The causes of the global financial crisis were multi-faceted but revealed still unresolved weaknesses in national and international financial oversight and resolution frameworks. In particular, many governments in the crisis-hit countries had to provide unprecedented levels of support to contain the crisis and protect financial stability. These interventions have not only contributed to a significant increase in sovereign exposures but, in many countries, they have also risked weakening market discipline and worsening moral hazard.
To address moral hazard and the problem that institutions can become too important to fail, proposals for contingent capital are gaining ground.3 Most recently, Switzerland has proposed a higher regulatory capital requirement (19 percent of risk-weighted assets) for its two largest banks, of which 9 percentage points may be held in the form of contingent-convertible debt. The Basel Committee on Banking Supervision (BCBS) has similarly proposed that all noncommon equity regulatory capital of internationally active banks be convertible to equity or subject to permanent write-downs when it is determined that the bank is no longer viable. The Financial Stability Board (FSB) and the European Commission, in their efforts to address risks associated with systemically important financial institutions (SIFIs), are also examining mechanisms that convert debt into equity or the write-off of debt (including unsecured senior debt), based on (i) contractual agreements between banks and investors; or (ii) supervisors’ statutory powers in the context of bank resolution.
Contingent capital provides an “automatic” mechanism for increasing equity capital and reducing debt of a financial institution in times of stress. It enables raising capital at times when other options are impossible, either owing to unfavorable market conditions or because they are unattractive to shareholders. Furthermore, automatic conversion by avoiding fire sales could help avoid contagion in times of systemic stress.
Concerns have been raised, however, about their operational aspects and implications for market dynamics. These instruments remain largely untested and could have unintended consequences, particularly in times of high market volatility and uncertainty. Their marketability, including whether there will be sufficient demand with traditional investors, is far from granted. Design features are key to ensuring their effectiveness and avoiding risks, including systemic ones. Others also warn that a conversion could have negative signaling effects, lead to contagion, and be subject to price manipulation (Sundaresan and Wang, 2010; Goodhart, 2010). For example, some have cautioned against using triggers based on systemic risks or regulatory discretion, since these would make pricing these instruments difficult.
The rest of the note is organized as follows. Section II discusses the economic rationale for contingent capital instruments. Section III focuses on the operational aspects of contingent capital instruments, especially the pros and cons of various triggers and conversion rates, and how they could influence the rating and pricing of these instruments. Section IV discusses some recent contingent capital proposals and the potential role of contingent capital in the framework of crisis prevention and crisis management. Section V concludes. Appendix I presents a simple two-period model on how the expectation of a public bail-out encourages excessive risk taking and how contingent-convertible bonds can mitigate such excesses. Appendix II details contingent-capital triggers and conversion options and conditions, and Appendix III compares contingent capital with hybrid and subordinated debt instruments.
II. Economic Rationale of Contingent Capital
Contingent capital provides an “automatic” mechanism for increasing capital and reducing debt of a financial institution in times of stress. It enables raising capital at times when other options are impossible, either owing to unfavorable market conditions or because they are unattractive to shareholders (Duffie, 2010). This latter “recapitalization gridlock” reflects the unwillingness of shareholders to dilute their equity by share issuance or by “fire sales” in unfavorable market conditions (Brunnermeier,2009 ; Adrian and Shin, 2010). Moreover, automatic conversion by avoiding fire sales could help avoid contagion in times of systemic stress.
Contingent capital instruments are expected to deal with the market failure associated with “too-important” or “too-connected” to fail. Depending on the choice of triggers and conversion rates, contingent capital instruments could be designed specifically to increase capital buffers, ensure prompt recapitalization, or increase loss absorbency before a bank-default event.
The credible threat of losses due to conversion and dilution could help reduce risk taking by managers, shareholders, and bond holders. The threat of heavy dilution should encourage shareholders to require more prudent corporate governance and risk-control procedures within a bank. Similarly, requiring bond holders to bear part of the cost of a future bank recapitalization would enhance their incentive to exercise greater market discipline. It also has been suggested that bank manager bonuses could be paid in the forms of contingent convertible debt instruments to reduce their incentives for excessive risk taking. We developed a simple two-period model to illustrate the benefits of CoCos in promoting market discipline (Appendix I). The model implies that the effect of CoCos on the bank’s risk-taking behavior is equivalent to that of a risk-based, pre-funded bank resolution fund.
From the banks’ perspective, contingent capital may be preferred to equity because (i) it may potentially be cheaper (if the interest expense is tax deductible); (ii) before conversion, it could be a nondilutive source of capital for existing shareholders, so that their issuance does not change corporate control; and (iii) it may be acceptable as Pillar 2 capital for supervisory stress tests.
Contingent capital instruments should be supplementary to common equity. While they may be used to replace the existing hybrid capital that had poor loss absorption during the crisis, they should not compromise the objective of capital transparency and their use should be based on an enhanced capital structure under the Basel III framework. Common equity is no doubt higher-quality capital in terms of loss absorption; however, an excessively high common-equity requirement could lead to perverse risk seeking, as bank managers struggle to maintain a return on equity demanded by investors.
Contingent capital differs from existing hybrid instruments in two important ways. First, contingent capital instruments are dated debt with debt/equity conversion or debt written-down contractual clauses, while the existing convertible hybrids are perpetual debt, which contractually offers the possibility to absorb losses primarily through the deferral of coupons (noncumulative) or extension of maturity. Second, the conversion of contingent capital instruments is automatic upon the activation of the predetermined conversion trigger, but the conversion of existing hybrids is largely at the discretion of banks, unless regulatory capital ratios are breached.
During the recent crisis, most hybrid capital instruments did not absorb losses as they were designed to do. This was partly due to banks’ reluctance to send negative signals to the markets and partly due to regulatory forbearance, overestimated capital ratios and/or capital injections from the governments, which prevented the breach of regulatory ratios. (See Appendix III for a more detailed discussion on pre-crisis hybrids). The crisis has exposed the weaknesses in capital structure and measurement, especially how Tier 1 capital ratios masked the underlying capital position of banks. Since then, the Basel III framework has raised the minimum common equity ratio from 2 percent to 7 percent (of which 2.5 percent is a conservation buffer) and substantially enhanced the quality of capital. It will be critical to build on past experience to avoid CoCos repeating the same failure as hybrids, including marring once again capital’s loss-absorbency availability and transparency.
III. Operational Aspects of Contingent Capital
The design of trigger and conversion rates will be crucial to ensure the effectiveness of contingent capital to achieve the intended objectives (Box 1). In this section, the pros and cons of various trigger conditions and conversion rates are discussed, including with regard to their impact on the pricing and marketability of these instruments and on market dynamics.
Triggers of contingent capital instruments determine the probabilities of conversion or conversion risks. To the extent that the conversion itself is not a default event, the idea is to impose losses upon creditors, who would otherwise be affected only by default. The trigger could be linked to system-wide conditions or bank-specific conditions, or a combination of both (Appendix II, Table 1).
Systemic triggers would be based on the condition of the whole financial system (for example, liquidity conditions, a market volatility index, or supervisory declaration of a systemic crisis). Such instruments may be more efficient at addressing systemic risks, as they increase capital across the banking system at the same time. However, systemic triggers would mean forgoing any benefits from the market-discipline effect, since there would be limited additional incentive for banks to improve their individual risk management. Moreover, given the difficulty of predicting systemic events and the discretionary element in calling such circumstances, the rating and pricing of these instruments would be complex.
Bank-specific triggers would be based on the state of an individual institution. These could include a bank’s capital ratio, share price or CDS price, or an assessment of nonviability by the supervisor. Instruments with market- or capital-ratio-based triggers are likely to be easier to rate and price compared to those with triggers based on supervisory discretion. Reported capital ratios seem to align better with a regulatory capital framework and objectives, but tend to be lagging indicators of a bank’s financial condition and, thus, may not trigger conversion sufficiently early. A concern with market-based triggers is that they can be more easily subject to market manipulation, though this problem may be overcome to some extent by, for example, basing the conversion trigger on the moving average of a market price (Flannery,2009).
Instruments with high-level triggers, i.e., set at capital levels well above distress thresholds, can be a useful tool for crisis prevention. Instruments with high triggers could help mitigate systemic risks by ensuring recapitalization well before a bank faces serious difficulties and a potential loss of broader market access. They would also provide strong encouragement to shareholders and bondholders to exercise market discipline, as they act earlier to reduce risk taking and improve management and align governance with the long-term sustainability of the bank.4
Box 1.A Schematic Exposition of Use of Contingent Capital Instruments
Crisis Prevention: Several tools work to limit the prospect of systemic crisis: (i) better management incentives to lower the risk appetite; (ii) higher capital buffers (e.g., Basel III) and additional loss-absorbing capital instruments that could include CoCos; (iii) revamped disclosure to inform markets of the true and fair view of the capital position of financial institutions; (iv) more intrusive supervision—impose restrictions on dividends mandate capital plans earlier; and (v) promote pre-emptive restructuring by virtue of effective resolution and recovery planning (e.g., sale of nonstrategic subsidiaries).
Crisis management (going concern): Revamped and more diligent prevention efforts serve to lower contagion by a distressed SIFI. Allow management and authorities to undertake progressively more aggressive restructuring measures to stave off insolvency as capital levels deteriorate (see below table) and prevent systemic crisis. Central bank emergency liquidity assistance could be made available under the conditions that equity solvency is sustained and borrowing is properly collateralized.
Orderly resolution (gone concern): Once all restructuring measures are exhausted, countries will require a framework that provides a menu of resolution-transaction options. Resolution-transaction options should include:
(i) a transaction for the purchase of viable assets and assumption of certain liabilities by an existing institution;
(ii) temporary creation of government-owned bridge financial institution (both necessary for SIFIs); and
(iii) liquidation of assets with deposit transfer/payout supported by depositor guarantee schemes (available for smaller nonsystemic institutions).
|Triggers||Objective||Degree of Stress||Debt Instruments||Approach|
|→→ Restructuring measures as capital declines →→||Examples of high triggers: - Common equity ratio 7 percent of risk-weighted assets (Swiss proposal) - 7 percent above plus any countercyclical buffer requirements||Recapitalization to stabilize the situation and build market confidence||Deteriorating financial situation (going concern)||High-trigger CoCos convert to equity||Conversion through ex ante contractual agreement between issuers and investors|
|Examples of low triggers: - Common equity 5 percent of risk- weighted assets, or ratio of equity to nonrisk- weighted assets||Provide additional recapitalization to prevent receivership||Threat of failure (going concern)||Low-trigger CoCos convert to equity||Conversion through ex ante contractual agreement between issuers and investors|
|- Point of nonviability or other resolution triggers||Compulsory restructuring to prevent insolvency resolution||Threat of insolvency (going/gone concern)||In order of claim priority, subordinated debt and, ultimately, unsecured senior debt could be subject to debt-to-equity conversion||Contractual possible, but most likely would require statutory powers|
Instruments with low-level triggers can be useful tools for orderly resolution. In this case, the trigger would be set at the point of nonviability5 as a way to ensure private sector involvement in the bank restructuring process, as was done in the case of Lloyds Banking Group (LBG).6 There is a possibility that market confidence in a bank’s financial condition could weaken and create liquidity pressures as bank capital approaches the conversion trigger. This argues for the careful integration of their use with emergency liquidity facilities and supervisory intervention techniques to ensure they increase, rather than decrease, the bank’s chance of survival and that they do not trigger a broader loss of confidence.
An important consideration would be the conversion rate, which determines the burden sharing between shareholders and bondholders. The impact on the incentives to monitor bank management would depend on which party loses most in a conversion: with a high rate of dilution, those who lose are the ex ante shareholders; with little dilution, those who lose are the holders of contingent debt. Possible options include the following (Appendix II, Table 2):
- Conversion into a predetermined number of shares based on the par value of an instrument divided by the issuing bank’s share price at the time of issuance. Upon conversion, the dilution to the shareholders is limited to lower share prices, but bondholders would suffer losses as if they were shareholders.
- Conversion into anex postdetermined number of shares, based on the par value of the instrument divided by the share price at the time of conversion. Upon conversion, the holders of CoCos would receive the notional amount in shares and would not suffer losses if they could sell the shares, but would absorb future losses as new shareholders. The ex ante shareholders, however, would suffer a much stronger dilution than in the previous case. Such instruments may give rise to an “infinite dilution” or “death spirals” when share prices are falling close to zero. Hence, they may be more prone to market manipulation. In this context, some restrictions (circuit breakers) on conversion rates may be justified (see De Martino et al,2010) if such risk is not reflected in prices.
Contingent capital instruments with debt write-off features would be more suitable for (but not limited to) cooperative and mutual banks that are prevented by their legal structure from issuing shares. Such instruments could be designed to impose a significant haircut on bondholders upon the trigger event. The Rabobank Senior Contingent Notes (SCN) provide an illustration whereby the SCN’s original principal amount can be written off (but not written back up) by 75 percent if the bank’s common equity ratio falls below 7 percent. Instruments with large haircuts would provide a powerful incentive for bondholders to monitor the bank’s performance and resilience closely. However, the stringency of the losses that would be imposed might mean the cost of these instruments could be steep and, thus, likely to be limited to issuance by strong banks (those with a very low probability of failure).
The pricing of and the demand for contingent capital instruments will depend on conversion triggers, types of conversion, and conversion rates. The level of the trigger determines the conversion risk; therefore, for a given conversion rate, the cost of issuing an instrument with a low trigger should be cheaper than one with a high trigger. Conservative and traditional real-money investors (whose investment horizon is longer-term) are more likely to be drawn to instruments with a lower probability of conversion (and loss sharing) and more dynamic/speculative investors (with higher risk returns investment strategies), and should be more willing to consider high-level trigger CoCos (with greater loss risk).
Figure 1.Issuance of Tier 1, Tier 2, and Lower Tier Capital in Europe
Like the hybrid market in the late 1990s, the contingent capital market would need some preconditions to become a deeper asset class. It took close to five years after its inception in 1997 for the European hybrid market to reach critical mass and become a mainstream asset class (Figure 1). Likewise, contingent capital may need several years before investors become more familiar and comfortable with the instruments. The main barriers that would also need to be overcome relate to (i) obtaining ratings; (ii) lifting mandate restrictions of certain investors to hold equity-like products;7 and (iii) obtaining the inclusion of CoCos in benchmark indices.
The tax and regulatory treatment of CoCos would be important in developing this instrument. Tax deductibility of interest would help lower the cost of contingent capital. After the crisis, two European banks have issued tax-deductible contingent capital instruments, but with different regulatory capital status—lower Tier 2 instrument by Lloyds and noncapital senior debt by Rabobank (as discussed earlier). However, there has been no further issuance of CoCos since then, possibly due to the regulatory uncertainty regarding the treatment of contingent capital instruments (Appendix III). Most existing hybrid securities will no longer qualify as regulatory capital under Basel III after January 2013 and will be phased out over time by 2023.8 The regulatory landscape—which will be shaped by the policy initiatives of the Basel Committee and the FSB (discussed below)—will be an important factor in the development of a market for contingent capital instruments.
A proper sequencing could help expand the CoCo market over time. Investors differentiate between strong and weak banks, and having strong banks issuing at first would generate confidence in the new instruments and market stability until investors are comfortable enough to increase exposure to “second tier” (riskier) issuers. The investor base could initially include a small pool of buyers, such as hedge funds, sovereign wealth funds, and high-yield or equity investors. Over time, traditional credit investors, real money, asset managers, and insurers could step in and provide depth to the market.
The depth of existing subordinated debt and equity markets (Figure 2) provides an indication of the pool of potential investors. If contingent capital is to target fixed-income investors, the outstanding Tier 1 and Tier 2 debt markets in Europe and the United States indicate that the current institutional market capacity is about $260 billion for the United States and €580 billion for Europe. European and U.S. equity markets for bank stocks, combined, amount to about $923 billion. An important issue to consider is the potential “crowding out” risk if equity investors were to prefer CoCos to equity—this may lead to higher costs of issuing common equity.
There may be a case for imposing certain restrictions on holders of convertible instruments. In particular, if the investors are other leveraged financial institutions of systemic importance, the potential write-off could have contagion effects. One of the reasons regulators did not enforce hybrid coupon deferrals and maturity extensions was to avoid penalizing insurance companies, who were large holders of those instruments and would have been forced to sell upon such event. The same political economy concerns could also apply to contingent capital. Regulators would also need to ensure that, after conversion, the new equity holders would be “fit and proper” owners of a bank. However, it is unclear whether it would be necessary for regulators to impose limits on SIFIs’ cross-holding of CoCo instruments, as the proposed more onerous liquidity requirements under Basel III are likely to limit cross-holdings in general.
Figure 2.Outstanding Bank Debt and Equity in Europe and the United States
IV. Contingent Capital Proposals
This section discusses the recent contingent capital proposals at international and national levels, as well as the potential role of contingent capital in a framework of crisis management.
A. Basel Committee Point of Nonviability Proposal
The Basel Committee issued minimum requirements on January 13, 2011, to ensure that all classes of capital instruments fully absorb losses at the point of nonviability before taxpayers are exposed to loss.9 More specifically, it proposed that all noncommon equity Tier 1 and all Tier 2 instruments must have a provision in their terms and conditions that requires them to be written off or converted into common equity upon the occurrence of the trigger event. The trigger event is the earlier decision by the relevant authority (i) that the firm would become nonviable without a debt write-off; or (ii) that the firm would become nonviable without a public sector injection of capital or equivalent support. The relevant authority in determining the trigger event is the authority where the capital is being given recognition for regulatory purpose. In the case of conversion, these capital instruments must be converted into common stock of the issuing bank or of the parent company of the consolidated group, including any successor in resolution.
This proposal can be considered as a low-level trigger contingent capital instrument similar to a resolution tool, as discussed earlier. It remains to be seen how the discretionary element may affect the pricing and marketability of these capital instruments. Regulators may be allowed to intervene before capital deteriorates to the point at which a capital-ratio-based, low-trigger CoCo would have been converted. Thus, the proposed trigger that is linked to the decision of public support would ensure that private creditors take losses before government intervention. It is important to note that the BCBS report also emphasizes that its proposal for contingent capital should not diminish the need for reforming national insolvency and bank-resolution schemes to create stronger incentives for private sector solutions for failing SIFIs.
B. Contingent Capital as Additional Loss-Absorbing Capital Buffer for SIFIs
Various proposals have been made to use contingent capital to meet part of the requirements for additional loss-absorbing capital. The countercyclical capital buffer is expected to be built up rarely (at times of bubbles); however, the additional loss-absorbing capital requirements for SIFIs may represent a substantial amount. The calibration of additional capital requirements for SIFIs is still under discussion and may be defined as a Pillar 2 add-on, calibrated by the systemic risk contribution of each bank. For illustrative purposes, for additional loss-absorbing capital requirements of 2 percent of risk-weighted assets, 25 global SIFIs10 would need to issue $300 billion of CoCo bonds. Should those same SIFIs use CoCos to meet the countercyclical buffer in bad times, they may need to issue another $400 billion.
The contingent capital proposal by the Swiss regulators represents the first national initiative (Box 2).11 The high-trigger CoCos are designed to meet a bank’s recapitalization needs when its capital situation is deteriorating, thus contributing to the stabilization of the bank before restructuring actions become necessary. The low-trigger CoCos are expected to generate capital to prevent the bank from being put into receivership and to ensure an orderly resolution that is funded first by private creditors, which should minimize the need for possible government support.
Contingent capital is not designed to deal with liquidity problems. The conversion would stop the interest payments on convertible bonds, but would otherwise not generate additional liquidity for banks. However, if perceived negatively by the market, the conversion may actually generate liquidity squeeze for the institution. To avoid the negative impact on market dynamics, triggers would need to be set off long before liquidity pressures start, in order to forestall a self-fulfilling presumption of a liquidity crisis (Duffie, 2010). However, forestalling a liquidity crisis with convertible debt would require large amounts of such debt and may require extending the coverage of debt-equity conversion to unsecured senior debt under the bail-in schemes (Box 3). In practice, the use of contingent capital instruments may require pre-committed liquidity support from the central bank or from a consortium of private banks.
Box 2.The Swiss Contingent Capital Proposal
The Swiss Commission of Experts released on October 4, 2010 a comprehensive framework to address risks associated with SIFIs, which should be adopted into law by early 2012. Part of the framework is to raise total capital ratio to 19 percent: 10 percent in Common Equity Tier 1 (CET1) and 9 percent in contingent-convertible bonds (CoCos), including 3 percent with a “high level trigger” of 7 percent of CET1 to meet the enhanced “additional capital conservation buffer,” and 6 percent with a “low level trigger” of 5 percent of CT1 as a SIFI additional loss-absorbing capital requirements or “progressive component.”
Comparison of Swiss Proposals and Basel III Proposals
Main features of Swiss contingent capital
- The triggers would be “contractually predefined” and based on common equity ratios in line with Basel III framework.
- The equity conversion prices would not be predefined and could be either set at the time of conversion or at the time of issuance. Write-down could be an alternative. Pre-conversion, these contingent capital instruments would be classified as dated subordinated debt with nondeferrable coupons, also known as (ex-lower) Tier 2 debt.
- For the implementation of the Swiss requirements, the same timeframe will apply as in the case of Basel III, by 2019. The accumulation of capital in the various categories will be overseen by FINMA and the SNB as part of capital planning.
Box 3.Bail-in” Proposals: A Statutory Approach to Debt Restructuring
“Bail-in” proposals mainly represent a statutory approach to debt write-downs or debt-equity conversion. This approach would likely require endowing regulators with statutory powers to cancel, writedown, or convert existing claims debt holders, or override pre-emption rights.
- Unlike regulatory contingent capital, “bail-in” schemes deal directly with the resolution of SIFIs. They provide supervisors with discretionary power to recapitalize an insolvent SIFI more quickly (a few days) than under the existing bankruptcy rules (a few months or longer), thus gaining a crucial time period to assess the viability of the bank and, when necessary, proceed with an orderly and rapid liquidation. Because the recapitalization will be based on the conversion of private debt into equity rather than a bailout, moral hazard will be reduced.
- “Bail-in” is better suited than contingent capital in dealing with larger shocks/tail risks, because financial institutions would ordinarily maintain a substantially large amount of unsecured debt that could be eligible under “bail-in” schemes for conversion into equity.
“Bail-in” could face tough legal challenges and strong political opposition. A clear and convincing legal procedure will be essential to its effectiveness. Regulators would need statutory power to write down existing claims of equity and debt holders, override pre-emption rights, and change management. Upon the occurrence of the trigger event, creditors would be forced to give up full legal claims, presumably in exchange for overall value maximization and so that business operations could continue normally. This would interfere with laws that guarantee property rights. Therefore, “bail-in” schemes will require changes in current legislation to legitimize the interference. The problem is further complicated by the lack of consensus on whether all unsecured debt should be subjected to the statutory power, though the general agreement is that collateralized or any other secured claims should be excluded.
There will also be challenges associated with cross-border implementation. Regulators in a home country with the “bail-in” statutory power may not be able to write down the debt that is booked in a foreign country or is governed by a foreign law. In this case, the effectiveness of the “bail-in” could be reduced significantly, unless states are willing to adopt laws recognizing the statutory power of the resolution authorities in other jurisdictions.
V. Concluding Remarks
Contingent capital instruments should be considered as part of a comprehensive and consistent crisis prevention and management framework. While these instruments could be useful additions to the crisis-management toolkit, they are not intended to be standalone tools. Instead they should be implemented within a comprehensive framework that includes strengthened supervision, an enhanced capital base, improved disclosure, and an effective resolution regime. Their design should also avoid adding more procyclicality during crisis times.
The design of the conversion trigger and the conversion rate will be crucial to ensure effectiveness. As each objective (e.g., prevention, resolution, and market discipline) entails a different design, it will be important to identify the design priorities. In particular, to avoid adverse unintended consequences and enhance market acceptance, several considerations should be at the fore.
Triggers based on capital ratios of individual institutions are preferable, since they are less prone to market manipulation and market contagion, and both shareholders and bondholders will have an incentive to avoid letting capital fall near the trigger point. As a tool for crisis prevention, conversion triggers should be set high enough, relative to the point of insolvency, to ensure conversion for individual institutions well ahead of the emergence of distress. As a tool for an orderly resolution of a failing bank, conversion triggers may be set at a low level, i.e., which convert close to the point of an institution’s insolvency, thereby providing a broader private sector investor-base for sharing the burden for bank failures with the public sector.
Ensuring consistency, transparency, and standardization will be important to avoid complex structures and to support the G-20 objective for higher capital transparency. Some standardization might be necessary to avoid complex structures. Regulators, issuers, and investors need to establish transparent criteria to ensure proper marketability of contingent capital instruments and allow sufficient depth and liquidity for their successful use. Despite their potential use in addressing moral hazard, contingent capital instruments are untested. Therefore, supervisors will need to be vigilant in monitoring (i) the design and issuance of contingent capital instruments; (ii) the implied transfer of risks within the financial system; and (iii) potential build-up of systemic risks, including liquidity risk.
A. The Basic Model
Consider a bank starts at period 1 with an asset consisting of its own cash capital of M and an investment of I (i.e., loans), which is financed with outside equity (NP) and debt (D):
where P is the price of each share and N is the number of issued shares.
At period 2, the repayment on debt is (1+r) D, where r is the interest rate on debt. For simplicity we assume that risk-free interest rate is equal to zero. The gross payoff on the investment I is stochastic R.
- i) If R< (1+r) D, the bank defaults on its debt. Assuming default is costless, creditors (bond investors) receive R while shareholders receive zero.
- ii) If R> (1+r) D, the bank pays off debt, and each share gives right to residual [R-(1+r) D].
In the equilibrium, the interest r on debt is determined by:
The value of each share is given by
And the gross return on the bank’s own capital M is
Result 1. For a given expected profit, the gross return increases when M decreases. Therefore, the bank’s incentive is to leverage as much as possible (thus increase default probability), so that potentially high profits from its investment will be spread over a common equity base.
B. The Model with Government Bail-out
How does a government bail-out encourage excessive risk taking by a bank? The simple intuition is that the bank now expects that when R< (1+r)D, it could receive a transfer T from the government to prevent it from default on its debt: T = (1+r)D -R. The bank therefore has one more reason to take excess risks: to maximize the T.
We now introduce a risk measure in investment in the basic model to illustrate how bail-out could lead to more risky behavior. For simplicity, we assume the bank holds only loans in its assets; hence, cash M is equal to zero. Assume
Assuming that there is a default cost d, if the bank defaults the creditors now will only receive R-d, while shareholders receive nothing. The bank defaults when
The equilibrium interest rate r on debt is determined by:
and the value of capital is given by
Therefore the total value of the bank is
Result 2. To maximize its value, it is optimal for the bank to take the minimum risk because of default cost d.
However, if the interest rate does not respond to risks because of a bail-out guarantee that prevents the default of the bank ex post, then
Result 3. With the expectation of bail-out, the bank will take maximum risk as long as the probability of default is greater than zero.
C. The Model with a “Bail-Out” Insurance Fund
Assume that the government imposes a “bail-out” premium ex ante on the bank for the expected value of bail-out. How would this change the bank’s risk-taking behavior? A practical example would be the bank resolution fund that has been set up in several advanced countries. With the “bail-out” insurance, the bank will receive a transfer T= [(l+r) D -R]+ in period 2—but it all goes to repay the creditors. In period 1, the bank must pay a premium that is equal to the expected value of the transfer, which depends on risk σ
Hence, a bank perceived as riskier would pay a higher insurance premium.
The value of capital net of the insurance premium is given by
which is not increasing with risk σ
Result 4. A “bail-out” insurance fund reduces the incentives for banks to take excessive risk.
D. A Model with Contingent Convertible Bonds
In this section, we will show that CoCos can be equivalent to optimal capital insurance. Following the basic model, now the bank issues convertible debt (Dc) in period 1, in addition to regular debt (D), so that
Consider three possible cases:
- i) If R< (1+r) D: the bank defaults (as before). The creditors of regular debt will receive R-d, while shareholders and CoCo investors receive nothing.
- ii) If (1+r) D< R < (1+r) D + (1+rc) Dc, CoCos are converted and CoCo investors receive a share (φ) of total capital (φ=1 implies total dilution to the original shareholders). φ = N’/(N+N’), where N’ is the number of new shares.
- iii) If R> (l+r) D + (1+rc) Dc, there is no default and no CoCo conversion.
We first assume that a default does not trigger a bail-out and, in this case, the interest rates are determined by:
One can show that the interest rate is higher for convertible debt than for plain vanilla debt. The value of capital is
Result 5. The value of capital is maximized by the bank taking minimum risk. In this case, contingent convertible bonds reduce the bank’s incentives to take excessive risks by eliminating the bail-out or reducing the probability of default.
Appendix II.Summary of Contingent Capital Trigger Conditions
|Bank-specific triggers||Institution-specific factors||✓|
|Clear incentives for shareholders and managers|
More targeted and focused on distressed institutions only
|May be insufficiently responsive to systemic risks (e.g., to the need to recap entire banking system simultaneously)|
|1. Financial soundness indicators (e.g., capital ratios)||✓|
|Easy to price|
Intuitive and simple to understand and implement
|Low publication periodicity means indicators likely to lag actual financial conditions |
Back-tested poorly against crisis; hard to determine a trigger level that is informative of financial distress ex ante
|Conversion occurs if a bank’s capital ratio (e.g., Tier 1 or Common Equity over RWA) falls below a pre-determined threshold|
Lloyds exchange: trigger set at “5% of published core Tier 1 capital to total RWA”
|2. Market indicators (e.g. share price, CDS spread)||✓|
|Forward-looking (assuming reasonably efficient markets)|
Better early warning indicator of distress than regulatory ratios
|Markets can be distorted, especially during times of stress, and therefore fail to provide the right signals|
Higher chance of false positives, i.e., premature conversions, which lead higher funding cost
Price manipulation (via short-selling) and the self-fulfilling threat of equity dilution could inflict a confidence-induced downward spiral that eventually triggers conversion
|✓||Conversion occurs if firm’s stock price drops below some level or the credit default spread (CDS) spread exceeds some level|
|3. Supervisors’ discretion to trigger conversion when financial stress is deemed sufficiently high (e.g. stress test results)||✓|
|Addresses the lag problem with capital ratios and other accounting/regulatory measures|
Uncertainty of outcome limits market manipulation and over-engineering of contingent capital with pre-determined conversion criteria
|Not automatic, which increases negative signaling problem|
Strong reliance on regulatory judgment and supervisory discretion might be counterproductive if negative signaling effect encourages forbearance Lower marketability: increases funding costs since investor may charge a premium for the uncertainty associated with the outcome of the stress test
|✓||Conversion can also occur if bank ‘fails’ a regulatory stress test (like U.S. SCAP or the European CEBS stress tests)|
|Systemic triggers||Based on ‘systemic’/broad market factors||✓||Increases capital across the banking system at the same time, and thus, may be most efficient at addressing and reducing systemic risk||✓|
|Virtually no incentive for bank management to take specific actions as trigger conditions are removed from direct control|
Might be too broad and costly ex ante unless trigger conditions are set sufficiently low
|1. Predetermined general conditions of financial sector (loss rates, capitalization, cash capital ratio)||✓|
|Intuitively attractive: increases banking system capitalization in response to systemic credit losses|
Automatic, no reliance on regulatory judgment and supervisory discretion (and does not breed entitlement for intervention)
|Lack of differentiation amongst banks may have unintended consequences|
Lack of supervisory interference removes judgment in situations with little or no precedent
Possibility that trigger is too narrow, i.e., systemic risk is caused by something unrelated to credit losses (and re-capitalization addresses only symptom not cause)
|✓||Credit loss trigger: conversion occurs if, say, residential or commercial mortgage delinquencies nationwide rise above a certain level|
|2. Supervisory discretion or supervisory declaration of a systemic crisis||✓||Broad-based recapitalization of the banking system when regulators want/need it||✓|
|Lack of differentiation amongst firms may have unintended consequences and makes recapitalization inefficient|
Strong reliance on regulatory judgment and supervisory discretion might be counterproductive if negative signaling effect encourages forbearance
Large-scale consequences might delay supervisory action
Lower marketability: increases funding costs since investor may charge a premium for the uncertainty associated with trigger conditions
|Dual triggers (bank specific + systemic)||Declaration of a systemic crisis and the realization of a single idiosyncratic trigger||✓||Broad-based recapitalization of the banking system while allowing for differentiation amongst banks||✓|
|Prone to produce mixed signals: pits regulatory judgment against market-perceived severity of systemic distress|
Risks combining the worst characteristics of triggers: strong reliance on supervisory discretion and lagging indicator subject to national variations
|✓||French et al (2010): declaration by regulators that the financial system is suffering from a systemic crisis|
|Relative to contingent capital holders(Conversion into fixed value of equity)(e.g.”reverse convertibles”)||Converts upon trigger breach into a variable number of shares based on the share price at the time of the trigger breach||✓|
|Provides clarity for holders of contingent capital (valuation certainty)|
Cost-effective issuance cost of contingent capital
|Possibility that conversion will be ineffective due to insufficient capital buffer|
Likely to increase the cost of and decrease access to new equity unless share amount is limited
|1. Conversion at par value||✓|
|Owners of contingent capital securities benefit from a negative shock to the share price (if it also increases the likelihood of conversion) (D)|
Lower issuance cost of contingent capital (D)
|Creates incentives for market manipulation (e.g., short-selling of stock), which could be mitigated by using some historical market-based trigger for conversion (e.g., the average stock price over a longer period) (D)|
High (and unknown) dilution to shareholders likely to create strong incentives to avoid trigger breach to a point when conversion might be too late.
Potentially large dilution risk since contingent capital is likely to trade at a discount for any reasonable trigger point (E)
Higher cost of equity capital due to potential dilution and dilution uncertainty, depending on trigger conditions relative to trading price of contingent capital (E)
Higher cost of equity but lower issuance cost (E/D)
|2. Conversion below par value||✓||Higher firm value due to reduced liabilities lowers the possibility of a lucrative speculative attack (E)||✓|
|Existing shareholders have incentive to trigger conversion prematurely (esp. in combination with market-based triggers) in order to reduce dilution risk (especially if debt does not trade at a discount) (E) |
Smaller dilution risk (than in the case of conversion at par) could fail to deter ex ante risk-taking by managers
Lower cost of equity but higher issuance cost (E/D)
|3. Conversion at/above/below trading price of contingent capital at time of conversion||✓||Lower dilution risk since conversion rate is pegged to a trading price that is likely to be discounted to principal (E)||✓|
|Higher overall cost of capital due to higher uncertainty for both holders of contingent capital and shareholders|
Creates incentives for market manipulation (e.g., short-selling of stock) to reduce debt payments when bank operates close to the default barrier, debt-equity correlation is high, and dilution risk is low (E)
High issuance cost of contingent capital (D)
|4. No conversion but principal write-down||✓||Most cost-efficient form of contingent capital (D) ✓No dilution risk (E)||✓||Creates incentives for market manipulation (e.g., short-selling of stock) to reduce debt payments|
|Relative to shareholders (Conversion into a fixed number of shares) (‘mandatory convertibles”)||Converts upon trigger breach or a certain date (if earlier) into a finite number of shares based on the share price at the time of original issuance||✓|
|No incentive for manipulation: pre-specified, fixed amount of dilution for shareholders; managers consider conversion only if the degree of dilution is smaller than the debt payments to holders of contingent capital (E)|
Encourages both parties to avoid trigger breach; conversion optimal only if share price is sufficiently low when bank operates close to its default barrier (as long as trigger conditions are idiosyncratic and not completely insulated from changes in the share price) (E/D)
|✓||Cost of funding will increase proportionately to the chance of the bank facing difficult circumstances||✓||Lloyds exchange: The conversion rate (as a number of shares) is determined by dividing the par value of the securities by the share price at the time of issuance.|
|Relative to issuer’s capital needs||Converts upon trigger breach into a variable number of shares based on the share price and the capital shortfall at the time of the trigger breach||✓|
|No clarity for holders of contingent capital (valuation uncertainty) (D)|
Conversion effectiveness, i.e., sufficient capital buffer creation upon conversion (assuming timely trigger and conversion formulation)
Greater incentive compatibility
|Higher overall cost of capital due to higher uncertainty for both holders of contingent capital and shareholders|
High issuance cost of contingent capital (D)
|✓||Conversion rate determined by book value multiple/ RWA multiple|
Advantages/disadvantages to shareholders (E) and holders of contingent capital (D).
Advantages/disadvantages to shareholders (E) and holders of contingent capital (D).
Similar to other hybrid capital instruments, contingent capital entails both equity and debtlike features. Bank regulators may allow these instruments to be treated as part of capital due to their loss-absorbing characteristics. They may be cheaper for issuers (thanks to tax deductibility) and shareholder dilution only occurs if conversion is triggered. Fixed-income investors may be attracted by the higher yields of these instruments.
The hybrid capital instruments did not provide meaningful loss absorption during the recent crisis. While they contractually entailed loss-absorbing clauses, primarily through the deferral of coupons and/or extension of maturity, this remained at the discretion of banks, unless regulatory capital ratios were breached. In most cases, governments’ rescue of banks meant that liquidation was averted and, thus, subordinated debt holders did not suffer liquidation-related losses either. At least in the initial stages of the financial crisis, the deferral of coupons/maturity extension was not the preferred route for issuers and regulators alike, as they feared negative signaling effects. However, as the crisis amplified, some regulators became more forceful in imposing some losses on subordinated creditors, and the European Commission introduced the concept of “burden sharing” between taxpayers, shareholders, and bondholders through the imposed suspension of dividends and coupons to preserve cash and capital in distressed banks.
However, by buying back bonds at heavy discounts to par, issuers managed to reduce their debt-servicing obligations and created capital gains that augmented their core capital. While this was beneficial for issuers, it changed the traditional priority of claims, as bondholders had to take permanent losses ahead of shareholders.
In response to the crisis, the BCBS strengthened the standards by increasing the quality, consistency, and transparency of the capital base. Higher minimum capital requirements were established for common equity from 2 percent to 4.5 percent, Tier 1 (from 4 percent to 6 percent), while maintaining capital ratio at 8 percent. In addition, the BCBS introduced additional capital buffers. All new Tier 1 and Tier 2 instruments will have to be loss-absorbing to be eligible as regulatory capital.
Accordingly, most existing hybrid and subordinated debt instruments will not qualify as capital after January 1, 2013. Innovative capital, such as instruments with an incentive to redeem like a step-up at the call date, will no longer be included in the calculation of Tier 1 capital. Other instruments that do not meet the new criteria for inclusion in Tier 1 and Tier 2 capital will be phased out by 2022.
Tier 1, Tier 2, and contingent capital will have to entail equity conversion clauses or permanent write-offs (Appendix Table 3). The order of loss absorption for these instruments is not yet clear (for example, whether conversion of contingent capital should occur prior to or after the deferral of hybrid Tier 1 coupons and hybrid Tier 1 conversion into equity). Regulatory clarification will be necessary to facilitate the pricing of these different instruments under the new Basel III capital regime.
|Hybrid Tier 1 with Conversion or Write-Down||Tier 2 With Conversion or Write-Down||High-Level Trigger Contingent Capital (>7%)||Low-Level Trigger Contingent Capital (5%)|
|Rank as||Tier 1||Tier 2||Possibly Tier 2 (pre-conversion)||Possibly Tier 2 (pre-conversion)|
|Status of missed coupons||Noncumulative (cancelled)||Not applicable||Not applicable||Not applicable|
|Loss absorption in gone concern 1/||Yes||Yes||No||No|
|&||Permanent writedown of principal||✓||✓|
|&||Conversion into equity||✓||✓|
|Loss absorption in going concern||Yes||No||Yes if Core Tier 1 below a||Yes if Core Tier 1 below a|
|&||Write-down of principal||Temporary||pre-determined level (e.g. 7%)||pre-determined level (e.g. 5%)|
|&||Conversion into equity||No||Same||Same|
|&||Mandatory cancellation of coupons||✓||No||No|
Gone concern at the point of nonviability.
Gone concern at the point of nonviability.
The investor base for contingent capital may be different from the traditional fixed-income investor base. Traditionally, the natural investor base for hybrid Tier 1 and Tier 2 securities was fixed-income investors. Going forward, the conversion into equity or permanent write-down features of contingent capital and other Basel III capital instruments may deter those investors. A new investor base may gradually emerge for CoCos, starting with hedge funds, high yields, or equity investors. Mapping investors yield targets and future Basel III instruments expected yields highlight the possible shift that could happen in the investor base (Figure 3).
Figure 3.Indicative Investors’ Yield Targets and Select Capital Instruments Yields in Europe
If a sample of 25 global SIFIs were to issue CoCo bonds to cover up to 2.5 percent RWA for the countercyclical buffer, this would represent $392 billion. If the SIFI additional loss-absorbing capital requirements (which remain to be defined) were to be established at 2 percent of RWA, for instance, the same sample could use CoCo (or equity, of course) to the extent of $314 billion. In addition, banks will have to replace existing Tier 1 and Tier 2 instruments that will be gradually phased out under Basel III, which will reduce their available non-equity Tier 1 and Tier 2 capital, and increase their needs to tap capital markets with loss-absorbing instruments. With over $1 trillion of outstanding subordinated and hybrid debt that will be phased out by 2022, banks may have to issue loss-absorbing instruments of this magnitude in coming years.13
AdmatiA.P.DeMarzoM.Hellwig and P.Pfleiderer2010“Fallacies Irrelevant Facts and Myths in the Discussion of Capital Regulation: Why Bank Equity is Not Expensive” Stanford Graduate School of Business Research Paper No. 2065R.
AdrianT. and H.Shin2010“The Changing Nature of Financial Intermediation and the Financial Crisis of 2007-2009,”Annual Review of Economics No.2pp. 603–618.
Association for Financial Markets in Europe (AFME) 2010“The Systemic Safety Net: Pulling Failing Firms Back From the Edge”August.
Bank of England2010Financial Stability Report No. 28December.
Basel Committee on Banking Supervision (BCBS)2010a“Basel III: A Global Regulatory Framework for More Resilient Banks and Banking Systems”December.
Basel Committee on Banking Supervision (BCBS)2010b“Proposal to Ensure the Loss Absorbency of Regulatory Capital at the Point of Nonviability” BCBS Consultative DocumentAugust.
Basel Committee on Banking Supervision (BCBS)2003“Markets for Bank Subordinated Debt and Equity in Basel Committee Member Countries,”BCBS Working Paper No. 12August.
BoltonPatrick and F.Samama2010“Contingent Capital and Long Term Investors: A Natural Match?”(mimeo) Columbia Business School.
BrennanM.2010“Bank Hybrid Capital Criteria: Methodology and Assumptions” Standard & Poor’sDecember.
BrennanM. and B.de Longevialle2010“Potential $1 Trillion Bank Contingent Capital-Style Issuance Faces Uncertain Investor Interest” Standard & Poor’sDecember.
BrunnermeierM. K.2009“Deciphering the Liquidity and Credit Crunch 2007-08,”Journal of Economic PerspectivesVol. 23 No. 1pp. 77–100.
ChanceClifford2009“Contingent Capital—The New Hybrid Capital?” Client briefingNovember.
De MartinoG.M.LibertucciM.Marangoni and M.Quagliariello2010“Countercyclical Contingent Capital (CCC): Possible Use and Ideal Design,”Banca d’Italia Occasional Paper (Forthcoming).
DuffieDarrell2010“A Contractual Approach to Restructuring Financial Institutions,” in Ending Government Bailouts as We Know Themedited byG.SchultzK.Scott and J.TaylorHoover Institute PressStanford University.
FlanneryMark2010“Stabilizing Large Financial Institutions with Contingent Capital Certificates,”Carefin Working Paper No. 4.
FlanneryMark2009“Market-Valued Triggers Will Work for Contingent Capital Instruments” Solicited Submission to U.S. Treasury Working Group on Bank CapitalNovember.
FrenchK.M.BailyJ.CampbellJ.CochraneD.DiamondD.DuffieA.KashyapF.MishkinR.RajanD.ScharfsteinR.ShillerH. S.ShinM.SlaughterJ.Stein and R.Stulz2010The Squam Lake Report: Fixing the Financial SystemPrinceton University Press.
GoodhartC.2010“Are CoCos from Cloud Cuckoo-Land?” VoxEU.
HenriquesRobertoChristianLeukers and EnricoLongato2010a“Applying the Swiss Finish”J. P.MorganOctober 28.
HenriquesRoberto2010b“European Bank Bail-In Survey Results”J. P.MorganEurope Credit ResearchOctober 14.
HenriquesRoberto2010c“The Future of Hybrid Bank Capital?”J. P.MorganEurope Credit ResearchOctober 6.
InekeJ.L.StreetF.Simpson and N.Blackman2010“European Banks Tier 1—A new paradigm,”Morgan Stanley ResearchNovember 4.
InekeJ.L.Street and F.Simpson2010“European banks: €49 billion of Swiss CoCos” Morgan Stanley ResearchOctober 4.
International Monetary Fund2010a“Resolution of Cross-Border Banks—A Proposed Framework for Enhanced Coordination”June.
International Monetary Fund2010bGlobal Financial Stability ReportApril.
LandierA. and K.Ueda2009“The Economics of Bank Restructuring: Understanding the Options” IMF Staff Position Note SPN/09/12International Monetary FundJune.
Senior Supervisors Group2009“Risk Management Lessons from the Global Banking Crisis of 2008”October.
SundaresanS. and Z.Wang2010“Designing of Contingent Capital with a Stock Price Trigger for Mandatory Conversion” Federal Reserve Bank of New York Staff Paper No. 448.
TiroleJean2006Theory of Corporate FinancePrinceton and Oxford: Princeton University Press.
ViñalsJ.J.FiechterC.PazarbasiogluL.KodresA.Narain and M.Moretti2010“Shaping the New Financial System” IMF Staff Position Note SPN/10/15OctoberInternational Monetary Fund.