of Good Bank-Bad Bank splits on non-core banks, ***assumes some OSI in debt-equity swaps. 2012/13 summary: three countries = three approaches. Complete individual path- dependency, only MoU changed was Cyprus, Greek OSI ratios >> Spain, Cyprus outlier (but not for smaller banks). Greek Banking Program**Spanish Banking Program, Group 1* Cyprus Banking Program***

29. Fiscal Expenditure Expanded Country Sample 29 Source: Finpolconsult. Creditor participation ratios increasing over time, however with outliers : Amagerbanken an early case (fall 2010) with senior unsecured creditor participation, o Note: In fall 2010, Ireland was not permitted to bail-in senior debt at Anglo Irish Bank, while Denmark did so, Dexia a late case (second recap 2012), even after Spain, with large junior bondholder bailout, junior debt investors will receive in total some EUR 2 billion. Estimated Private Sector Involvement in Capital Gap Financing

30. Fiscal Expenditure Expanded Country Sample 30 Source: Finpolconsult. Deterrents to bail-in : First government recap often followed by second rather than bail-in – > avoids stigma of wasting taxpayer money (HRE, Dexia), Fiscal capacity, program vs. non-program country (currently MPS Italy), Restructuring delay, may hit the wrong creditors  Bankia, but did not deter SNS Reaal. Dexia S.A. Junior Debt Structure

31. Fiscal Expenditure vs. Cost Loss Expectation of Recaps 31 Source: Bank reporting, national central banks, Finpolconsult. Question: - What share value does government acquire with a given expenditure? -Chart assumes 100% price/book ratio of acquired bank stock, i.e. full capital gap has been detected and adequately provisioned for. Result: 7 sample bank aggregate fiscal loss would have eaten up 30%-35% of ESM capital. Fiscal Loss Based on 2012 Book Value

32. Fiscal Expenditure vs. Cost Loss Expectation, Extended Country Sample 32 Source: Bank reporting, national central banks, Finpolconsult. Additional result: SNS Reaal with small fiscal profit, based on (possibly heroic) assumption of no additional losses, Deep bail-in cases (Laiki, Amagerbanken) show high loss for government as government ‘gives up’ initial recapitalization, Median expected loss ratio for government is 75%. Fiscal Loss Based on 2012 CT1 Book Value

33. Fiscal Cost of Bank Resolution Methodology Example, Spain 33 Source: FROB, Bank of Spain, Autonomous Research, Finpolconsult assumptions and computations. DES – Debt Equity Swap Recovery expectation matters !! ECB repo is both super-senior and collateralized, ECB ELA at least super-senior (Cyprus), Hybrid capital safer than shares, Share injection after bail-in safer than before bail-in, DESs share economic value with investors, haircuts don’t. Spain -Historic expenditure inflated by large guarantees. -Very large ECB/ELA exposure, e.g. Bankia alone EUR 74.5 bln, -Potential bad bank (Sareb) fiscal cost are not properly accounted for (guarantees). Spain, Banking Program Accounting under Author’s Subjective Expected Loss Assumptions

34. Discussion Topics Bank capital and liquidity policies during crisis Creditor participation before and during restructuring/resolution Lessons: policies and institutions for Europe 34

35. Bank Restructuring & Resolution Law What did take Europe so long? In the U.S, the 1991 FDIC Improvement Act marked the turning point – a full 9 years after Garn St Germain: Abolition of ‘Open Bank Assistance (‘direct recapitalization’), Least Cost Resolution Approach (from taxpayer perspective), ALL deposits became super- senior (separate act), FDIC became only required to transfer insured deposits in a P&A  bail-in of large deposits. Europe: IKB to Cyprus is 5.5 years.. (but there was a template!) 35 FDIC Least Cost Resolution Game Changer 1991

36. Lessons Learned Resolution and Restructuring A European Directive by 2018 is too late ! o 10 years after Lehman – exempts all likely current crisis cases, o EU KOM rule (Aug 1, 2013) disallowing state aid without prior junior bond bail-in less effective without the Directive (MPS case). Exceptions should be avoided o At least junior bond bail-in, o Include Covered Bonds overcollateralization, o ‘Systemic risk’ backdoor should be closed  main risk is fiscal collapse, o ECB ‘ precautionary recapitalizations ’? Complete Banking Union architecture o Build the European version of U.S. FDIC, to create vested interest in shorter time to restructuring and deeper creditor participation, o ECB has conflicts of interest as de-facto bank investor and cannot be a monopoly bank regulator from a fiscal perspective  shared supervision. 36

37. 11/05/201537 Internal Market Single Resolution Fund National Resolution Authorities manages puts under resolution / sets resolution framework contribute provides funding instructs resolve All banks Failed bank supervise notify ECB National Supervisors COM Single Resolution Board EU ‘FDIC’ Concept (Source DG Markt)

38. Lessons Learned Junior Bond Capital Availability 38 Reconcile Basel and fiscal definitions of bank capital !!

39. Challenges Ahead 39 Irish Mortgage Portfolio Funding, Stylized Source: above - Osnabrück University, below - Finpolconsult. Securities portfolios o Self-inflicted ‘sovereign-bank’ doom-loop by banks rejecting diversification, o ECB has not enforced diversification, LTRO without conditionality, o Whichever-is-lower ratings are accepted - usually Canadian DBRS, o Arbitrary repo policies (e.g. Cyprus sov). Loan portfolios o ‘Kick-the-can’ corporate loans (EBIT<interest; 40-50% in ES, IT, PT), o Insufficient mortgage profitability related to indexed loans (Euribor, ECB refi), o Underperformance and market risk are NOT generally covered by Asset Quality Review. Slow bank recapitalization due to rejection of bail-in means that Eurosystem imbalances could remain high.

40. USD Target Debate 40

41. Slides in Reserve 41

42. Discussion Topics Historic bank capital and central bank liquidity policies o Central bank liquidity as a substitute for bank capital, o Initial bank recapitalizations & Basel III reaction. Creditor participation in European bank restructurings o Before restructuring/resolution: voluntary liability management, o During restructuring/resolution: haircuts vs. debt equity swaps, Good Bank vs. Bad Bank resolution approach, o Government view: fiscal expenditure vs. fiscal cost. Lessons learned: banking sector policies and institutions o Policies: sequencing of recaps, rank of government in recaps, options for tying fate of historic assets to historic liabilities, bank bond guarantees, o Institutions: Supervisory and Resolution Mechanisms, deposit insurance. 42

43. Bank Capital Policies Basel II Definitions Proved Obsolete 43 Basel II diagnosis o There were EUR 1 trillion in subordinated debt and hybrid capital per 2007 for 100 of the world’s largest banks (Fitch). o Most of which was not used for recapitalization. Reasons o Weak instrument setup create legal ambiguity (contractual, revaluation and termination clauses, tied to GAAP), o Connected insiders demand Open Bank Assistance (“Unser /nuestro /μας Lehman”), o Insufficient options for bank supervisors to intervene outside the insolvency scenario. Source: FitchRatings (2010)

44. Bank Capital Policies Basel III Transition - A Problem During Crisis 44 Basel III (Dec 2010) o New Core Tier 1 definition : no revaluation, all capital is permanent, hybrid is limited. o Strongly rising CT 1 demands while second leg of financial crisis unfolds. o Incentivizes buybacks, early calls of Tier 2 and non-eligible Tier 1 instru- ments. The PROBLEM: o This is all fine for going concern banks.. o.. while it somehow ignores progress in bank restructuring / resolution law (UK, DE), which has widened the bail-in definition. o But what happens, when a bank in difficulties starts “buying” core capital by selling extended capital?

45. Greek Bank Subordinated Debt Pull to Par ! 45 Source: Onvista.de. *Nationalized EFT Eurobank is the only core bank that has offered a debt equity swap per May 2013 to sub bond investors, but it was voluntary. Other banks offered cash LME (Alpha 55c, Piraeus 55c). Alpha, Piraeus and NBG also offered cash to hybrid capital investors (35-40c). EFG Eurobank* Alpha Bank Piraeus Bank Expected gross 600 million CT1 effect (disregarding the opportunity cost) from voluntary LME looks ambitious. Piraeus May sub LME had 10% acceptance. Compare to Anglo Irish 2010 sub LME : investors demanded 35-40%, government offered 20%, demanding exit consent to punish holdouts. Comparable legal situation (offshore trust SPV-Issued) “Liquidity management exercises” (Nov 12 CB report)

46. When Gone Concerns Buy Core Tier 1 Liability Management Exercises in Spain Sources: Banco de Valencia, Bankia reporting, Finpolconsult computations. Note: senior bonds include covered bonds, which however largely disappear from reporting as a technical effect from large scale repo with the ECB. Voluntary LMEs became a popular sport in Spain Late 2011 to June 2012 all banks got involved. Great diversity of offers. Private banks partly aggressive (Santander with lg. haircuts) partly concessionary (Popular). Some Cajas rather aggressive (Bankia voluntary DES, overleaf). Yet, many Cajas offered or tried to offer concessionary deals to investors Liberbank with EUR 1.2 billion capital gap in October offered in June 2012 to swap EUR 627 mln hybrids (‘preferentes’) into 5 year deposits without haircut. Late sweet Caja LME offers can be seen as having shifted mood in banking program sponsor countries in direction of SLE. Liberbank LMEs were intercepted by the Spain MoU. Bankia Liability StructureBanco de Valencia Liability Structure

47. When Gone Concerns Buy Core Tier 1 Liability Management Exercises in Slovenia Large LMEs of the two largest Slovenian banks during 2012 let to severe loss of bail-inable capital NLB: 2012 gain in CT1 of EUR 110 million. EUR 430 million reduction of subordinated debt. Also reduction of senior unsecured debt by EUR 1,130 million. Maribor: 2012 loss of EUR 90 million in subordinated debt. Insufficient subordinated and senior unsecured debt, ex deposits, by restructuring date to support Good Bank approach.  Government-sponsored bad bank. Novo Ljubljanska Banka Liability Structure Source: Bank Reporting, Finpolconsult. Novo Ljubljanska Banka Cash Flow

48. Bank Capital Policies Bad Bank Is Often Stealth Public Recap Problems: Systematically unties the fate of junior (and senior unsecured) debt from the fate of assets. Requires new equity provider for the bad bank ( government, incl implicit – e.g. Sareb). Demands definition of a transfer prices today, despite great uncertainty  risk of subsidies Germany Landesbanken: 90% transfer price with gaps to be amortized by the selling bank over 20 years. Ireland/Spain: far lower transfer prices, against full protection of the bank. Low asset price growth environment creates risk, there are major differences between the 2010s and 1990s (Sweden success story). 48 Bad Bank Model (Asset Swap Model)

49. Fiscal Expenditure vs. Cost Loss Expectation, Extended Country Sample 49 Source: Bank reporting, national central banks, Finpolconsult. Same methodology as before Additional result: SNS Reaal with small fiscal profit, based on (possibly heroic) assumption of no additional losses, Deep bail-in cases (Laiki, Amagerbanken) show high loss for government as government ‘gives up’ initial recapitalization in order to claw back junior and partly senior debt, Shallow bail-in cases (HRE, Dexia) vice versa, Median expected loss ratio for government is 75%. Deep bail-in legislation may cure impact of long restructuring delay, but at high costs SNS Reaal, Laiki, Bankia have hit new junior or senior creditor generations. Fiscal Loss Based on 2012 CT1 Book ValueCash Drain to Junior Debt Investors and Policy Score (Stylized)

50. Holding on or Selling Now? Greek Warrant Policy for Core Banks 50 Source: Bank Reporting, national central banks, Finpolconsult. Target CT 1 includes pre-provision income estimate for 2012 – 2014/15. Greek banks have optimistic PPI estimates compared to Spain, prompting government to price warrants above stressed book values. E.g. Alphabank at target CT 1 level + 10% (44c) = 150% 2012 book (30c). Is forgiving a potentially large upside optimal fiscal policy ? Greek government locks in historic recapitalization losses, but cuts additional loss expectation, GGB haircuts essentially treated as if they were an external shock to banks. Alpha Bank HFSF Warrant PricingFiscal Loss Based on Target CT1 Book Value

51. Fiscal Cost of Bank Resolution Quasi PSI-Reversal in Greece, Cyprus MoUs 51 Source: Bank and press reporting, Finpolconsult assumptions and computations. Note: very limited data officially provided by HFSF. Without bond guarantees. Greek program has produced both far higher cost / GDP and even higher cost / expenditure than Spain. Cost expenditure inflated by clear political intention to shield system from GGB PSI losses. Calculatory PSI reimbursement between 60% (NBG) and 100% (rest, based on 2012 CT1) Result will likely be OSI. Cyprus MoU II (Mar 13) saved ca half of the fiscal cost over MoU I (Nov 12) EUR 5 bln under MoU II (mostly smaller banks & EUR 1.8 bln Laiki) compared to EUR 10 bln under MoU I. Subject to additional evaluation of large Cypriotic banks Q II 2013. Greece, Banking Program Accounting under Author’s Subjective Expected Loss Assumptions

52. Some Counterfactuals Early Intervention is Key to Protect Government and Depositors 52 Source: Bank reporting. *7 caja reporting for 2010, Bankia senior bonds minus 50% assumed covered bonds, **assumed to be the average uninsured deposit ratio outside Cyprus. CT 1 needs include early gov recaps. Finpolconsult. Assume a 10% deposit bail-in**/ zero fiscal bail-out rule. Then the break-even restructuring dates were: Bankia – Q IV 2010 … Anglo Irish bank was nationalized in Jan 2009 !! Banco de Valencia – Q IV 2010 EFG Eurobank – Q IV 2011 Alphabank – Q I 2012 Laiki Bank – 10% deposit threshold always missed, but the uninsured deposit ratio was 50%.. Bank of Cyprus – 10% deposit threshold always missed, but…. Piraeus – only case in a 7 bank list where the deposit insurer would have had to disburse in a super-senior rank scenario. Bankia/BFA* Alpha Bank Laiki Bank

53. Lessons Learned Bank Resolution Policies Establish clear limits for government open bank assistance o ‘Direct recapitalization’ so far means transfers to bank investors. Invest in higher ranks only (e.g. senior hybrid capital) to avoid deep subordination. Ex: U.S. in Fannie Mae. o Generally avoid early government recapitalizations before at least junior debt has been bailed in. Early public recaps raise the threshold for future private investor bail-in, if losses widen. Minimize delay to restructuring o Acting late may cost government and in the extreme case depositors dearly (Cyprus), o Acting late hits new junior/senior debt investor generations, and bails out old ones. o The macroeconomic benefits of delay (‘soft landing’) are dubious. 53

54. Lessons Learned Bank Resolution Policies Increase depth of bail-in while stabilizing investor relations o Promote explicitly layered and granular liability side (by law or contractual conditions), stabilizing ex-ante expectations, o Do not expropriate investors during bail-in Limit haircuts to demonstrable losses (always observing that shares are haircut first), Preferable bail-in approaches to match uncertainty are debt equity swaps into separate share classes for bailed in debt, or credit default swaps referencing the original portfolio, Observe initial ranks. E.g. swap senior bonds/uninsured deposits into dated junior bonds, junior bonds into hybrid capital, hybrid capital into shares, Depending on the situation after bail-in, permit shareholders to regain full control, e.g. through warrants or reconversions or prepayments of bailed-in capital, o If significant parts of the bank are to be unwound making bail-in infeasible, prefer the Good Bank over the Bad Bank approach, o Avoid ad-hoc rank changes in the neighborhood of any mandatory liability management, even if government is hit. 54

55. Lessons Learned General Bank Capital / Liquidity Policies General bank capital policy Reconcile Basel III and Eurozone-fiscal definitions of bank ‘capital’ (8% RWA?), Revoking excessive focus on Core Tier 1 for gone concern banks is sine-qua non (Slovenia!!), rules for voluntary LME policies. Bank liquidity policy European Central Bank to curb lending to zombie banks, enforce restructuring. Discussion over Target II can be largely retired if lending is to banks with sufficient capital levels (but see portfolio issues). 55

56. Banking Union Institutions Banking Union architecture Build the Second Pillar now (European version of U.S. FDIC), to create vested interest in shorter time to restructuring and deeper creditor participation, ECB has conflicts of interest as de-facto bank investor and cannot be a monopoly bank regulator from a fiscal perspective  shared supervision, Three-party division of labor : competition/state aid policy (KOM), liquidity policy (ECB), resolution/deposit insurance (TBD). European resolution authority detail Initial resolution authority (= de-facto deposit insurer) permitted to XX years learning curve, Is (ex officio) bank supervisor from the start (existing models, e.g. Bafin / Bundesbank), Handles pan-European bank-sponsored funds, on top of national funds, Possible ESM backup under strict conditions (e.g. credit lines as U.S. Treasury to FDIC), In the long-term, single European fund, Integration of regional banks is paramount (given crisis causes) ! 56