Michael Mainelli, The Z/Yen Group
[A version of this article originally appeared as “Enter the Specialists” (credit derivative market transformation), Risk Professional, Issue 2/10, Informa Group plc (December 2000/January 2001) pages 32-35.]
Expanding Credit Markets
Growth in credit derivatives during both 1998 and 1999 exceeded expectations markedly. The British Bankers' Association estimated the global market at $586 billion for 1999 and predicts more than a 50% increase during 2000 to $893 billion. London has about 45% of the market, but the real competition may be cyberspace. Some commentators estimate that the two leading online credit exchanges, CreditTrade (London) and Creditex (New York), despite their caginess about figures, may have up to 5% of the global market between them. Credit derivatives form a complex, multi-product market. Plain vanilla credit default products accounted for 38% of the 1999 market and are growing rapidly, yet their market share is predicted to decline slightly this year. The most intriguing development may be the rise in market share of portfolio products and credit linked obligations which are already 18% of the total global market. This development, together with others in the portfolio management arena, could be significant for the future direction of today's banks.
Originators of credit risk, typically banks, create comparative advantage like any business. For banks, this means understanding customers' businesses well enough to assess the credit risks better than other lenders. The resulting specialisation by industry or geography means, even for the largest banks, that natural portfolios are subject to significant concentration by industry, by geography or by size. At the same time, returns from credit risk are only attractive if they can be highly leveraged. A more diversified portfolio can accommodate more leverage, hence banks seek diversification beyond their natural portfolio. The tools to achieve diversification are still limited: participations in the loan market, securitisations and capital market underwriting (but typically investment banks get the choicest mandates) can reduce concentration risk. Examples of geographical diversification by these means (e.g. European banks in the US loan market) show that the lower level of knowledge about the borrowers can destroy the improvements from diversification. Investment portfolios can build diversification as well but typically at lower absolute return levels. Credit derivatives are becoming useful on both sides (acquiring diversifying exposure and reducing the largest exposures) but are largely limited to high quality obligors and the impact at the portfolio level is, as yet, modest.
The margins on credit risk for a bank on underwriting a big issue, say US$ 1 billion, can be large. Although a complicated assessment, typical margins are between 2% and 3%. However margins may shrink rapidly in a fast-growth market. An analogy could be made with the growth of US Commercial Paper in the mid-1980's. At that time a commercial paper dealing mandate typically had a margin of 12.5 basis points, but as banks and investment banks were dis-intermediated by more liquid markets, created by direct dealing technology, mandates were reduced to 2 or 3 basis points. By way of comparison, all this points to stress building in the traditional credit market for the traditional players.
Buyers of credit risk, or potential investors, are those for whom credit risk diversifies their existing risk portfolios: insurance companies, pension funds, other investors. For them, building credit risk one obligor at a time creates concentration risk which negates the advantage. Only securitisations such as collateralised debt obligations make sense as they transfer the risk on a portfolio of obligors. Here, though, liquidity, income levels, credit risk first loss, second loss and tail risk are often intertwined in ways that do not suit the potential investors. Unwinding the intertwining may lead to breaking up the traditional bank.
An Ideal World
What should a solution for trading credit portfolios easily look like? From the seller's perspective, it needs to provide regulatory capital relief as well as transfer economic risk. The seller needs to cap the remaining risk in order to provide both economic risk transfer and regulatory capital relief. The risk that needs transferring comes from concentrations, by geography, industry or individual obligor: portfolios that are artificially smooth across large numbers of obligors are less attractive. From the buyer's perspective, the solution should allow the rapid build up of a sufficiently diversified portfolio such that concentration risk is manageable, perhaps via derivatives.
Efficiency is important too. Banks vary in their structural liquidity and some will want to free up cash/liquidity as well as transferring credit risk, but others may prefer the keep the funding element. This suggests both funded (on balance sheet) and unfunded (off balance sheet and derivative) solutions will continue to exist.
For any market to take off, there need to be standards around which a price and other characteristics can be described. In options markets, this is the venerable Black-Scholes pricing formula. Even though traders may use more complex and proprietary models to price and manage risk in option books, the trades are often posted and traded as if the Black-Scholes model were used. At the asset level, credit risk is more difficult to price than market risk. Although the credit market is becoming more liquid, it is still almost impossible to price a specific obligor or tenor confidently using market data. Further, default probabilities and correlations are impossible to observe or measure. Also, it is not just default: a credit instrument also declines in value when an obligor is downgraded. Finally, to calculate capital/equity reserves requires a good estimate of asymmetric, fat-tailed loss distributions. While credit risk does not yet have an equivalent to Black-Scholes, we can be fairly sure what it would look like.
A key component is a simple, objective, and 'good enough' credit rating system. It should produce an explicit probability of default, over a useful time horizon (similar to the maturity of most deals, so probably 3 to 7 years), be sufficiently granular to allow meaningful differentiation across the entire spectrum of undoubted investment grade to nearly bankrupt. 100 users analysing a single company should find the same answer, not 10 different ones, which means its inputs must be objectively verifiable. This credit rating system needs to be based on a publicly disclosed methodology, be open for testing by many participants and be maintained and calibrated regularly. There is a competitive market for credit rating systems, although the market is overly-focused on the larger obligors.
The second key component is a standard to describe the degree of diversification. This means a model to measure correlation across a portfolio, commonly termed a portfolio model. Again, the model must be open, transparent and objective. At the portfolio level, a number of models have been proposed: KMV's Portfolio Manager (1993), JP Morgan's CreditMetrics (1997), CSFP's CreditRisk+ (1997) or McKinsey's CreditPortfolio View (1998). While they differ in details, for instance using insurance-based continuous random variables or incorporating macroeconomics, in the end they all create a distribution of possible credit portfolio values. Nevertheless, it is fair to say there are no market standards yet. While we can look forward in the near future to having the analytical building blocks ready, are the players?
Preparing for a Brave New World
Banks, in this framework, originate, warehouse, administer and distribute credit risk. They will, in addition to a pure warehouse, hold a certain amount of credit risk for their own portfolio as one of the investments of surplus liquidity generated by their other business lines and relationships. Both the warehouse and the own portfolio will be managed along true portfolio lines. Initially this will simply highlight the contribution the various functions make, but eventually this will lead to setting up distinct units to manage the four functions.
Managing the resulting organisation requires that appropriate incentive structures are in place, in order to ensure that all functions contribute positively to shareholder value creation and are not operated parochially. The interaction among functions in the organisation, e.g. between the financial markets and trading businesses, needs careful consideration to avoid unnecessary duplication of skills and analyses, as well as too many players trying to earn a margin along the path from origination to sale.
Banks will also begin to analyse their source of competitive advantage more clearly. In most cases, competitive advantage probably arises from traditional relationships combined with the associated skills in origination, cross selling of other banking products, and credit administration. For some banks, warehousing may be a source of competitive advantage where their short line of communication with potential investors coupled with detailed understanding of the credit risks could be difficult for others to achieve. The regulatory environment with its emphasis on capital adequacy will restrict banks' ability to use their own portfolio as a source of competitive advantage. As several of these functions require similar skills, efficiency will become important. A single set of consistent analytical tools, working with an efficient data set, can ensure an advantage in profitability as margins reduce. Some banks are already involved in marketing these analytical tools, perhaps one or two may become specialists in this area.
Investors will extend their activities in building portfolios for themselves and clients to include the use of leverage. Credit without leverage is not an attractive asset class to hold, or trade, due to the highly asymmetrical nature of the returns. The determinant of the amount of leverage which can be employed is the 'tail risk' of the portfolio, which is driven by the structures of the underlying transactions (secured/unsecured, the value of security, collateral usage, etc.) and diversification. Investors will therefore need to build skills to understand those elements to compete efficiently and this may include significant economies of scale. A further development will be the use of insurance to limit the 'tail risk' and manage the safe leverage levels. This may in effect become a second source of risk capital, next to the investors' own.
A particular class of intermediate investor is likely to arise. These are the consolidators and aggregators who will provide new ways of creating portfolio components. By seeking out desirable risk groupings from originators, in conjunction with keen marketing and pricing, they will onward-sell portfolio components of risk needed by the larger warehouses. Consolidators and aggregators will threaten some of the credit origination and much of the credit distribution of today's banks.
Obligors will not be pure bystanders, certainly when compared to the equity markets. In the past two years the proportion of credit derivatives written against corporate assets has risen from 35% of the market to 55% while sovereign assets are declining. Consolidation among larger banks means that smaller firms find their credit lines shrinking. Shrinking credit lines are forcing them to be more creative, to look to other sources of credit, to look at new structures and to talk with new types of organisations directly about credit risk transfer. Insurers and non-bank banks are also developing products for smaller organisations which will increase portfolio trading opportunities. Technology is widening the potential pool of clients for credit products. Moves towards direct credit exchanges could broaden the potential pool of clients from large corporates in wholesale markets to smaller corporates globally.
Inevitably, some credits will go bad. Here, ownership takes on a different aspect as skills in recovery become more important than those of portfolio analysis. In future, in the event of default, ownership will probably transfer from the holder to a recovery specialist, who may well be the originating bank, or eventually one of several specialist companies. Going back to the originator has implications for pricing (who determines fair value?) and for the controls necessary in the origination process. In addition, accounting issues can impede this process if the financial consequences in published accounts are not closely aligned between the various players. Jurisdiction plays a great role as creditor rights vary markedly between countries. There could be complex interactions with other corporate creditors, particularly as receivables can often constitute up to 35% of their balance sheets. It could be that the rise of credit markets promotes more international uniformity in creditor rights and bankruptcy procedures as more firms seek access to credit markets.
Deal structures will likely evolve for quite some time. It is useful to look at the experience in the US Collateralised Mortgage Obligation (CMO) market. CMO structures in the early 1990's became increasingly complex with 20 or more tranches dividing the cash flows and the prepayment risk in more and more ingenious ways. The prepayment boom and bust cycles that followed led quickly to simplification of deal structures, as players realised little or no value is created by over-complication. Credit portfolio deals should therefore be structured as simply as possible, and players should strive to avoid the spiral of increasing complexity so common in financial markets. While the end point is becoming more clear, what is unclear is whether getting there will be straightforward or a painful evolution.
Competitive markets require competitive skills; typically competition promotes specialisation for most and enormous size for a very few. In the late 1980's Lowell Bryan, James Burnham and Robert Litan, to name a few, envisioned 'narrow banking'. Narrow banking foresaw a set of low-risk banks which would constitute the payment system for the economy. Other banks and specialists would take on high risk financing. While Bryan et al's vision of narrow banks never arrived, specialisation arising from competitive credit markets could create 'narrow credit banks'. Credit specialisation would shatter the comfortable links between origination, warehousing, administration and portfolio management which exist today in many banks. Narrow credit banks would have to know where their competitive advantage lay - origination, warehousing or credit administration.
Jan-Peter Onstwedder is an independent consultant specialising in market and credit risk measurement and management. He is the former Head of Group Market Risk for the Royal Bank of Scotland. His current interests include portfolio management methodologies and their application in financial services.
Michael Mainelli is a Director of Z/Yen Limited. Michael is a former Big 5 accountancy partner who has worked on cutting edge risk systems, including risk visualisation. Z/Yen is a risk/reward firm which specialises in the wholesale markets, providing strategy, systems, organisational, benchmarking and commercialisation solutions in order to improve performance.