PrefBlog

The Bank for International Settlements has released a working paper by Allen B Frankel titled The risk of relying on reputational capital: a case study of the 2007 failure of New Century Financial:

The quality of newly originated subprime mortgages had been visibly deteriorating for some time before the window for such loans was shut in 2007. Nevertheless, a bankruptcy court’s directed ex post examination of New Century Financial, one of the largest originators of subprime mortgages, discovered no change, over time, in how that firm went about its business. This paper employs the court examiner’s findings in a critical review of the procedures used by various agents involved in the origination and securitisation of subprime mortgages. A contribution of this paper is its elaboration of the choices and incentives faced by the various types of institutions involved in those linked processes of origination and securitisation. It highlights the limited roles played by the originators of subprime loans in screening borrowers and in bearing losses on defective loans that had been sold to securitisers of pooled loan packages (ie, mortgage-backed securities). It also illustrates the willingness of the management of those institutions that became key players in that market to put their reputations with fixed-income investor clients in jeopardy. What is perplexing is that such risk exposures were accepted by investing firms that had the wherewithal and knowledge to appreciate the overall paucity of due diligence in the loan origination processes. This observation, in turn, points to the conclusion that the subprime episode is a case in which reputational capital, a presumptively effective motivator of market discipline, was not an effective incentive device.

The end of the road for New Century came when:

Purchasers of New Century’s loan production normally conducted a due diligence examination after a sales agreement had been reached. The investor, or a due diligence firm hired by the investor, would review loan files to determine whether the loan was underwritten according to the pool’s guidelines. Loans not meeting guidelines could be excluded from the loan bundle (kicked out) and returned to the originator.

Once kicked out, the mortgages were known as a “scratch and dent” (S&D) loans, which were purchased by specialised investors at a large discount to their principal balance. Consequently, one measure of the deterioration of the quality of New Century’s loan production is the percentage of S&D loan sales. In 2004 and 2005, such sales amounted to less than 0.5% of New Century’s secondary market transactions. By contrast, in the first three quarters of 2006, S&D loan sales accounted for 2.1% of such transactions (Missal (2008, p. 68)).

The upsurge in loan repurchase requests to New Century coincided with a change in the methodology employed to estimate its allowance for loan repurchase losses. New Century’s board learned of the change after a considerable delay. This discovery was followed, after a few days, by a public announcement on 7 February 2007 that New Century’s results for the three quarters of 2006 needed to be restated. It also noted an expectation that losses would continue due to heightened early payment default (EPD) rates.

New Century’s announcement prompted margin calls by many of its warehouse lenders and requests for accelerated loan repurchases. Soon, all of New Century’s warehouse lenders ceased providing new funding. Because simultaneous margin calls by its warehouse lenders could not be met, New Century filed for bankruptcy on April 2, 2007. It ceased to originate mortgages and entered into an agreement to sell off its loan servicing businesses.

Amusingly, in the light of the current bonus hysteria:

The examiner’s access to internal New Century documents provided valuable insights into how the appearance of the warning flags influenced, or did not influence, management. For example, the examiner could find no reference to loan quality in the internal documents that described New Century’s bonus compensation system for regional managers for 2005 and 2006 (Missal (2008, p. 147)). The examiner says that the compensation of New Century’s loan production executives was directly and solely related to the amount of mortgage loans originated, loans that, in turn, were subsequently sold or securitised.32 Likewise, the examiner found no mention of penalties (reduced commission payments to loan production staff) that would be assessed against defective loans that required price discounts for secondary market sale.

Heightened investor concerns about the performance of subprime loans were reflected in changes in their due diligence processes (Missal (2008, p. 165)). Historically, investors would ask due diligence firms to examine, on their behalf, only a small sample of loans in a particular pool. The character of the process first changed in 2006 when most investors began to look at the appraisal documents in all loan files in a loan pool. Investors then increased the share of loan files examined. This intensification of due diligence efforts was responsible for a sharp increase in New Century’s kickout rate from 6.9% in January 2006 to 14.95% in December 2006 (Missal (2008, p. 161)).

The author concludes:

The examiner’s report suggests that some of the actions undertaken to improve loan quality in late 2006 and early 2007 were designed to anticipate new credit risk concerns among New Century’s counterparties. Nonetheless, when New Century announced a need to recast its financial reports, there had not yet been a defection by any of its largest counterparties. Not surprisingly, defections ensued immediately after the announcement. In those circumstances, the bunching of defections probably signalled an absence of attention on the part of counterparties to the mounting risks of ongoing transactions with New Century. In turn, the evidence of ineffective counterparty risk management has led to concerns about the effectiveness of existing governance structures (corporate and regulatory) and, in particular, reputational capital as an incentive device. Can those structures now be relied on to discipline the risk-taking incentives of those involved in underwriting securities backed by subprime (and other risky) assets?

The Bank of Engand has released the December 2009 Financial Stability Report, with the usual tip-top analysis.

The first chart puts things into perspective: the UK is a smaller economy than the US or the Continent:

Who but the Old Lady of Threadneedle Street would dare produce an equity returns graph dating back to 1693?

UK banks are strongly encouraged to sell equity:

Despite inevitable short-term costs, there is a strong case for banks acting now to improve balance sheet positions while conditions are favourable. Retaining a higher share of current buoyant earnings could significantly increase banks’ resilience and ability to lend. If discretionary distributions had been 20% lower per year between 2000 and 2008, banks would have generated around £75 billion of additional capital — more than provided by the public sector during the crisis. It is also an opportune time for banks to raise capital externally, extend the maturity of their funding, and develop and implement plans for refinancing substantial sums as official sector support is withdrawn.

The bank is also throwing its weight behind Contingent Capital and leverage caps:

Capital buffers will need to rise, possibly substantially, over the coming years. The quality of banks’ capital also needs to improve. To absorb losses, capital should comprise equity or instruments that convert to equity automatically under pre-defined conditions. To avoid excessive reliance on refined regulatory risk weights, risk-based capital requirements should be accompanied by a mandatory maximum leverage ratio (Box 6).

They’ve done some work to see how much capital should be required going forward:

On average, a pre-crisis Tier 1 capital ratio of around 8.5% would have been needed by banks in the sample to avoid going below a Tier 1 capital ratio of 4% during the crisis (Chart A). Minimum capital requirements are likely to be higher in the future.

A feature of this analysis is the wide variation in results across banks, shown by the distributions in Chart A. Banks with similar pre-crisis Tier 1 capital ratios faced different outcomes in some cases. Even if all banks in the sample had a pre-crisis capital ratio of 8.5%, 40% of the banks would still have breached the 4% Tier 1 capital ratio in-crisis. The highest pre-crisis Tier 1 capital ratio that would have been needed across the sample of banks to maintain a 4% Tier 1 capital ratio in-crisis is around 18%. This variation across banks suggests the need for flexibility in their future capital structure and potentially a higher average buffer. In principle, this could be achieved through greater use of contingent capital (see Section 3).

Oddly, they have a chart decomposing credit spreads according to the BoC methodology, which I dislike, as opposed to the Webber & Churm methodology used in the past:

There is no explanation in the report regarding the change.

There are big problems with loan-to-value ratios on commercial property … or there would be if they were recognized!

The sharp declines in capital values have triggered breaches of loan to value (LTV) covenants, with some loans in negative equity. Estimates from the Property Industry Alliance (PIA) suggest that average LTVs could reach 114% by end-2010.

As well as causing covenant breaches, declines in values (and rises in LTVs) will also have reduced firms’ access to credit by reducing the value of the commercial property that they might use as collateral for secured borrowing. Market contacts suggest that banks have been willing, to date, to show forbearance in respect of breaches of LTV covenants. In addition, research by De Montfort University suggests that, while loans are still performing, some lenders have not sought to revalue underlying properties. As a result, the sharp declines in capital values alone had a fairly limited impact on banks.

Footnote: The PIA is an alliance of five property bodies — the British Council for Offices, British Council of Shopping Centres, British Property Federation, Investment Property Forum and Royal Institution of Chartered Surveyors.

Many will find the commentary on the composition of Tier 1 Capital interesting, particularly given the Canadian limits following OSFI’s debasement of bank capital:

Ahead of the crisis, the composition of banks’ capital shifted away from common equity and reserves (core Tier 1 capital) towards lower-quality instruments (Chart 3.4). Experience during the crisis in the United Kingdom and elsewhere has revealed that these instruments were not always able to absorb losses for going-concern banks.

There is now broad agreement internationally that equity and reserves should form a much larger part of banks’ capital in the future. The Bank believes that no instrument should be classified as going-concern capital if it does not have the same loss-absorbing characteristics as common equity. In practice, this means either that the principal of the instrument can be written down at the same time and to the same extent as common equity, or that the instrument is convertible into equity — so-called ‘contingent capital’.

And now we’re getting into the meaty bit! Contingent Capital is a vital concept for preferred shareholders: I am convinced that the preferred share as we know it is dead; it will all be contingent at some point in the future. Bet a nickel.

On what terms private non-bank investors would be willing to provide such insurance remains unclear. For example, investor appetite may initially be restricted if these instruments are excluded from benchmark indices or are not permitted under certain investment mandates. If, over time, an investor base for such instruments did not develop, this would provide a useful signal that debt investors were unwilling to accept losses on their investments in banks. For contingent capital instruments to be loss-absorbing, their design needs careful consideration. In this respect, the definition of the conversion trigger is crucial. Contingent capital would need to convert automatically, or at the discretion of the regulator, rather than on the initiative of the issuer. Setting the trigger involves balancing the risk of conversion too soon (before capital is needed) and too late (when funding problems may already have emerged). The acceptable level of contingent capital within banks’ capital structure also needs to be considered carefully. Too much convertible debt could increase the risk of a bank equity price ‘death spiral’ — whereby investors may short-sell the stock in anticipation of dilution as the trigger for conversion comes closer.

This constant harping on regulatory discretion really gets on my wick. In a crisis, the pronouncement by a regulator that Bank X is sufficiently endangered that it needs to trigger conversion will be a death sentence. Conversion needs to be automatic, predictable and hedgeable: as I have argued countless times, these conditions are met by setting a trigger-and-conversion price at the time of issue of the non-equity capital (maybe 50% of issue-time common price for Tier 1; 25% for Tier 2). If the common price falls below the trigger price (on a well-defined exchange in a well defined way for a well defined period) then the contingent capital converts at that particular trigger price. Holders who wish to hedge will be able to buy options with their income payments … alternatively, options players may wish to buy the CoCo from the existent holders in order to get the embedded option; this will depend on the market price of the CoCo.

Death Spirals are not an issue if the conversion price is fixed; and become less important as a minimum coversion price increases (e.g., in Canadian Operating Retractible issues, the minimum conversion price is $2, which prevented the IQW.PR.C conversion from becoming a death spiral … although it ultimately made no difference).

The BoE states flatly:

The Bank would support the introduction of a leverage ratio
and this being hard-wired into regulatory rules through Pillar 1, provided that it can be well defined. It will be difficult to set a single standard applicable across different business models and accounting regimes, but it is important to achieve consistent implementation across jurisdictions.

… which, I think, can be taken at face value. Given the tenor of the rest of their discussion, I don’t think they intend to invent problems regarding implementation specifics to mask a distaste for the idea.

I’m not so enthralled with the following:

In a recent Discussion Paper (DP), the Bank contributed to emerging ideas on how such a macroprudential regime could be made operational.(2)
The DP examined the possibility of applying time-varying capital surcharges on banks to dampen cyclical exuberance (the orange bars in the stylised example in Chart 3.13). Raising capital requirements in a credit boom would offer greater self-insurance for the financial system against a subsequent bust. It could also provide incentives for banks to restrain exuberant lending by raising its marginal cost. In addition, the DP suggested that capital surcharges could be imposed on firms to better reflect their individual contribution to systemic risk (the magenta bars in Chart 3.13). These would be based on factors such as firms’ size, complexity, interconnectedness and propensity to cause losses to others through asset fire sales. The key objective would be to lower the probability of default of banks whose failure would impose a large spillover cost on the financial system. Systemic surcharges could also provide incentives for banks to alter their balance sheets or business models, supporting structural initiatives in this area (see Section 3.2).

I don’t support the nod-and-wink model of regulation by any means, and the proposals to give regulators discretion in such matters will only enhance the attractiveness of regulatory capture. Counter-cyclical requirements, yes: have a surcharge on asset growth over the past 5-10 years. Systemic surcharges, yes: have a surcharge on a progressive schedule based on risk-weighte assets. Regulatory discretion? No. Not only can’t the regulators be trusted with that degree of power (NOBODY can be trusted with that degree of power), but it raises the spectre of single-point failure even higher and will make the regulators the mutual plaything of the banks and politicians.

The next part is good. In Canada, we denigrate the shadow-banking sector:

With these objectives in mind, HM Treasury has announced that it intends to publish a discussion paper on developing non-bank lending channels in the United Kingdom, drawing on advice from the FSA and the Bank.(1) Key issues to be considered include identifying necessary improvements to market infrastructure that will help corporate borrowers to access non-bank investors.

One thing I haven’t seen before is the proposal to treat all retail deposits as covered bonds:

One way of ensuring continuity of payment services could be to require banks to invest retail deposits solely in risk-free assets such as government bonds — an approach commonly referred to as ‘narrow banking’. A number of commentators have put forward proposals along these lines in response to the crisis.(1) This could be seen as an extension of arrangements already in place for private banknotes issued by some Scottish and Northern Irish banks. These banks are required to hold cash or credit balances with the Bank of England fully backing their note issuance. These assets cannot be used for any other purpose and would be excluded — or ‘ring-fenced’ — from any insolvency proceeding and reserved for satisfying the claims of note holders.

An arrangement where retail deposits are backed by risk-free assets need not require the creation of dedicated narrow banks, although this could conceivably occur naturally over time. Existing banks could instead be required to segregate their retail deposit books and the assets backing them within their internal structures. The segregated part of a bank would effectively be subject to a 100% liquidity requirement, and would need to be easily extractable from the wider group using available resolution tools. In this way, the integrity of the payment system would be assured, while still allowing banks to exploit economies of scope between payment services and other types of banking activity.

The Federal Reserve Bank of New York has published a staff report by Michael J. Fleming and Neel Krishnan title The Microstructure of the TIPS Market. The authors examined the electronic records of inter-dealer broker trading, which provided quotes and trade sizes, everything time-stamped to the second. Nothing particularly useful here, perhaps, but there are some items of interest.

Trading activity for on-the-run TIPS is substantially higher than it is for off-the-run TIPS (Table 3). Daily trading in the on-the-run 10-year note thus averages $137 million, more than six times higher than average trading volume ($22 million) of individual off-the-run 10-year notes. The comparable ratio for the 5-year note is just over three ($87 million versus $27 million) and it is somewhat less than five for the 20-year bond ($30 million versus $6 million). Such on-the-run/off-the-run differentials are just as striking in the nominal market (Fleming (2002), Fabozzi and Fleming (2005), Goldreich, Hanke, and Nath (2005), and Barclay, Hendershott, and Kotz (2006)), reflecting a concentration of liquidity in just a few securities, and in those securities that tend to have the largest floating supplies.

While there is a similar on-the-run/off-the-run divergence in daily trading frequency, such a pattern is not evident in trade size. In fact, average trade sizes are actually slightly higher for off-the-run TIPS.

While bid-ask spreads and quoted depth are similar for on-the-run and off-the-run securities, “quote incidence” is markedly higher for on-the-run securities. Quote incidence gauges the percent of time there are two-sided quotes in a security (that is, both a posted bid price and a posted offer price). This proportion averages close to 60% for the on-the-run 10-year note (during New York trading hours, defined as 7:30 a.m. to 5 p.m.), but only about 15% for any given off-the-run 10-year note. That is, for off-the-run 10-year notes, there is a one-sided quote, or no quote, about 85% of the time.

There are pronounced day-of-week effects in trading activity in the TIPS market, as there are in the nominal market. In particular, trading volume is lowest on Monday, averaging $424 million, highest on Wednesday and Thursday, averaging $615 and $658 million, respectively, with Tuesday and Friday in between, at $552 and $546 million, respectively. These patterns remain when controlling for the announcements examined in this paper.

The Bank of Canada has released Discussion Paper 2009-14 by Jonathan Witmer, Market Timing of Long-Term Debt Issuance:

The literature on market timing of long-term debt issuance yields mixed evidence that managers can successfully time their debt-maturity issuance. The early results that are indicative of debt-maturity timing are not robust to accounting for structural breaks or to other measures of debt maturity from firm-level data that account for call and put provisions in debt contracts. The author applies the analysis from some recent U.S. studies to aggregate Canadian data to determine whether the market-timing results are robust. Although the relation between debt maturity and future excess returns is in the same direction as in the United States, it is not statistically significant. This mixed evidence, combined with the difficulties in interpreting predictive regressions of this nature, provides little support for the notion that firms can effectively reduce their cost of capital by varying the maturity of their debt issuance to take advantage of market conditions. Managers do, however, try to time their debt-maturity issuance, given that long-term corporate debt issuance in both Canada and the United States is negatively related to the term spread.

One possible mechanism of interest is:

An argument against [firms that are successfully timing an inefficient market as an explanation for the relation between future excess long-term bond returns and the long-term share [return]] is that, as a whole, corporate managers should not have inside information on the evolution of future market interest rates, so the evidence that their issuance decisions predict interest rates raises the question as to how firms in the aggregate have some sort of advantage over other sophisticated market participants, such as banks and institutional investors, in recognizing market mispricings. To explain this, Greenwood, Hanson, and Stein (2009) propose a “gap-filling” theory, whereby corporate issuers act as macro liquidity providers (e.g., by issuing long-term debt) in a segmented bond market where certain groups of investors have fixed maturity preferences for long-term assets.19 Consistent with this theory, they show that corporations issue more long-term debt when the government issues relatively less long-term debt.20 Moreover, they use firm-level data to show that larger firms and firms in better financial position are more likely to engage in “gap filling.” If long-term Treasuries provide a lower expected return when their supply decreases relative to short-term Treasuries,21 this provides an explanation of the apparent ability of corporations’ aggregate issuing characteristics to predict future bond returns.

This would be the flip-side of “crowding out”.

However, the author concludes:

In Canada, the relation between debt maturity and future excess returns is in the same direction as in the United States, but it is not statistically significant. This mixed evidence, combined with the difficulties in interpreting predictive regressions of this nature, provides little support for the notion that firms can effectively reduce their cost of capital by varying the maturity of their debt issuance to take advantage of market conditions. Managers do, however, try to time their debt-maturity issuance, given that longer-term corporate debt issuance in both Canada and the United States is negatively related to the term spread. In the United States, corporations also may be providing liquidity at a macro level, since corporate debt issuance is negatively related to the proportion of government long-term debt outstanding. In Canada, there is less evidence for a relation between these two variables. Hence, while managers are not successful at forecasting future returns, they at least attempt to do so, and changing their maturity structure in such a way could increase the risk of liquidation if managers issue more short-term debt in an attempt to time interest rates. But the increased liquidation risk at the end of the sample period caused by debt-maturity timing is probably minimal, given that Canadian corporations had a long-term share of corporate debt outstanding comparable to historic norms.

One possible mechanism that I was sorry to see not tested or discussed is the idea that managers out-perform the market because they have inside information about their own firms and projects. Under this hypothesis, managers would issue 30-year paper to fund a long-term project simply because the numbers work for them – e.g., future operational profits will exceed the cost of funding (hopefully substantially). Their ability to fund long term profitable projects should, in aggregate, affect macroeconomic spreads and bond returns, since project will be funded when the required yield works, and not funded when it doesn’t.