PrefBlog

A question in the comments to my old post A Structural Model of Contingent Bank Capital led me to look up what Prof. George Pennacchi has been doing lately; together with Theo Vermaelen and Christian C. P. Wolff he has written a paper titled Contingent Capital: The Case for COERCs:

In this paper we propose a new security, the Call Option Enhanced Reverse Convertible (COERC). The security is a form of contingent capital, i.e. a bond that converts to equity when the market value of equity or capital falls below a certain trigger. The conversion price is set significantly below the trigger price and, at the same time, equity holders have the option to buy back the shares from the bondholders at the conversion price. Compared to other forms of contingent capital proposed in the literature, the COERC is less risky in a world where bank assets can experience sudden, large declines in value. Moreover, the structure eliminates concerns of an equity price “death spiral” as a result of manipulation or panic. A bank that issues COERCs also has a smaller incentive to choose investments that are subject to large losses. Furthermore, COERCs reduce the problem of “debt overhang,” the disincentive to replenish shareholders’ equity following a decline.

The basic justification for the COERCs is:

In contrast to the Credit Suisse coco bond [with accounting and regulatory triggers], the trigger is based on market value based leverage ratios, which are forward looking, rather than backward looking, measures of financial distress. It also means that at the time of the triggering event the stock price is known, unlike in the case of coco bonds with accounting based capital ratio triggers. As the trigger is driven by the market and not by regulators, regulatory risk is avoided. The conversion price is set at a large discount from the market price at the time of conversion, which means that conversion would generate massive shareholder dilution. However, in order to prevent this dilution, shareholders have an option to buy back the shares from the bondholders at the conversion price. In practice, what will happen is that when the trigger is reached, the company will announce a rights issue with an issue price equal to the conversion price and use the proceeds to repay the debt. As a result, the debt will be (almost) risk-free. In our simulations, we show that it is possible to design a COERC in such a way that the fair credit spread is 20 basis points above the risk-free rate. So although the shareholders are coerced to repay the debt, the benefit from this coercion is reflected in the low cost of debt as well as the elimination of all direct and indirect costs of financial distress. Although at the time of the trigger, the company will announce an equity issue, there is no negative signal associated with the issuance as the issue is the automatic result of reaching a pre-defined trigger.

Market based triggers are generally criticised because they create instability: bond holders have an incentive to short the stock and trigger conversion. Moreover, the fear of dilution may encourage shareholders to sell their shares so that the company ends up in a self-fulfilling death spiral. However, because in a COERC shareholders have pre-emptive rights in buying the shares from the bondholders, they can undo any conversion that is result of manipulation or unjustified panic. Moreover, because bondholders will generally be repaid, they have no incentive to hedge their investment by shorting the stock when the leverage ratio approaches the trigger, unlike the case of coco bonds where bondholders will become shareholders after the triggering event. The design of the contract also discourages manipulation by other bondholders. Bolton and Samama (2010) argue that other bond-holders may want to short the stock to trigger conversion, in order to improve their seniority. However, because the COERCs will be repaid in these circumstances such activity will not improve other bondholder’s seniority.

Further justification is given as:

Our objective is to propose an alternative, an instrument that a value maximizing manager would like to issue, without being forced by regulators. Companies are coerced to issue equity and repay debt by fear of dilution, not by the decision of a regulator. Imposing regulation against the interest of the bank’s shareholders will encourage regulatory arbitrage and may also reduce economic growth.6 If bankers, on the other hand, can be convinced that issuing contingent capital increases shareholder value, then any regulatory “encouragement” to issue these securities will be welcomed. Our proposal is therefore more consistent with a free market solution to the general problem that debt overhang discourages firms from recapitalizing when they are in financial distress. Hence the COERC should be of interest to any corporation where costs of financial distress are potentially important.

It seems like a very good idea. One factor not considered in the paper is the impact on equity investors.

Say you have an equity holding in a bank that has a stock price (and the fair value of the stock price) slightly in excess of the trigger price for its COERCs. At that point, buyers of the stock (and continuing holders!) must account for the probability that the conversion will be triggered and their will be a rights issue. Therefore, in order to avoid dilution, they must not only pay the fair market value for the stock, but they must also have cash on hand (or credit lines) available that will allow them to subscribe to the rights offering; the necessity of having this excess cash will make the common less attractive at its fair market value. This may serve to accelerate declines in the bank’s stock price.

It is also by no means assured that shareholders will be able to sell the rights anything close to their fair value.

A Goldman Sachs research report titled Contingent capital Possibilities, problems and opportunities is also of interest. Canadians panic-stricken by the recent musings in the federal budget (see discussion on April 1, April 2 and April 5) will be fascinated by:

Bail-in is a potential resolution tool designed to protect taxpayer funds by converting unsecured debt into equity at the point of insolvency. Most bail-in proposals would give regulators discretion to decide whether and when to convert the debt, as well as how much.

There is an active discussion under way as to whether bail-in should be a tool broadly applicable to all forms of unsecured credit (including senior debt) or whether it should be a specific security with an embedded write-down feature.

Naturally, this discussion is not being held in Canada; we’re too stupid to be allowed to participate in intelligent discussions.

As might be expected, GS is in favour of market-based solutions and consequent ‘high-trigger’ contingent capital:

Going-concern contingent capital differs substantially from the gone-concern kind. It is designed to operate well before resolution mechanisms come into play, and thus to contain financial distress at an early stage. The recapitalization occurs at a time when there is still significant enterprise value, and is “triggered” through a more objective process with far less scope for regulatory discretion. For investors to view objective triggers as credible, however, better and more-standardized bank disclosures will be needed on a regular basis. Because this type of contingent capital triggers early, when losses are still limited, it can be issued in smaller tranches. This, in turn, allows for greater flexibility in
structuring its terms.

When the early recapitalization occurs, control of the firm can shift from existing shareholders to the contingent capital holders, and a change in management may occur. The threat of the loss of control helps to strengthen market discipline by spurring the firm to de-risk and de-leverage as problems begin to emerge. As such, going-concern contingent capital can be an effective risk-mitigating tool.

GS further emphasizes the need to appeal to fixed income investors:

Contingent capital will only be viable as a large market if it is treated as debt

Whether “going” or “gone,” contingent capital will only be viable as a large market if it is treated as debt. This is not just a question of technical issues like ratings, inclusion in indices, fixed income fund mandates and tax-deductibility, though these issues are important. More fundamentally, contingent capital must be debt in order to appeal to traditional fixed income investors, the one market large enough to absorb at least $925 billion in potential issuance over the next decade.

Surprisingly, GS is in favour of capital-based triggers despite the problems:

A capital-based trigger would force mandatory conversion if and when Tier 1 (core) capital fell below a threshold specified either by regulators (in advance) or in the contractual terms of the contingent bonds. We think this would likely be the most effective trigger, because it is transparent and objective. Investors would be able to assess and model the likelihood of conversion if banks’ disclosure and transparency are enhanced. Critically, a capital-based trigger removes the uncertainty around regulatory discretion and the vulnerability to market manipulation that the other options entail.
…
Capital-based triggers are also vulnerable to financial reporting that fails to accurately reflect the underlying health of the firm. Lehman Brothers, for example, reported a Tier 1 capital ratio of 11% in the period before its demise – well above the regulatory minimum and a level most would have considered healthy. The same was true for Bear Stearns and Washington Mutual before they were acquired under distress. We think this issue must be resolved for investors to embrace capital-based triggers.

Fortunately there are several ways to make capital ratios more robust, whether by “stressing” them through regulator-led stress tests or by enforcing more rigorous and standardized disclosure requirements that would allow investors to better assess the health of the bank. Such standardized disclosures could relieve regulators of the burden of conducting regular stress tests, and would significantly enhance transparency. The value of stress testing and greater disclosures is one lesson from the financial crisis. The US Treasury’s 2009 stress test illustrates this point vividly. While not perfect, it offered greater
transparency and comparability of bank balance sheets than investors were able to derive from public filings. With this reassurance, investors were willing to step forward and commit capital. The European stress test proves the point as well: it did not significantly improve transparency and thus failed to reassure investors or attract capital.

That is the crux of the matter and I do not believe that the Gordian Knot can be cut in the real world. The US Treasury made their stress test strict and credible because it knew in advance that its banks would pass. The Europeans made their stress test ridiculous and incredible because they knew in advance that their banks would fail.

I liked their succinct dismissal of regulatory triggers:

While flexibility can be helpful, particularly given that no two crises are alike, recent experience shows that some regulators may be hesitant to publicly pronounce that a financial firm is unhealthy, especially during the early stages of distress. There is, after all, always the hope that the firm’s problems will be short-lived, or that an alternative solution to the triggering of contingent capital can be found. Thus a regulator may be unlikely to pull the trigger – affecting not only the firm and all of its stakeholders, but also likely raising alarm about the health of other financial firms – unless it is certain of a high degree of distress. By then, losses may have already risen to untenable levels, which is why this type of trigger is associated with gone-concern contingent capital.

GS emphasizes the importance of the indices:

The inclusion of contingent capital securities in credit indices will also be an important factor, perhaps even more important than achieving a rating. This is because the inclusion itself would attract investors, who otherwise might risk underperforming benchmarks by being underweight a significant component of the index. Credit indices currently do not include mandatorily convertible equity securities, although they can include instruments that allow for loss absorption through a write-down feature. This again contributes to the appeal of the write-down feature (rather than the simple conversion to equity) to most fixed income investors. If contingent capital securities were included in credit indices, this addition would be likely to drive a substantially deeper contingent capital market.

Here in Canada, of course, the usual benchmark is prepared by the TMX, which the regulators allowed to become bank-owned on condition that it improved the employment prospects for regulators. It’s a thoroughly disgraceful system which will blow up in all our faces some days and then everybody will pretend to be surprised.

GS is dismissive of regulatory triggers and NVCC:

A discretionary, “point of non-viability” trigger would likely be attractive to many regulators as it helps them to preserve maximum flexibility in the event of a financial crisis. This can be useful given that no two crises are exactly alike. It could also allow regulators to consider multiple factors – including the state of the overall financial system – when making the decision to pull the trigger. Discretion also gives regulators the opportunity to exercise regulatory forbearance away from the public spotlight.

Yet we believe this preference for discretion and flexibility makes it difficult for regulators to meet one of their most important – yet mostly unspoken – goals, which is to develop a viable contingent capital market. Regulators have certainly solicited feedback from investors, but some seem to believe that simply making contingent capital mandatory for issuers means that investors will buy them. However, from conversations with many investors, we believe that regulators may need to move toward a more objective trigger; if not, the price of these instruments may be prohibitive.

There is another set of participants in a potential contingent capital market: taxpayers. Regulators represent taxpayers’ interests by promoting systemic stability and requiring robust loss-absorption capabilities at individual banks. But the interests of regulators and taxpayers may not always be fully aligned. If taxpayers’ principal goal is to avoid socializing private-sector losses, and to prevent the dislocation of a systemic crisis even in its early stages, then they should want a stringent version of contingent capital – one that converts to equity at a highly dilutive rate, based on an early and objective trigger. The discretion and flexibility inherent in regulatory-triggered gone-concern contingent capital may have less appeal to taxpayers. From their standpoint, gone-concern contingent capital might well have allowed a major financial firm to fail, causing job losses and other disruptions across the financial system. Taxpayers may find the potential risk-reducing incentives created by going-concern contingent capital to be a more robust answer to the problem of too big to fail.

Goldman’s musings on investor preference can be taken as an argument in favour of COERCs:

Traditional fixed-income investors will likely want contingent capital to have a very low probability of triggering, which leads them to prefer an objective, capital-based and disclosure-enhanced trigger. Many investors have indicated their concerns about the challenges of modeling a discretionary trigger: it is very difficult to model the probability of default, the potential loss given default or even the appropriate price to pay for a security that converts under a discretionary and opaque process. Greater transparency is a prerequisite for a capital-based trigger to be seen as credible by investors, because they will need to have greater confidence that banks’ balance sheets reflect reality. We also believe that investors would be more likely to embrace a capital-based trigger if the terms were quite stringent, thereby lowering the probability of conversion.

Under pressure from the Swiss bank regulator Credit Suisse is issuing High-Trigger CoCos:

Credit Suisse today announced a number of measures to accelerate the strengthening of its capital position in light of the current regulatory and market environment. An immediate set of actions will be implemented to increase the capital by CHF 8.7 billion. Additional capital actions and earnings related impacts are to increase the capital by a further CHF 6.6 billion by year-end 2012.

The measures will result in an expected end-2012 look-through Swiss Core Capital Ratio of 9.4%, compared to the 2018 requirement of 10%. Look-through Swiss Core Capital includes look-through Basel III Common Equity Tier 1 (CET1) and existing participation securities (“Claudius notes”) that qualify as part of the Swiss equity requirement in excess of the 8.5% Basel III G-SIB Common Equity Tier 1 (CET1) ratio.

The measures will result in an expected look-through Swiss Total Capital Ratio of 10.8% at end 2012. This broadly compares to the figure of 5.9% calculated by the Swiss National Bank (SNB) at the end of 1Q12 and published in its 2012 Financial Stability Report. Look-through Swiss Total Capital includes look-through Basel III CET1 and the participation securities (“Claudius notes”). Additionally it includes the Group’s Buffer Capital Notes (“CoCos with high trigger”).

There are no details available on the projected notes, but they have some Tier 2 Buffer Capital Notes outstanding.

For example, there is a USD 2-billion issue of 7.875 per cent. Tier 2 Buffer Capital Notes due 2041:

Interest on the BCNs will accrue from and including 24 February 2011 (the ‘‘Issue Date’’) to (but excluding) 24 August 2016 (the ‘‘First Optional Redemption Date’’) at an initial rate of 7.875 per cent. per annum, and thereafter at a rate, to be reset every five years thereafter, based on the Mid Market Swap Rate (as defined herein) plus 5.22 per cent.
…
If a Contingency Event or a Viability Event (each as defined herein) occurs, the BCNs shall, subject to the satisfaction of certain conditions, mandatorily convert into Ordinary Shares (as defined herein) which shall be delivered to the Settlement Shares Depository (as defined herein) on behalf of the Holders, as more particularly described in ‘‘Terms and Conditions of the BCNs—Conversion’’. In the event of a Contingency Event Conversion (as defined herein), such Ordinary Shares may, at the election of CSG, be offered for sale in a Settlement Shares Offer as described herein.
…
Contingency Event means that CSG has given notice to the Holders that CSG’s Core Tier 1 Ratio (prior to the Basel III Regulations Date) or the Common Equity Tier 1 Ratio (on or after the Basel III Regulations Date) is below 7 per cent. as at the date of the financial statements contained in a Quarterly Financial Report and that a Contingency Event Conversion will take place.
…
Viability Event means that either: (a) the Regulator has notified CSG that it has determined that Conversion of the BCNs, together with the conversion or write off of holders’ claims in respect of any other Buffer Capital Instruments, Tier 1 Instruments and Tier 2 Instruments that, pursuant to their terms or by operation of laws are capable of being converted into equity or written off at that time, is, because customary measures to improve CSG’s capital adequacy are at the time inadequate or unfeasible, an essential requirement to prevent CSG from becoming insolvent, bankrupt or unable to pay a material part of its debts as they fall due, or from ceasing to carry on its business; or (b) customary measures to improve CSG’s capital adequacy being at the time inadequate or unfeasible, CSG has received an irrevocable commitment of extraordinary support from the Public Sector (beyond customary transactions and arrangements in the ordinary course) that has, or imminently will have, the effect of improving CSG’s capital adequacy and, without which, in the determination of the Regulator, CSG would have become insolvent, bankrupt, unable to pay a material part of its debts as they fall due or unable to carry on its business.
…
The BCNs will be converted into a number of Ordinary Shares determined by dividing the principal amount of each BCN by the Conversion Price in effect on the relevant Conversion Date. ‘‘Conversion Price’’ means (i) at any time when the Ordinary Shares are admitted to trading on a Recognised Stock Exchange, in respect of any Conversion Date, the greatest of (a) the Reference Market Price of an Ordinary Share on the fifth Zurich Business Day prior to the date of the relevant Contingency Event Notice or, as the case may be, the Viability Event Notice translated into United States dollars at the Exchange Rate, (b) the Floor Price on the fifth Zurich Business Day prior to the date of the Contingency Event Notice or, as the case may be, the Viability Event Notice; and (c) the nominal value of each Ordinary Share on the Share Creation Date (being, at the Issue Date, CHF 0.04) translated into United States dollars at the Adjusted Exchange Rate, or (ii) without prejudice to ‘‘Takeover Event and De-listing’’ below, at any time when the Ordinary Shares are not admitted to trading on a Recognised Stock Exchange by reason of a Non-Qualifying Takeover Event or otherwise, the greater of (b) and (c) above.

Very good. There’s a high trigger and conversion at market price. The part I dislike is that the conversion trigger is a regulatory ratio – we found during the crisis that regulatory ratios aren’t worth much in the course of a panic. Still – much better than anything we’re ever likely to see in Canada!

Following the finalization of the OSFI NVCC Advisory and basically simultaneously with the Review-Negative slapped on NVCC-eligible extant preferreds, DBRS has announced that it:

has concluded today that it expects it can rate Canadian subordinated debt with a non-viability contingent capital clause (sub debt NVCC) and Canadian preferred shares with a non-viability contingent capital clause (pref NVCC) following the review of the Office of the Superintendent of Financial Institutions Canada (OSFI) Advisory on Non-Viability Contingent Capital, issued on August 16, 2011 (NVCC Advisory).

In this document, all references to non-viability contingent capital (NVCC) instruments are based on our expectations that non-viability (as determined by OSFI) is the only contingent event, that the contingent event triggers permanent conversion to common equity and that over time, as NVCC becomes the major instrument with respect to subordinated debt and preferred shares, any trigger event for sub debt NVCC holders would cause these holders to become meaningful owners of the bank in question. These considerations are also consistent with our ability to rate the NVCC instruments. According to DBRS criteria, the triggers are well defined and permit an assessment of the risks.

Both the sub debt NVCC and pref NVCC ratings will have wider notching, based on the global standard notching for preferred shares, because of additional risk associated with tripping the trigger. The expected losses resulting from tripping the trigger would have an impact on the relative rating of sub debt NVCC and pref NVCC. As guidance, sub debt NVCC will likely be rated no higher than the standard rating for preferred shares and the pref NVCC will likely be rated one notch below the standard rating for preferred shares.
…
For clarity, global standard notching for preferred shares means the starting point for notching preferred share ratings is the intrinsic assessment (IA) rating rather than the final senior debt rating, and the degree of notching from the IA rating to the preferred share rating widens to reflect our perception of the increased risk in these capital instruments. The base notching policy is three notches for AA, four notches for “A” and five notches for BBB and lower IA ratings. Note that when DBRS initiated the criteria on June 29, 2009, most banks in Canada had their preferred share ratings downgraded to only one notch above the global standard notching for preferred shares
…
DBRS has determined that the likelihood of tripping the trigger event (i.e., non-viability as determined by OSFI) would be very hard or remote. DBRS’s decision was based on the assessment of the criteria to be considered in triggering conversion of NVCC instruments that was spelled out in the NVCC advisory by OSFI. Lower-rated banks suggest an increased probability of conversion as a result of tripping the trigger given the greater need for a bank to generate regulatory capital. This would result in higher notching from the intrinsic assessment, as set out in the DBRS methodology Rating Bank Preferred Shares and Equivalent Hybrids. As such, both the sub debt NVCC and pref NVCC ratings would be tied to the preferred share rating of the bank.

The expected losses as a result of the conversion would affect the rating for sub debt NVCC relative to pref NVCC. It is the economic entitlement each receives post-trigger that is the significant factor in the relative ratings as opposed to the host security’s pre-trigger features. This economic entitlement can be assessed only after terms are provided in a contractual agreement between the issuing bank and the purchaser.

OSFI has released a final Advisory on NVCC, with some changes from the draft advisory which was discussed on PrefBlog. The draft advisory has been removed from OSFI’s website in accordance with their policy to ensure that the rationale behind their policies and their development is not understood by investors. Naturally, no comment letters have been published, nor have any documents been referenced that might provide any vestiege of support for their arbitrary and capricious rule-making.

The final advisory begins with a non-sequiter that would not be tolerated in Grade 4:

All regulatory capital must be able to absorb losses in a failed financial institution. During the recent crisis, however, this premise was challenged as certain non-common Tier 1 and Tier 2 capital instruments did not absorb losses for a number of foreign financial institutions that would have failed in the absence of government support.

Principle 3 from the draft advisory, giving the Superintendent the right to trigger conversion if she feels like it, with no appeal, has been retained. Banks are urged to hire lots of former OSFI employees.

There is now a requirement that there be a floor on the conversion price – this did not exist before:

Principle # 4: The conversion terms of new NVCC instruments must reference the market value of common equity on or before the date of the trigger event. The conversion method must also include a limit or cap on the number of shares issued upon a trigger event.

On the one hand, this will prevent so-call “death spirals”. On the other hand, it may make NVCC instruments harder to issue during times of crisis. The necessity of such an unprincipled principle is necessary due to OSFI’s insistence on “low-trigger” NVCC, at a time when the rest of the world has determined that “high-trigger” NVCC is the way to go (see, for example, statements by officials of S&P, more from S&P, Switzerland, the UK, respected academics, and other respected academics, and an equivocal view from IMF staff).

Principal #8, which throws contract law into the same garbage bin as bankruptcy law, has been retained:

Principle # 8: The terms of the NVCC instrument should include provisions to address NVCC investors that are prohibited, pursuant to the legislation governing the DTI, from acquiring common shares in the DTI upon a trigger event. Such mechanisms should allow such capital providers to comply with legal prohibitions while continuing to receive the economic results of common share ownership and should allow such persons to transfer their entitlements to a person that is permitted to own shares in the DTI and allow such transferee to thereafter receive direct share ownership.

Section 2 seeks to ensure permanent employment and many future job opportunities for OSFI employees:

Section 2: Information Requirements to Confirm Quality of NVCC Instruments

While not mandatory, DTIs are strongly encouraged to seek confirmations of capital quality from OSFI’s Capital Division prior to issuing NVCC instruments11. In conjunction with such requests, the DTI is expected to provide the following information….

The Bank for International Settlements has announced:

the Group of Governors and Heads of Supervision (GHOS), the oversight body of the Basel Committee on Banking Supervision (BCBS), agreed on a consultative document setting out measures for global systemically important banks (G-SIBs). These measures include the methodology for assessing systemic importance, the additional required capital and the arrangements by which they will be phased in. These measures will strengthen the resilience of G-SIBs and create strong incentives for them to reduce their systemic importance over time.

The GHOS is submitting this consultative document to the Financial Stability Board (FSB), which is coordinating the overall set of measures to reduce the moral hazard posed by global systemically important financial institutions. This package of measures will be issued for consultation around the end of July 2011.

The assessment methodology for G-SIBs is based on an indicator-based approach and comprises five broad categories: size, interconnectedness, lack of substitutability, global (cross-jurisdictional) activity and complexity.

The additional loss absorbency requirements are to be met with a progressive Common Equity Tier 1 (CET1) capital requirement ranging from 1% to 2.5%, depending on a bank’s systemic importance. To provide a disincentive for banks facing the highest charge to increase materially their global systemic importance in the future, an additional 1% surcharge would be applied in such circumstances.

The higher loss absorbency requirements will be introduced in parallel with the Basel III capital conservation and countercyclical buffers, ie between 1 January 2016 and year end 2018 becoming fully effective on 1 January 2019.

The GHOS and BCBS will continue to review contingent capital, and support the use of contingent capital to meet higher national loss absorbency requirements than the global minimum, as high-trigger contingent capital could help absorb losses on a going concern basis.

I have mixed views on this. I reported last August that the push towards surcharges was gaining ground and have been advocating surcharges based on size since (at least!) March 2009.

However, I am unfavourably disposed towards the narrow focus of the plan, which affects only “global systemically important banks” as defined by the regulators and then uses an as-yet untested formula “based on an indicator-based approach and comprises five broad categories: size, interconnectedness, lack of substitutability, global (cross-jurisdictional) activity and complexity” to assess the surcharge imposed. There’s a lot of room for error there, and a lot of room for lobbying. There’s also a lot of cliff effect: what will be the effect on the markets when a bank’s G-SIB status is changed? What if it changes during the height of a crisis? What if a well capitalized medium sized bank is interested in purchasing a failing medium sized bank during a crisis? We saw that during the crisis, a lot of the American banks bulked up – will they be willing to bid next time? And finally, of course, the subjective nature of the G-SIB status determination opens up the door for a lot of lobbying and corruption.

I would be much happier with a system that was formula-based and applied to all banks on a progressive basis.

I was very pleased to see that the committees “support the use of contingent capital to meet higher national loss absorbency requirements than the global minimum, as high-trigger contingent capital could help absorb losses on a going concern basis”. The critical part of that phrase is high-trigger contingent capital, which is really one in the eye for those morons at OSFI, who have decided that the lowest possible trigger is the best. However, the “low-trigger” policy was enacted during the reign of the Assistant Croupier; now that he has departed for a greener pastures with a company he used to regulate (see June 14), the new incumbent may have different ideas.

I use the term “McDonald CoCo” to describe a hybrid security that is initially debt-like coverts into equity when the issuer’s common equity price declines below a preset floor. The conversion is performed at the equity trigger price.

I will note immodestly that, were there any justice in the world, they would be called Hymas CoCos, since I published first, but there ain’t no justice and McDonald has the union card.

Anyway, Andrew G Haldane, Executive Director of the Bank of England, has published remarks based on a speech given at the American Economic Association, Denver, Colorado, 9 January 2011:

For large and complex banks, the number of risk categories has exploded. To illustrate, consider the position of a large, representative bank using an advanced internal set of models to calibrate capital. Its number of risk buckets has increased from around seven under Basel I to, on a conservative estimate, over 200,000 under Basel II. To determine the regulatory capital ratio of this bank, the number of calculations has risen from single figures to over 200 million. The quant and the computer have displaced the clerk and the envelope.

At one level, this is technical progress; it is the appliance of science to risk management. But there are costs. Given such complexity, it has become increasingly difficult for regulators and market participants to vouch for the accuracy of reported capital ratios. They are no longer easily verifiable or transparent. They are as much an article of faith as fact, as much art as science. This weakens both Pillars II and III. For what the market cannot observe, it is unlikely to be able to exercise discipline over. And what the regulator cannot verify, it is unlikely to be able to exercise supervision over. Banks themselves have recently begun to voice just such concerns.

… and complexity is Bad:

This evidence only provides a glimpse at the potential model error problem viewed from three different angles. Yet it suggests that model error-based confidence intervals around reported capital ratios might run to several percentage points. For a bank, that is the difference between life and death. The shift to advanced models for calibrating economic capital has not arrested this trend. More likely, it has intensified it. The quest for precision may have come at the expense of robustness.

Hayek titled his 1974 Nobel address “The Pretence of Knowledge”. In it, he highlighted the pitfalls of seeking precisely measurable answers to questions about the dynamics of complex systems. Subsequent research on complex systems has confirmed Hayek’s hunch. Policy predicated on over-precision risks catastrophic error. Complexity in risk models may have perpetuated Hayek’s pretence in the minds of risk managers and regulators.

Like, for instance, in the run-up to the height of the crisis:

To see that, consider the experience of a panel of 33 large international banks during the crisis. This panel conveniently partitions itself into banks subject to government intervention in the form of capital or guarantees (“crisis banks”)
and those free from such intervention (“no crisis banks”).

Chart 5 plots the reported Tier 1 capital ratio of these two sets of banks in the run-up to the Lehman Brothers crisis in September 2008. Two observations are striking. First, the reported capital ratios of the two sets of banks are largely indistinguishable. If anything, the crisis banks looked slightly stronger pre-crisis on regulatory solvency measures. Second, regulatory capital ratios offer, on average, little if any advance warning of impending problems. These conclusions are essentially unchanged using the Basel III definitions of capital.

Click for Big

Got any better ideas?

What could be done to strengthen the framework? As a thought experiment, consider dropping risk models and instead relying on the market. Market-based metrics of bank solvency could be based around the market rather than book value of capital. The market prices of banks are known to offer useful supplementary information to that collected by supervisors when assessing bank health.8 And there is also evidence they can offer reliable advance warnings of bank distress

To bring these thoughts to life, consider three possible alternative bank solvency ratios based on market rather than accounting measures of capital:

• Market-based capital ratio: the ratio of a bank’s market capitalisation to its total assets.
• Market-based leverage ratio: the ratio of a bank’s market capitalisation to its total debt.
• Tobin’s Q: the ratio of the market value of a bank’s equity to its book value.

The first two are essentially market-based variants of regulatory capital measures, the third a well-known corporate valuation metric. How do they fare against the first principles of complex, adaptive systems?

Click for big

Having set the stage, he starts talking about CoCos:

Alongside equity, banks would be required to issue a set of contingent convertible instruments – so-called “CoCos”. These instruments have attracted quite a bit of attention recently among academics, policymakers and bankers, though there remains uncertainty about their design. In particular, consider CoCos with the following possible design
characteristics.

• Triggers are based on market-based measures of solvency, as in Charts 6–8.
• These triggers are graduated, stretching up banks’ capital structure.
• On triggering, these claims convert from debt into equity.

Although novel in some respects, CoCos with these characteristics would be simple to understand. They would be easy to monitor in real time by regulators and investors. And they would alter potentially quite radically incentives, and thus market dynamics, ahead of banking stress becoming too acute.

He points out:

CoCos buttress market discipline and help lift the authorities from the horns of the timeconsistency dilemma. They augment regulatory discretion at the point of distress with contractual rules well ahead of distress. Capital replenishment is contractual and automatic; it is written and priced ex-ante and delivered without temptation ex-post. Because intervention would be prompt, transparent and rule-based, the scope for regulatory discretion would be constrained. For that reason, the time-consistency problem ought to be reduced, perhaps materially. A contractual belt is added to the resolution braces.

These are the most important things. As investors, we want as much certainty as possible. Contractual conversion with a preset trigger and conversion factor removes the layer of regulatory uncertainty that bedevils most other approaches.

He highlights one concern that has been of interest to the Fed, and which seems to be the thing that industry professionals focus on when I discuss this with them:

If such a structure is for the best in most states of the world, why does it not already exist? At least two legitimate concerns have been raised. First, might market-based triggers invite speculative attack by short-sellers? The concern is that CoCo holders may be able to shortsell a bank’s equity to force conversion, then using the proceeds of a CoCo conversion to cover their short position.

There are several practical ways in which the contract design of CoCos could lean against these speculative incentives. Perhaps the simplest would be to base the conversion trigger on a weighted average of equity prices over some prior interval – say, 30 days. That would require short-sellers to fund their short positions for a longer period, at a commensurately greater cost. It would also create uncertainty about whether conversion would indeed occur, given the risk of prices bouncing back and the short-seller suffering a loss. Both would act as a speculative disincentive.

A second potential firewall against speculative attack could come from imposing restrictions on the ability of short-sellers to cover their positions with the proceeds of conversion.

I like the first solution and am particularly gratified that he chose essentially the same VWAP measurement period that I chose as a basis for discussion.

I don’t like the second firewall. Stock is stock is stock. Everybody knows you can’t cheat an honest man, right? Well, you can’t manipulate a healthy stock, either. Not on the scale of a 30-day VWAP, you can’t.

A related concern is that CoCos alter the seniority structure of banks’ capital, as holders of CoCos potentially suffer a loss ahead of equity-holders. But provided the price at which CoCos convert to equity is close to the market price, conversion does not transfer value between existing equity-holders and CoCo investors. And provided conversion is into equity it need not imply investor loss. If a market move really is unjustified, prices will correct over time towards fundamentals. The holder of a converted CoCo will then garner the upside.

I don’t understand this bit. As long as the trigger/conversion price is set well below the market price at CoCo issue time (I suggested that “half” was a good figure), then CoCos will retain significant first-loss protection.

Mr Paul Tucker, Deputy Governor of the Bank of England, made a speech at the Clare Distinguished Lecture in Economics, Cambridge, 18 February 2011 titled Discussion of Lord Turner’s lecture, “Reforming finance – are we being radical enough?”:

But none of what I have said makes a case for placing all of our eggs in the resolution basket. Switching metaphors, we need belt and braces. Which is why the G20 agreed that the so-called Global Systemically Important Financial Institutions (G-SIFIs) should carry greater loss absorbing capacity (or GLAC) than implied by Basel III.

First best would be equity. Indeed, Adair has argued this evening that ideally Basel 3 would have set a higher equity requirement. But that did not happen. In practice, we are going to have to be open-minded, but also principled, about quasi-equity instruments contributing to GLAC for SIFIs (sorry about the acronyms!). Currently, the leading candidate is so-called Contingent Capital bonds (CoCos), which convert from debt into equity in certain states of the world. It seems to me that to serve the purpose of GLAC for large and complex firms, such instruments would need to convert when a firm was still fundamentally sound, which is to say that they should have high capital triggers. For a large and complex firm, a low capital trigger would be dangerous, as funders and counterparties would be likely to flee before reaching the point at which the firm would be recapitalised through the CoCos’ conversion.

Moreover, high-trigger CoCos would presumably get converted not infrequently which, in terms of reducing myopia in capital markets, would have the merit of reminding holders and issuers about risks in banking.

Lord Turner’s speech discusses a particular hobby-horse of mine:

It is therefore crucial that our answers to the SIFI problem cover also the more difficult but more likely scenario of multiple bank systemic stress. And in such conditions, bail-inable bonds will only enable us to avoid the dilemma of Autumn 2008, if the following vital conditions are met:

• • If regulators could be confident that those bonds are held outside the banking system; and
• • in addition, confident that the bonds are held by investors who have so arranged their assets and liabilities that they could face the imposed losses without that in turn inducing systemic effects.

And it may be very difficult to be confident that those conditions we met.

There are two ways to gain that confidence – the first relies on empirical observation, the second on an assumption of fully informed investor rationality. Neither route may be entirely robust.

• • The first way to seek such confidence, would be for regulators to understand, or to regulate, which investors hold bank medium-term debt. Our information on this today is imperfect. We believe a significant proportion is initially held by other banks, and a larger proportion still by a broadly defined group of ‘fund managers’. (Slide 7). But ownership after secondary market trading could be significantly different. And some of these ‘fund managers’ may be in turn financed by banks (e.g. hedge funds by prime brokers), or linked to the banking system by complex repo and derivative relationship so that losses suffered by one bank, could indirectly impose losses or confidence shocks on others. And our ability to track these complex inter-connections, and as a result to predict the knock-on consequences of initial losses in conditions of systemic fragility is imperfect today and likely to remain so. We need to improve our understanding of the complex interconnections of our financial system: but it is unclear that understanding will ever be good enough for us confidently to impose large losses simultaneously on the senior debt of multiple large banks (or indeed multiple small banks), in conditions of macro-systemic stress.
• • The other route to confidence, would be based on faith in market and investor rationality, assuming axiomatically that investors who buy bail-inable bonds will only do so on the basis of rational assessments of their ability to absorb risks in all possible future states of the world, including those of macroeconomic stress. As Section 3 will discuss, this axiomatic assumption was at the core of the pre-crisis conventional wisdom, the reason why public authorities thought they could sleep easy in the face of an explosive growth in financial scale, complexity and interconnectedness. But it relies on an assumption of fully informed rationality, which may be simply untrue, and indeed impossible. For as Andrei Shleifer et al (2010) have argued in an extremely perceptive recent paper, it may be inherent to human nature that in the good times investors systematically fail to take rational account of the tail of low probability adverse events.

A bail-inable bond will have a highly skewed probability distribution of pay-outs. (Slide 8 ) Over a long period of time, only the zero-loss segment of the distribution will be observed. A low probability of significant loss continues to exist, but Gennaioli, Shleifer and Vishay argue that that low probability will be wholly discounted through a behavioural process which they label ‘local thinking’ – the reality, deeply rooted in human nature, that not all contingencies are represented in decision makers’ thought processes. After a period of good times, investors will assume that senior bank debt is effectively risk-free: as indeed they did, in the years before the crisis (Slide 9). Regulators cannot therefore rely on free-market discipline to ensure that the debt is only held by investors who can suffer loss without that causing knock-on systemic disruption.

If therefore we can neither perfectly and continuously monitor or regulate who owns bail-inable debt, nor rely on free-market discipline to ensure that it is always appropriately held, contractually bail-inable debt and technical resolvability will be valuable but still imperfect solutions to the ‘too big to fail’ problem. We can only be sure that losses can be smoothly absorbed if we are sure that the investors who provide funds do not suffer from ‘local thinking’ but remain perpetually aware of the full distribution of possible results. Subordinated debt which can convert to equity well before potential failure (‘early trigger CoCos’) may approach what is required since the price will presumably vary with probabilistic expectations of future conversion. But only with pure equity can we be fully confident that the dangers of ‘local thinking’ will not creep in over time, and that investors, facing day-by-day price movements up and down will remain continually aware that they hold a potentially loss absorbing instrument. The implication of Shleifer’s ‘local thinking’ theory is that if investors are to remain continuously aware of the full frequency distribution of objectively possible results the observed frequency distribution of returns needs to include negatives and well as positives. This is achieved by equity returns but not by low risk debt.

OSFI, in its infinite wisdom, is going in entirely the opposite direction: the lowest possible conversion triggers for CoCos, and seeking to include CoCos in the regular bond indices so that investors will be fooled into buying them.

Credit Suisse is issuing contingent capital:

The bank agreed to sell $3.5 billion of contingent convertibles with a coupon of 9.5 percent, and 2.5 billion francs with a coupon of 9 percent, it said. The sale will happen no earlier than October 2013, which is the first call date on $3.5 billion of 11 percent and 2.5 billion francs of 10 percent Tier 1 capital notes the bank sold in 2008.

The notes will convert into shares if the bank’s Basel III common equity Tier 1 ratio falls below 7 percent. The conversion price will be the higher of the floor price of $20 or 20 francs per share or the daily weighted average sale price of ordinary shares over the trading period preceding the notice of conversion, the bank said.

The transaction is subject to the implementation of Swiss regulations and the approval of shareholders, the bank said. The Swiss committee proposed that the country’s two biggest banks should hold common equity equal to at least 10 percent of their assets, weighted according to risks. In addition, the companies may hold up to 3 percent in so-called high-trigger CoCos that would convert into shares if the bank’s common equity ratio falls below 7 percent, plus 6 percent in CoCos that would convert at a 5 percent trigger.

Credit Suisse said the 6 billion-franc sale would satisfy about 50 percent of the high-trigger requirement. The bank said it would like to see the market for contingent convertible bonds expand to a wider group of buyers and is pursuing an additional offering of such notes to potential investors outside the U.S. and certain other countries.

On the positive side, conversion occurs well before the the point of non-viability. On the negative – the trigger is based on Capital Ratios, which I have strongly criticized in the past and continue to criticize.

The Financial Times comments:

Switzerland’s other big bank, UBS, takes a diametrically opposed view to Credit Suisse, on cocos, arguing that they will be excessively expensive because no one knows how to price them properly. UBS prefers the “haircut bond” as an instrument.

But investors believe that other UK banks, such as HSBC, could be drawn to cocos. “That would really seal cocos’ reputation,” said one London-based investor. “But in the meantime, we expect the Nordics, particularly Sweden, to be big issuers. We also think this will take off in the US.” In spite of a lack of enthusiasm from US regulators, the likes of Morgan Stanley and Goldman Sachs are privately intrigued by cocos.

Senior bankers at BNP and Société Générale have similarly signalled a willingness to consider coco issuance to finance buffers. Analysts at Barclays Capital said the market for European cocos alone could be close to €700bn ($945bn) by 2018.

Many traditional fixed-income investors are barred from owning instruments such as cocos that can convert into equity.

Update, 2011-2-23: The deal was a huge success:

Investors rushed to take up the benchmark issue by Credit Suisse of a new financial instrument hailed by regulators as a key tool for rebuilding the capital strength of banks, placing orders of $22bn – 11 times the $2bn on offer.

The deluge of orders represented a big vote of confidence in the nascent market for contingent capital bonds, dubbed cocos.
…
Asset managers took about two-thirds of Credit Suisse’s cocos, while private banks took a third on behalf of their clients. A total of 550 different investors – an unusually large number – put in orders for the bonds. The strong demand from asset managers was particularly important since they will form the backbone of any sustainable market for the products.

Credit Suisse’s deal was helped by the fact the bank anchored its coco deal by simultaneously announcing a agreement to swap $6.2bn of its existing hybrid debt for cocos – covering in one go about half the total cocos the bank needs to issue.

OSFI has released a Draft Advisory titled Non-Viability Contingent Capital (NVCC):

OSFI has determined that, effective January 1, 2013 (the Cut-off Date), all non-common Tier 1 and Tier 2 capital instruments issued by DTIs must comply with the following principles to satisfy the NVCC requirement:

Principle # 1: Non-common Tier 1 and Tier 2 capital instruments must have, in their contractual terms and conditions, a clause requiring a full and permanent conversion [Footnote 4] into common shares of the DTI upon a trigger event.[Footnote 5] As such, original capital providers must not have any residual claims that are senior to common equity following a trigger event.

Footnote 4: The BCBS rules permit national discretion in respect of requiring contingent capital instruments to be written off or converted to common stock upon a trigger event. OSFI has determined that conversion is more consistent with traditional insolvency consequences and reorganization norms and better respects the legitimate expectations of all stakeholders.

Footnote 5 The non-common capital of a DTI that does not meet the NVCC requirement but otherwise satisfies the Basel III requirements may be, as permitted by applicable law, amended to meet the NVCC requirement.

Some extant contingent capital has a “write-up” clause, whereby amounts written down can be recovered if the company squeaks through its troubles.

The minimum condition reveals that OSFI is more interested in political posturing than averting a crisis. If they wanted to avert a crisis, they would insist that conversion took place long before the point of non-viability, when the common still had value.

Principle # 3: All capital instruments must, at a minimum, include the following trigger events:

• a. the Superintendent of Financial Institutions (the “Superintendent”) advises the DTI, in writing, that she is of the opinion that the DTI has ceased, or is about to cease, to be viable and that, after the conversion of all contingent capital instruments and taking into account any other factors or circumstances that she considers relevant or appropriate, it is reasonably likely that the viability of the DTI will be restored or maintained; or
• b. a federal or provincial government in Canada publicly announces that the DTI has accepted or agreed to accept a capital injection, or equivalent support [Footnote 6], from the federal government or any provincial government or political subdivision or agent or agency thereof without which the DTI would have been determined by the Superintendent to be non-viable [Footnote 7]

  Footnote 6: OSFI, after consulting with its FISC partner agencies, will provide guidance to DTIs upon request whether a particular form of government support being offered to such DTI is considered equivalent to a capital injection. For example, the Bank of Canada’s Emergency Liquidity Assistance (ELA) does not constitute equivalent support as it is targeted at solvent institutions experiencing temporary liquidity problems.

  Footnote 7: Any capital injection or equivalent support from the federal government or any provincial government or political subdivision or agent or agency thereof would need to comply with applicable legislation, including any prohibitions related to the issue of shares to governments.

So the Superintendent, an employee of the federal Ministry of Finance, has absolute power – there is no appeal. There is nothing to prevent the Superintendent from saying tomorrow that the Royal Bank is non-viable, the Government is buying a hundred-billion shares for a dollar, fuck you suckers, goodbye. Five hundred years of bankruptcy law out the window.

Principle # 8: The issuing DTI must provide a trust arrangement or other mechanism to hold shares issued upon the conversion for non-common capital providers that are not permitted to own common shares of the DTI due to legal prohibitions. Such mechanisms should allow such capital providers to comply with such legal prohibition while continuing to receive the economic results of common share ownership and should allow such persons to transfer their entitlements to a person that is permitted to own shares in the DTI and allow such transferee to thereafter receive direct share ownership.

Since we’re ignoring bankruptcy law, why not ignore every other law and contract while we’re at it?

Section 3: Issuance of Capital Instruments prior to the Cut-off Date
…
3. DTIs are encouraged to consider amending the terms of existing non-common instruments that do not comply with the NVCC requirement to thereby achieve compliance, or to otherwise take actions, including exchange offers, which would mitigate the effects of such non-compliance.

It’s possible that some issuers might try this, but it’s awfully hard to imagine the kind of coercion that would be required to get something like this to pass for a PerpetualDiscount, given the reasonable expectation of redemption at par within ten-odd years.

Section 4: Criteria to be considered in Triggering Conversion of NVCC

In triggering the conversion of NVCC, the Superintendent will exercise his or her discretion to maintain a financial institution as a going-concern where it would otherwise become non-viable. In doing so, the Superintendent will consider the below list of criteria and any other relevant OSFI guidance [Footnote 16]. These criteria may be mutually exclusive and should not be viewed as an exhaustive list.[Footnote 17]

The exercise of discretion by the Superintendent will be informed by OSFI’s interaction with the Financial Institutions Supervisory Committee (FISC)[Footnote 18] (and any other relevant agencies the Superintendent determines should be consulted in the circumstances). In particular, the Superintendent will consult with the FISC member agencies and the Minister of Finance prior to making a non-viability determination.

Footnote 16: See, in particular, OSFI’s Guide to Intervention for Federally-Regulated Deposit-Taking Institutions.

Footnote 17: The Superintendent retains the flexibility and discretion to deal with unforeseen events or circumstances on a case-by-case basis.

Footnote 18: Under the OSFI Act, FISC comprises OSFI, the Canada Deposit Insurance Corporation, the Bank of Canada, the Department of Finance, and the Financial Consumer Agency of Canada. Under the chairmanship of the Superintendent of Financial Institutions, these federal agencies meet regularly to exchange information relevant to the supervision of regulated financial institutions. This forum also provides for the coordination of strategies when dealing with troubled institutions.

Full discretion, no judiciary, no appeal. Goodbye Canada, hello Soviet Union.

Update, 2011-2-7: DBRS says:

OSFI has also issued a draft advisory on non-viable contingent capital. Again, the draft advisory was consistent with the BCBS’s release on minimum requirements to ensure loss absorbency at the point of non-viability (January 13, 2011). The NVCC Draft Advisory sets out the governing principles, information requirements and criteria to be considered in triggering a conversion of non-viable contingent capital. DBRS will state its views on non-viable contingent capital when OSFI publishes a final release of the advisory, expected in 2011.

Notwithstanding the NVCC Draft Advisory, DBRS’s global bank rating methodology continues to deem the five largest Canadian banks (Bank of Montreal, Bank of Nova Scotia, Canadian Imperial Bank of Commerce, Royal Bank of Canada, and The Toronto-Dominion Bank) systemically important in Canada, which positively impacts DBRS’s senior and subordinated debt ratings of these banks.