## _spn0912 — Executive Summary

## Source details

**Canonical URL:** [_spn0912 — Executive Summary](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/spn/2009/_spn0912.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/spn/2009/_spn0912.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/spn/2009/_spn0912.pdf.json)

---

### Framework and objectives
- Government objective: lower the probability of a bank’s default while minimizing taxpayer burden.
- Analysis starts from a Modigliani-Miller frictionless benchmark and introduces frictions: inability to renegotiate debt, costs of financial distress, asymmetric information, and managerial incentive problems.
- Key conflict: when debt contracts cannot be renegotiated, restructuring raises debt value and lowers equity value, so shareholders oppose restructuring unless compensated.

### Frictionless benchmark and implication for restructuring
- Debt-for-equity swap mechanism:
  - Partial conversion of debt D into lower-face-value debt D' = A* plus equity can lower default probability to target p* = F(A*) while leaving debt and equity holders indifferent.
  - In the benchmark, no taxpayer transfers are needed: V(A) = V(E) + V(D); total firm value unchanged.
- When debt cannot be renegotiated:
  - Any restructuring that reduces default probability increases market value of debt V'(D) relative to V(D) and reduces equity value.
  - Minimum government transfer required to achieve p = p* equals T = V'(D) – V(D).
  - Transfer requirements vary by scheme and by whether private surplus C exists to offset T.

### Comparative evaluation of restructuring instruments (findings)
- Optimal (state-contingent) subsidy:
  - Government transfers only ex post when A1 is between A* and D: transfer = D – A1 in that region; no transfer when A1 < A* or A1 > D.
  - Minimizes expected taxpayer cost for achieving p*.
- Recapitalization with common equity:
  - Required cash to achieve p*: Cash = D – A*.
  - Government transfer needed to leave initial equity unchanged: T = V'(D) – V(D).
  - Debt recovery for default realizations A1 < A* becomes D – A* + A1 after recapitalization.
- Recapitalization with preferred stock or convertibles:
  - When convertibles/preferred do not trigger default, impact on preexisting debt value and required transfer equals that of common-equity recapitalization.
  - If convertibles are pari passu and conversion is automatic or at holder’s discretion, recovery of preexisting debt equals the equity-issue case.
- Subsidized debt buybacks:
  - Bank issues equity and uses proceeds to buy back fraction α of debt so that (1 – α) D = A*.
  - Less costly to taxpayers than recapitalization that keeps proceeds as cash; transfer still equals V'(D) – V(D), but implied recovery schedule is closer to first-best.
- Simple (capped) asset guarantees:
  - Government insures assets falling below D up to cap D – A*; guarantees reduce default probability to p* but pay transfers even when A1 < A* (more costly than the optimal state-contingent scheme).
  - Cost to taxpayers equals that of a subsidized recapitalization with identical implied debt recovery.
- Caballero’s scheme:
  - Government gives new equity investors a put (buyback at fixed future price); equivalent in transfer cost to subsidized equity recapitalization but requires no up-front transfer.
- Above-market-price asset sales:
  - Government buys fraction a at price (1 + m) a V(A).
  - More costly for taxpayers than recapitalization or guarantees because recovery for debt holders is higher across realizations; recovery slope in the default zone becomes (1 – a) < 1.
- Sachs proposal (asset sales at book value with recourse via warrants):
  - Lowers default probability and raises debt recovery, but initial equity holders are worse off (dilution or losses ex post); equity holders would oppose.
- General ranking intuition:
  - Asset sales (above-market) are most costly to taxpayers.
  - Guarantees or recapitalizations that target the payoff region near solvency are more efficient.
  - Debt buybacks that change liability structure are closest to first-best (debt-for-equity swap) and reduce transfer magnitude.

### Private surplus, social benefit, and intervention thresholds
- Restructuring can generate private surplus C by reducing costs of financial distress (e.g., restoring interbank confidence, enabling projects).
  - Literature estimate for costs of financial distress for typical nonfinancial firms: "about 10 percent to 23 percent of ex post firm value."
- Social benefit B from preventing systemic failure sets upper limit of government willingness to pay.
- Three decision cases based on total surplus S = C + B:
  - If C > V'(D) – V(D): private surplus suffices; government transfer not needed.
  - If intervention is needed and B > T: government willing to pay T; minimum government cost = V'(D) – V(D) – C.
  - If B < V'(D) – V(D) – C: aggregate surplus positive but insufficient to leave debt contract unchanged; renegotiation of debt contracts needed.

### Incentive, governance, and moral-hazard considerations
- Separating bad assets (good bank / bad bank) can increase managerial productivity by focusing talent and span of control.
- Governments often lack operational expertise; recommended use of private-sector expertise to run asset management funds or nationalized banks; consider auctions for management contracts.
- Moral hazard and timing of transfers:
  - Up-front cash injections risk "free cash flow" misuse (looting, excessive bonuses, dividends).
  - Ex post instruments (guarantees) reduce ex ante looting risk but may weaken managers’ incentives to maximize A1.
  - Optimal design balances ex ante looting risk and ex post incentive provision; contingent transfers and minimal up-front transfers recommended.
- Convertibles:
  - Automatic or holder-discretion conversion can serve as automatic future recapitalization and increase surplus.
- Long-run accountability:
  - Insufficient punishment for managers and shareholders combined with large taxpayer transfers increases moral hazard for future crises.

### Limits of market pricing and arbitrage gains
- Market prices V(A) may be below fundamental value J(A) due to limits of arbitrage and funding liquidity constraints.
- Government can capture arbitrage gains V_GOV(A) – V(A) by buying undervalued toxic assets above market price but below fundamental value; gains larger when government takes equity-like claims.
- Arbitrage gains can be used to incentivize private managers if private investors participate in asset purchases.

### Participation and asymmetric information
- Voluntary participation signals adverse selection and raises required subsidies.
- Two-type example (high-quality and low-quality banks):
  - Both types can have default rates pL > pH > p*; voluntary recapitalization at market prices forces high-quality banks to require high subsidies; low-quality banks obtain informational rents.
  - Outcome: high subsidy needed to induce participation of high-quality banks; inefficiency arises from oversubsidizing low-quality banks.
- Role of hybrid securities:
  - Hybrids (convertible notes, preferred shares) are less information-sensitive than equity and can reduce taxpayer subsidies needed for voluntary participation.
  - Convertible notes trade off conversion ratio (cash raised and default reduction) against increased information sensitivity.
- Remedies to participation problems:
  - Use hybrid instruments and well-designed asset guarantees to reduce signaling costs and required transfers.
  - Compulsory programs, when feasible, eliminate signaling concerns and associated subsidy premia.
  - Rigorous bank examination and public disclosure of asset-quality information mitigate asymmetric-information problems.

### Design rules under asymmetric information and adverse selection
- Asset guarantees or state-contingent transfers that do not issue new securities mitigate signaling costs because no return is asked from banks (no adverse signal).
- Government should compute A* using the payoff distribution of the low-quality bank:
  - Overinsures high-quality banks (they attain default probability below p*) but avoids overtransferring to low-quality banks and informational rents under recapitalization.
- Asset sales and the "lemons" problem:
  - Managers will tend to sell lower-quality assets; government pricing should reflect anticipated quality (auction mechanism recommended).
  - Government purchase price affects regulatory book values; marking all banks to a low purchase price can force regulatory insolvency and justify subsidies for stabilizing prices despite lack of direct economic motive.
- Use of government information:
  - Bank examinations and stress tests with disclosed results reduce restructuring costs by informing investors.
  - Guarantees without caps are persuasive signals but put taxpayers at risk; Caballero’s insurance-on-stock-price scheme can credibly signal commitment.

### Political constraints and practical considerations
- Opportunity cost of government resources and fiscal rules may favor mechanisms with only ex post transfers (asset insurance, Caballero’s scheme) if credible commitment to honor transfers exists.
- Political influence on management risks mismanagement under long-term government ownership; hybrids without voting rights reduce this risk.
- Even without formal voting rights, governments may influence management when they are large owners.

### If bankruptcy is inevitable
- If realized asset value A1 < A*, optimal response is to let the bank go bankrupt (A* set accordingly).
- Under heavy liquidity pressure, temporary nationalization often inevitable:
  - Government holds large common equity share, controls management, acquires information, and can ask stakeholders to share burden.
  - Temporary nationalization can limit taxpayer burden, expedite resolution, and facilitate resale to private investors.
- Scope of honored debt during temporary nationalization:
  - At minimum, transaction-purpose instruments should be honored (e.g., interbank borrowings).
  - Less justification exists for honoring long-term debt.
  - Further discussion required on instruments such as guarantees on securities backed by credit card debt and accounts payable.

### Case studies (selected facts and mechanics)
- Switzerland (UBS good bank / bad bank split, fall 2008):
  - Up to $60 billion in toxic assets could be removed from UBS (total assets almost $2 trillion).
  - UBS provided 10 percent of asset value (up to $6 billion) for bad-bank equity but transferred equity to the Swiss government for $1, covering first 10 percent of losses.
  - Swiss National Bank lent the StabFund up to $54 billion at LIBOR plus 250 basis points.
  - UBS ultimately transferred $39 billion of assets to the StabFund.
  - StabFund loan carried 12.5 percent interest.
  - StabFund was immediately valued at $1.
  - UBS transferred about $4 billion (not the $6 billion maximum), allowing UBS to retain $2 billion worth of capital net.
  - Repurchase option: once loan fully repaid, UBS can repurchase fund equity by paying the Swiss National Bank $1 billion plus 50 percent of the equity value in excess of $1 billion.
  - UBS completed a voluntary rights-issue recapitalization in April 2008; old management exited at the annual general meeting.
- United Kingdom (RBS and Lloyds-HBOS):
  - Recapitalization: government owned 58 percent (₤20 billion) of RBS and about 44 percent (₤17 billion) of Lloyds-HBOS after injections.
  - Asset Protection Scheme (January 2009):
    - ₤325 billon (14.5 percent) of end-2008 assets for RBS.
    - ₤260 billon (23.4 percent) of end-2008 assets for Lloyds-HBOS.
    - Government would compensate losses up to 90 percent if valuation fell below tailored threshold.
    - First-loss thresholds: ₤19.5 billon (6 percent) for RBS; ₤25 billion (9.6 percent) for Lloyds-HBOS.
  - Insurance fees paid in preferred shares; government economic stake could approach 50 percent for Lloyds-HBOS and more than 80 percent for RBS after fees and injections (voting rights about 70 percent if converted).
  - Corporate governance: government obtained rights to appoint independent nonexecutive directors and to limit executive compensation and dividend payouts.
- United States (Geithner Financial Stability Plan, as of May 2009):
  - Capital Purchase Program (TARP) allowed banks to receive cash by selling preferred shares to Treasury; Citibank and Bank of America received additional targeted support (Targeted Investment Program and Asset Guarantee Program).
  - February 2009 Financial Stability Plan components:
    - Supervisory Capital Assessment Program: compulsory stress test for 19 largest banks to evaluate extra capital needs.
    - Capital Assistance Program: banks required to raise extra capital; government option to issue preferred shares that automatically convert after seven years (earlier conversion possible with issuer discretion and regulator approval).
    - Public-Private Investment Program: funds to purchase legacy loans and securities; prices determined by competitive bidding.
      - Each fund: 50 percent equity participation from Treasury without voting rights.
      - Loan-purchasing funds: can issue FDIC-guaranteed debt with leverage up to 6; FDIC supervises funds.
      - Securities-purchasing funds: Treasury lends to each fund at up to a 2-to-1 leverage ratio; loans nonrecourse.
  - Liquidity measures: FDIC Temporary Liquidity Guarantee Program; Federal Reserve TALF lending up to $200 billion on certain AAA-rated ABS.
  - Corporate governance constraints: restrictions on executive compensation, dividends, stock repurchases, acquisitions; Treasury to make contracts public.
  - Evaluation concerns: moral hazard from government-sponsored inexpensive leverage; calibration challenge in setting subsidies to secure participation of high-quality managers; proposal to auction management and cash-flow rights.

### Overarching policy recommendations and institutional reforms
- Combine asset-side and liability-side interventions case-by-case to address multiple trade-offs and frictions.
- Aim to limit taxpayer transfers by:
  - Avoiding unnecessary subsidies to debt holders.
  - Maximizing economic value created by restructuring (capture private and arbitrage gains where feasible).
  - Favoring liability-structure changes (debt buybacks, convertibles, preferred shares) when they move recovery closer to first-best.
- Design contingent transfers to align manager incentives with maximizing future profits; minimize up-front transfers to prevent misuse.
- Ensure transparency of taxpayer costs and final beneficiaries of subsidies.
- Hold managers and shareholders accountable and impose punitive consequences where appropriate to reduce future moral hazard.
- Institutional reforms to reduce frictions over time:
  - Better legal framework for rapid debt renegotiation with limited systemic risk.
  - More timely and in-depth disclosure requirements for bank asset information and counterparty exposures.
  - Regulation to encourage conversion clauses in long-term debt so debt automatically converts into equity in distress.

*Source: EXECUTIVE SUMMARY and Section II E) of _spn0912 (IMF staff note).*

### Executive Summary ......................................................................................................

### Executive Summary

### I. Introduction
- Presents the scope and motivation for the analysis. (See full section: I. Introduction, p.4)

### II. A Benchmark Frictionless Framework
- A. Setup (p.6)
- B. First Best—Voluntary Debt Restructuring (p.7)
- Includes figures:
  - 1a. Assets and Liabilities of the Bank
  - 1b. Cumulative Distribution Function of Ex Post Asset Value
  - 1c. Sharing Rule

### III. Restructuring with No Debt Renegotiation
- A. Difficulty of Voluntary Restructuring (p.9)
- B. Government Subsidy and Debt Recovery (p.10)
- C. State-Contingent Insurance: Optimal Subsidy (p.10)
- D. Recapitalization with Common Equity (p.11)
- E. Recapitalization by Issuing Preferred Stock or Convertible Debts (p.12)
- F. Subsidized Debt Buybacks (p.14)
- G. Simple Asset Guarantees (p.15)
- H. Caballero’s scheme (p.16)
- I. Above-Market-Price Asset Sales (p.16)
- J. The Sachs Proposal (p.18)
- K. Combining Several Schemes (p.18)
- Figures related to restructuring:
  - 2. Debt-for-Equity Swap
  - 3. Restructuring and Debt Recovery
  - 4a. Transfer of the Optimal Subsidy
  - 4b. Recovery Rate
  - 5. Recapitalization
  - 6a. Same Seniority Convertible
  - 6b. Recapitalization with Hybrid Securities
  - 7. Debt Buyback
  - 8. Transfer under Capped Asset Guarantee
  - 9a. Assets and Liabilities after Asset Sales of a Fraction a
  - 9b. Debt Recovery after Asset Sales of a Fraction a

### IV. Private and Social Surplus from Restructuring
- A. Key Concepts (p.19)
- B. Endogenous Surplus and Restructuring Design (p.20)

### V. Participation Issues under Asymmetric Information
- A. Recapitalization with Asymmetric Information on Across-Bank Asset Quality (p.23)
- B. Asset Sales with within-Bank Adverse Selection (Lemons Problem) (p.26)
- C. Use of Government Information (p.27)
- Figure:
  - 10. Convertible Note

### VI. Other Considerations
- A. Political Constraints (p.27)
- B. If Bankruptcy Is Inevitable (p.28)

### VII. Case Studies
- A. Switzerland: Good Bank/Bad Bank Split in the Case of UBS (p.29)
  - Figure 11. UBS Restructuring (Announced Plan)
- B. United Kingdom: Recapitalization and Asset Guarantee for RBS and Lloyds-HBOS (p.31)
- C. United States: The Geithner Plan as of May 2009 (p.32)

### VIII. Conclusion
- Summarizes findings and implications (p.35)

Key supporting material
- References (p.38)
- Table:
  - 1. Pros and Cons of Various Policy Options (p.37)

*Source: Executive Summary (document table of contents and figure/table list) from the provided PDF content.*

### EXECUTIVE SUMMARY

### _spn0912 - EXECUTIVE SUMMARY

### Framework and objectives
- Assumes government objective: lower the probability of a bank’s default while minimizing taxpayer burden.
- Analysis starts from a Modigliani-Miller frictionless benchmark and then introduces frictions: inability to renegotiate debt, costs of financial distress, asymmetric information, and managerial incentive problems.
- Key conflict highlighted: when debt contracts cannot be renegotiated, restructuring raises debt value and lowers equity value, so shareholders oppose restructuring unless compensated.

### Frictionless benchmark (Modigliani-Miller)
- Debt-for-equity swap (partial conversion of debt D into lower-face-value debt D' = A* plus equity) can lower default probability to target p* = F(A*) while leaving both debt and equity holders indifferent.
- In this benchmark, no taxpayer transfers are needed: V(A) = V(E) + V(D) and total firm value is unchanged by reallocation of claims.

### When debt cannot be renegotiated: need for taxpayer transfers
- Any restructuring that reduces default probability increases the market value of debt V'(D) relative to V(D) and reduces equity value; shareholders will oppose unless transferred T = V'(D) – V(D).
- The minimum government transfer to achieve p = p* equals the increase in debt value absent private surplus; transfer requirements vary by scheme.

### Comparative evaluation of restructuring instruments (key findings)
- Optimal (state-contingent) subsidy: government transfers only ex post when A1 is between A* and D, i.e., transfer = D – A1 in that region; no transfer when A1 < A* or A1 > D. This minimizes taxpayer cost for achieving p*.
- Recapitalization with common equity:
  - Required cash to achieve p*: Cash = D – A*.
  - Government transfer needed to leave initial equity unchanged: T = V'(D) – V(D).
  - Recovery schedule after recapitalization: debt recovery for default realizations A1 < A* becomes D – A* + A1.
- Recapitalization with preferred stock or convertibles:
  - Impact on preexisting debt value and required government transfer is the same as for common-equity recapitalization when convertibles/preferred do not trigger default.
  - If convertibles are pari passu and conversion is automatic or at holder’s discretion, recovery of preexisting debt equals the equity-issue case.
- Subsidized debt buybacks:
  - Bank issues equity and uses proceeds to buy back fraction α of debt so that (1 – α) D = A*.
  - Less costly to taxpayers than recapitalization that keeps proceeds as cash; transfer still equals increase in debt value V'(D) – V(D), but implied recovery schedule is closer to first-best.
- Simple (capped) asset guarantees:
  - Government insures assets falling below D up to cap D – A*; guarantees reduce default probability to p* but pay transfers even when A1 < A* (more costly than optimal scheme).
  - Cost to taxpayers equals that of a subsidized recapitalization with identical implied debt recovery.
- Caballero’s scheme:
  - Government gives new equity investors a put (buyback at fixed future price); equivalent in transfer cost to subsidized equity recapitalization but requires no up-front transfer.
- Above-market-price asset sales (government buys fraction a at price (1 + m) a V(A)):
  - More costly for taxpayers than recapitalization or guarantees because recovery for debt holders is higher across realizations; slope of recovery in default zone becomes (1 – a) < 1.
- Sachs proposal (asset sales at book value with recourse via warrants):
  - Lowers default probability and raises debt recovery, but initial equity holders are worse off (they bear dilution or losses ex post); equity holders would oppose.

### General ranking and intuition
- Asset sales (above-market) are most costly to taxpayers because they raise debt recovery over a wide range of realizations.
- Guarantees or recapitalizations that target the payoff region near solvency are more efficient.
- Debt buybacks that change liability structure are closer to first-best (debt-for-equity swap) and reduce transfer size.

### Private and social surplus considerations
- Restructuring can generate private surplus C by reducing costs of financial distress (e.g., attracting customers, restoring interbank confidence, enabling positive-value projects).
  - Literature estimate for costs of financial distress for typical nonfinancial firms: "about 10 percent to 23 percent of ex post firm value."
- Social benefit B from preventing systemic failure sets the upper limit of government willingness to pay.
- Three cases for government intervention when total surplus S = C + B:
  - If C > V'(D) – V(D): private surplus suffices; government transfer not needed.
  - If government intervention is needed and B > T, government willing to pay T; minimum government cost = V'(D) – V(D) – C.
  - If B < V'(D) – V(D) – C: aggregate surplus positive but too small to leave debt contract unchanged; renegotiation of debt contracts needed.

### Design features that affect surplus and incentives
- Separating bad assets (good bank / bad bank) can increase managerial productivity by focusing talent and span of control on typical bank operations.
- Expertise: governments often lack operational expertise; use private sector expertise to run asset management funds or nationalized banks; consider auctions for management contracts.
- Moral hazard and timing of transfers:
  - Up-front cash injections risk "free cash flow" misuse (looting, excessive bonuses, dividends).
  - Ex post instruments (guarantees) reduce ex ante looting risk but may weaken managers’ incentives to maximize A1.
  - Optimal design balances ex ante looting risk and ex post incentive provision; contingent transfers and minimal up-front transfers recommended.
- Convertibles have positive medium-term effects: automatic or holder-discretion conversion can serve as automatic future recapitalization and increase surplus.
- Long-run considerations: insufficient punishment for managers and shareholders combined with large taxpayer transfers increases moral hazard for future crises.

### Limits of market pricing and arbitrage gains
- Market prices V(A) may be below fundamental value J(A) due to limits of arbitrage and funding liquidity constraints.
- Government can capture arbitrage gains V_GOV(A) – V(A) by buying undervalued toxic assets above market price but below fundamental value; gains larger when government takes equity-like claims.
- Arbitrage gains can be used to incentivize private managers if private investors participate in asset purchases.

### Participation under asymmetric information
- When bank managers know asset quality J(A) but markets see average V(A), voluntary participation signals adverse selection and raises required subsidies.
- Two-type example (high-quality and low-quality banks):
  - Both types have default rates pL > pH > p* in the analyzed case.
  - Voluntary recapitalization at market prices forces high-quality banks to require high subsidies (they suffer informational discount); low-quality banks receive informational rents.
  - Result: high subsidy needed to induce participation of high-quality banks; low-quality banks may be oversubsidized and high-quality banks overrecapitalized.
- Role of hybrid securities under asymmetric information:
  - Hybrid instruments (convertible notes, preferred shares) are less information-sensitive than equity and can reduce taxpayer subsidies needed for voluntary participation.
  - Convertible notes trade off: higher conversion ratio raises cash and lowers default probability but increases information sensitivity; if conversion is at issuer discretion, signaling benefits reduce.
  - Preferred shares are less information sensitive than common equity and less than pure debt; can be counted as regulatory capital.

### Participation remedies and policy guidance
- Use hybrid instruments and well-designed asset guarantees to reduce signaling costs and taxpayer transfers.
- Compulsory programs, if feasible, eliminate signaling concerns and associated additional subsidies.
- Rigorous bank examination and public disclosure of asset-quality information can mitigate asymmetric-information problems.

### Overarching recommendations and principles
- Best practice: combine several restructuring elements (asset-side and liability-side interventions) case-by-case to address multiple trade-offs and frictions.
- Design contingent transfers to align manager incentives with maximizing future profits; minimize up-front transfers to prevent misuse of taxpayer funds.
- Ensure transparency of taxpayer costs and final beneficiaries of subsidies.
- Managers and shareholders should be held accountable and face punitive consequences to reduce moral hazard going forward.
- Reduce frictions over the long term to make systemic bank restructuring quicker, less complex, and less costly.

*Source: EXECUTIVE SUMMARY of _spn0912 (IMF staff note).*

### Section II E).

### Section II E)

### Optimality of Asset Guarantees and State-Contingent Transfers
- Asset guarantees or state-contingent transfers that do not involve issuing new securities can mitigate asymmetric information costs by being contingent on realized asset values.
- Among restructuring options discussed, asset guarantees (either the optimal partial insurance scheme or the second-best capped transfer contingent on default) dominate plans that issue claims on assets from the perspective of asymmetric information.
- If a plan asks for nothing from banks in return, every bank will participate (no signaling).
- Government should compute A* using the payoff distribution of the low-quality bank.
  - Doing so overinsures high-quality banks (they will have a default probability lower than p*), but avoids overtransferring to low-quality banks and the informational rent they obtain under recapitalization.
- Asset guarantees without a cap on transfers can serve as a credible signal of government confidence in asset downside risk; a public statement alone is not a credible signal.
- Insurance on the stock price (Caballero’s 2009 proposal) can signal government confidence and commitment to future policy, making nationalization with high dilution a more costly option and potentially increasing equity value when private investors doubt government denials of future nationalization.

### Compulsory Programs
- Voluntary participation creates asymmetric-information costs because participation can be a bad signal about asset quality.
- A compulsory program targeting specific banks (e.g., mandatory equity issuance proposed by Rajan, 2008; and Diamond and others, 2008) can largely mitigate asymmetric information.
- Compulsory programs may require legal changes and may not be feasible when systemic risk is imminent.

### Asset Sales and Within-Bank Adverse Selection (“Lemons” Problem)
- Managers have private information about heterogeneous asset quality and will tend to sell lower-quality assets if given the opportunity.
- Government should pay a price reflecting anticipated quality, which could be determined by an auction mechanism (see Ausubel and Cramton, 2008).
- Balance sheet externalities:
  - The price at which the government purchases assets determines the book value of bank assets under mark-to-market accounting and thus affects regulatory solvency ratios.
  - If the government purchases at the price of low-quality assets, banks may be forced into regulatory insolvency because all banks must book assets at this price, leading to equity write-downs or limits on asset growth (credit crunch).
  - This regulatory-driven outcome can justify subsidized sales prices despite lacking an economic motive beyond regulatory constraints.
- Correlation between amount of toxic assets sold and overall asset quality:
  - If a bank sells more toxic assets to the government than average, market participants may infer below-average overall asset quality, creating a negative signal and reluctance to participate.

### Use of Government Information
- Bank examination (e.g., rigorous stress tests) can yield more accurate information on bank assets; disclosing results can reduce restructuring costs by informing investors.
- Asset guarantees can operate as credible commitments; guarantees without caps put taxpayers at risk but can be persuasive signals.
- Caballero’s scheme (insurance on stock issuances) can credibly commit the government regarding future policy and asset quality, altering incentives around nationalization and dilution.

### Political Constraints and Other Considerations
- Opportunity cost for the government:
  - If government resources are limited, restructuring schemes that constrain other investments should be evaluated for their fiscal trade-offs.
  - Political pressures and fiscal rules may make mobilizing liquid resources immediately costly (see Johnson and Kwak, 2009), favoring mechanisms with only ex post transfers such as asset insurance or Caballero’s scheme—provided there is a credible plan to honor those transfers.
- Political influence on management:
  - Long-term government ownership can cause mismanagement because the government is ill-equipped to monitor and may face political pressures on lending policy.
  - Recapitalization by hybrids without voting rights reduces the risk of government-induced mismanagement.
  - Holding common equity without voting rights can avoid some inefficiencies, but large disparities between control rights and cash flow rights can invite tunneling by other shareholders.
  - Even without formal voting rights, the government or parliament may still partially influence managerial decisions when a bank participates in a government-led restructuring.

### If Bankruptcy Is Inevitable
- If realized asset value A1 < A*, the optimal government response is to let the bank go bankrupt (A* chosen accordingly).
- Under heavy liquidity pressure from markets or depositors, other options may be unavailable.
- To make bankruptcy less destructive, temporary nationalization is often inevitable because:
  - Banks are highly leveraged and opaque; due diligence by private investors requires more time (likely a half year or more) and may not be feasible before collapse.
  - Temporary nationalization mimics private solutions for distressed firms (vulture funds): government holds large common equity share, controls management, acquires necessary information, and is positioned to ask debt holders and stakeholders to share the burden.
  - Government can thus limit taxpayer burden, expedite resolution, and sell the bank to private investors.
- Scope of honored debt during temporary nationalization:
  - At a minimum, transaction-purpose instruments should be honored to save the payment system in the short term (for example, interbank market borrowings).
  - There is less justification for honoring long-term debt.
  - Further discussion is needed on other transaction-purpose instruments, such as bank guarantees on securities backed by credit card debt and accounts payable.

### Case Study — Switzerland: UBS Good Bank/Bad Bank Split (fall 2008)
- Context:
  - Two systemically important banks in Switzerland; only UBS had substantial exposure to U.S. subprime mortgage securities by fall 2008.
  - Swiss authorities focused restructuring efforts on UBS; UBS voluntarily participated. Credit Suisse was offered participation but declined.
- Overview of the plan:
  - Combination of asset sales to an asset management fund (“bad bank”) and recapitalization by convertible notes.
  - Almost all transfers are up-front.
  - UBS was not liable for future losses on transferred assets but kept a partial share of upside.
  - Two potential subsidy sources: (1) price of transferred assets above fundamental value net of buyback option, (2) issuance price of convertible notes above fundamental value.
- Asset-side restructuring (asset sales):
  - October 2008: Swiss authorities and UBS created a special purpose vehicle (StabFund) under the Swiss National Bank to hold toxic assets (the “bad bank”); remainder of UBS intended to be the “good bank.”
  - Up to $60 billion in toxic assets were allowed to be removed from UBS, whose assets totaled almost $2 trillion at that time.
  - UBS provided 10 percent of asset value (i.e., up to $6 billion) for equity of the bad bank but immediately transferred the equity ownership to the Swiss government for $1, covering the first 10 percent of losses of the StabFund.
  - The Swiss National Bank lent the StabFund additional funding, up to $54 billion at the London interbank offered rate (LIBOR) plus 250 basis points.
  - The StabFund’s future loss would not be charged to UBS (nonrecourse condition), but UBS retained some upside option.
  - In the end, UBS transferred only $39 billion worth of its assets to the StabFund.
  - The price was set by an independent valuation process.
- Liability-side restructuring (recapitalization):
  - UBS received $6 billion from the Swiss government by issuing mandatory convertible notes.

*Source: _spn0912 - Section II E).*

### 12.5 percent interest. The proceeds were intended to finance UBS’s equity injection in the

### _spn0912 - 12.5 percent interest. The proceeds were intended to finance UBS’s equity injection in the

### UBS StabFund transaction details
- Loan carried 12.5 percent interest.
- The StabFund was immediately valued at $1.
- If UBS had transferred the assets limit of $6 billion, there would have been no increase in the book value of capital for UBS.
- UBS transferred only about $4 billion to the StabFund, allowing UBS to retain $2 billion worth of capital in net.
- Once the loan is fully repaid by the StabFund, UBS can exercise its option to repurchase the fund equity by paying the Swiss National Bank $1 billion plus 50 percent of the equity value at the time of exercise in excess of $1 billion.

### Corporate governance implications (UBS)
- UBS shareholders already completed a voluntary recapitalization in April 2008 via a rights issue without public help.
- Old management exited in April 2008 at the annual general meeting when UBS asked for the rights issue.
- As a majority owner, a public authority obtains the right to appoint independent nonexecutive directors and to limit executive compensation and dividend payouts (noted in other cases below).

### United Kingdom: recapitalization and asset guarantees for RBS and Lloyds-HBOS
- The U.K. approach combined recapitalization (fall 2008) and asset protection (introduced January 2009).
- Recapitalization:
  - Government injected capital with preferred shares.
  - Government ended up owning 58 percent (₤20 billion) of RBS total ownership.
  - Government ended up owning about 44 percent (₤17 billion) of Lloyds-HBOS.
- Asset Protection Scheme (January 2009):
  - ₤325 billon (14.5 percent) of end-2008 assets for RBS.
  - ₤260 billon (23.4 percent) of end-2008 assets for Lloyds-HBOS.
  - If valuation of assets fell below a tailored threshold, the government would compensate the loss up to 90 percent.
  - First loss (threshold) amounts: ₤19.5 billon (6 percent) of protected assets for RBS and ₤25 billion (9.6 percent) of protected assets for Lloyds-HBOS.
- Costs and ownership after guarantees:
  - Insurance fees were paid in preferred shares.
  - For Lloyds-HBOS, government would own close to 50 percent of the total economic stake.
  - For RBS, government economic stake would rise to more than 80 percent after insurance fees and extra preferred-share capital injection; conversion to common equity made government voting rights about 70 percent.
- Corporate governance measures:
  - Government obtained right to appoint new independent nonexecutive directors.
  - Government limited executive compensation and dividend payouts.
- Complementary U.K. measures noted:
  - Credit Guarantee Scheme (October 2008) provided insurance to debt holders.
  - Asset-Backed Securities Guarantee Scheme (January 2009) guaranteed newly issued AAA-rated mortgage-backed securities issued after January 2008.
  - Asset Purchase Facility introduced in January funded by the Treasury and established within the Bank of England.

### United States: Geithner Plan (Financial Stability Plan) as of May 2009
- Rationale:
  - In fall 2008, asset quality of systemically important banks was not fully known; policy design accounted for informational shortfalls.
- Initial measures:
  - Capital Purchase Program (part of Troubled Assets Relief Program) in October–November 2008 allowed banks to receive cash by offering preferred shares to the Treasury; across-the-board recapitalization with few conditions.
  - Targeted exceptions: Citibank received additional recapitalization and asset guarantees in November 2008; these measures were formalized in January 2009 for Citibank and Bank of America as the Targeted Investment Program and Asset Guarantee Program.
- February 2009 Financial Stability Plan (Geithner Plan) components:
  - Information gathering: a compulsory stress test (Supervisory Capital Assessment Program) for the 19 largest banks to evaluate asset risks and determine extra capital needs.
  - Recapitalization (Capital Assistance Program):
    - Banks required to raise extra capital identified by the stress test.
    - Government option: issue of preferred shares that automatically convert into common equity after seven years; earlier conversion possible with issuer discretion and regulator approval.
    - Treasury investments managed under a separate entity (Financial Stability Trust).
  - Asset-side restructuring (Public-Private Investment Program):
    - Several funds to purchase legacy loans and legacy securities.
    - Funds buy pools of legacy loans sold by banks; prices determined by competitive bidding.
    - Financing support: each fund will receive 50 percent equity participation from the Treasury without voting rights.
    - For loan-purchasing funds: funds can issue debt guaranteed by the FDIC with leverage ratio up to 6; FDIC will supervise the funds.
    - For securities-purchasing funds: Treasury provides 50 percent equity stake without voting rights and lends money to each fund at up to a 2-to-1 leverage ratio.
    - Loans to funds are nonrecourse: if asset values are very low, funds can default on the Treasury; fund managers’ responsibility limited to losses on their own investment.
- Other U.S. liquidity measures noted:
  - FDIC temporarily insured new debt holders of FDIC-member banks under the Temporary Liquidity Guarantee Program.
  - Federal Reserve started lending up to $200 billion on a nonrecourse basis to holders of certain AAA-rated asset-backed securities under the Term Asset-Backed Securities Loan Facility (TALF).
- Corporate governance and constraints:
  - Plan requires banks to restrict executive compensation, dividends, stock repurchases, and acquisitions.
  - Plan prohibits political interference in investment decisions; Treasury will make all contracts public.
- Evaluation concerns highlighted:
  - Moral hazard in asset management funds due to government-sponsored inexpensive leverage.
  - Calibration challenge: determining subsidy needed to secure participation of high-quality managers and how subsidy will be shared between banks and fund managers.
  - Proposal: auction management and cash flow rights to better calibrate subsidies (citing Bebchuk, 2009; Spence, 2009).

### Key findings, trade-offs, and policy recommendations (from Conclusion)
- General principles:
  - Restructuring should aim to limit transfers from taxpayers by avoiding unnecessary subsidies to debt holders and maximizing economic value created by restructuring.
  - No single solution fits all; combining asset-side (e.g., sales of toxic assets) and liability-side (e.g., recapitalization with preferred shares) measures is common and often necessary.
- Specific analytical findings:
  - In a Modigliani-Miller framework with cash flows independent of capital structure, converting some debt into equity can theoretically restructure a firm, but this is difficult in practice.
  - Without changing debt contracts, all restructuring involves government transfers. A plan subsidizing common equity issues and buying back debt is close to optimal.
  - Subsidized asset sales are more costly to taxpayers because debt holders benefit more.
  - Restructuring design must account for value created or destroyed by changes in participant behavior; manager expertise and incentives are key concerns.
  - If assets are undervalued due to liquidity or “lemons” problems, government can profit by buying assets above market value but below fundamental value; assessing such undervaluation is difficult.
  - Asymmetric information on future payoffs makes equity holders reluctant to support restructurings involving new-claim issues.
- Ways to avoid restructuring impasse caused by asymmetric information:
  - Conduct stress tests with credibly publicized results.
  - Use compulsory rather than voluntary schemes.
  - Provide contingent guarantees for banks to avoid new-claim issues.
  - Make banks issue low-information-sensitive claims such as convertible debt or preferred stocks.
- Governance and accountability:
  - From a long-run perspective, managers and shareholders of bailed-out banks should be punished in a way that discourages excessive future risk taking.
- Institutional reforms suggested to reduce frictions over time:
  - Better legal framework to handle debt renegotiation quickly and with smaller systemic risk.
  - Reduce opacity via more timely and in-depth disclosure requirements for bank asset information and counterparty exposures.
  - Regulation to encourage conversion clauses in long-term debt contracts so debt automatically converts into equity in distress.

*Source: Excerpt from IMF PDF chapter _spn0912.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/spn/2009/_spn0912.pdf_