Up to last year, most people regarded valuation as a necessary but unexciting aspect of running an investment operation. Spotting future price changes was the skill to which traders aspired, not the humbler job of correctly measuring today’s price. Assigned to lowly, dealer-support and risk quants, valuation was like plumbing: an essential activity but not one most people were desperate to exercise themselves.

This year, perceptions of valuation have changed radically. The difficulty of pricing illiquid securities has been the main factor driving the increasingly drawn out market turmoil. The levels of confusion and disagreement about valuations have been unprecedented. Investment banks have, reportedly, valued particular illiquid securities internally at levels 20% apart. The values of credit structured products have fluctuated massively. In particular, Asset Backed Securities (ABS) with triple or double A ratings have traded at enormous discounts.

Discrepancies between valuations remain but the unease runs deep

If agents disagree on prices, one may usually expect them to transact and for values to converge. But few investors have been sure enough of their own valuations of the most distressed asset classes (such as real estate ABS) in recent months to want to buy or sell aggressively on the basis of them. So liquidity has dried up and volumes have been pitiful despite the gyrations in price and the discrepancies between the valuations of different market participants.

The data I present in Figs.1,2 & 3 demonstrates the extent of the problem. In normal times, individual ABS may be reasonably precisely valued using spreads extracted from the current cross-section of prices for similarly-rated securities. Fig 1 shows the distribution of values around the ‘model’ or ‘fitted price’ yielded by such a fitting exercise for a large cross-section of Home Equity Loan ABS. (The fits allow for Constant Prepayment Rates and cash-flows generated by contractual repayments of principal appropriate for such securities.)

The data in the Figs. show the distribution of gaps between actual and model or fitted prices for three dates – May, August and November of 2007 and demonstrates how uncertainty about pricing grew progressively as the crisis took hold. For AAA-rated ABS, the differences between the actual and fitted values for AAA (Fig 1) are mostly within a range of plus and minus 3 before the crisis began. (The notional of these securities is adjusted to be 100.) They changed little by August but by November had increased significantly.

The AA-rated (Fig.2) and A-rated (Fig.3) securities appear tightly priced in May but the ranges of value grew very substantially in the period up to November. The range of values for AA-rated securities was breath-taking by that date.

The problem of achieving consensus on ABS values is worsened by the opaqueness of these products. The most careful pricing analysis requires a bottom-up calculation that takes into account the quality of the collateral and the nature of the cash-flow waterfall. This is the type of calculation that investment banks advising at origination would perform. Even if one has sufficient information to carry this out, different analysts are likely to disagree about the quality of the underlying collateral.

To illustrate the complexities involved, not so long ago, a deal was issued in which a AAA tranche was subsequently discovered to be almost underwater because a quant had incorrectly programmed the spreadsheet that the parties to the deal had used at issue. The valuation and the rating of the deal had been subject to the most egregious error and yet no one had noticed at the time of issue. A debt issue by a large commercial company could never be affected by such problems because the value of the debt is much more obvious from common sense analysis of yields and basic knowledge about the credit standing of the borrower.

Don’t rely on ratings alone

Some highly-specialised investors have always made money by carefully inspecting ABS deals including the quality of their pools to see if they are mispriced. But most mere mortals have taken a more broad brush approach, relying on the assessments of the ratings agencies and on comparisons of relative value across different ABS sectors and issuers. Few have invested in the infrastructure of analysis (involving developmentof models, data sources and expertise) needed to react in a timely and effective way to the current crisis.

For the less specialised investors who relied on ratings agency assessments, the events of the last year have profoundly dented confidence. What became evident as prices tumbled was that the agencies’ subsequent surveillance of structured deals issued in the past left much to be desired. The agencies had got into the habit of exerting themselves a lot prior to the initial issue but then failed to keep track of things as the market deteriorated.

Who to trust?

The consequence of the uncertainty investors have faced in valuing illiquid credit exposures has been major anxieties about counter-parties and a collapse in market liquidity. Individual hedge funds and banks, keenly aware of their in-house problems in placing values on their own books, have become highly distrusting of counter-parties. This distrust has generated the counter-party risk worries that have so dramatically undermined the interbank market since August of last year and led hedge fund prime brokers to demand ever larger risk haircuts on repo deals.

But if confidence in the ratings agencies is no longer there, on what can investors rely? A partial solution is third-party valuations. The advisory firms are currently being run off their feet by requests from their clients to assess illiquid structured product books.

But privately supplied, third-party valuations are a limited blessing. Valuations can have the features of a public good. All the investors in an issue would benefit from learning assessments of its value. Indeed, the investors in the other tranches of a deal or even investors in other comparable deals (say in the same sector) would benefit from receiving this information.

To the extent that publicising third-party valuations helps to create greater consensus on values, market liquidity might be partially restored. So, third-party valuations that enhance liquidity would ideally be made public. But of course there is right now no mechanism for this. So, third party valuations remain (inefficiently) private.

No more secrets

If valuations themselves cannot right now be publicised, at least the inputs on which valuations are based might be more widely disseminated. Here, the initiatives currently under consideration by bodies like the European Securitization Forum are important. Pool performance data together with information about prepayment and basic contractual information if widely distributed for a large number of deals would blow open the current opaqueness of ABS markets.

Participants in these markets, especially those involved in the new issue market, have major incentives to reconsider their arrangements. The current crisis in residential mortgage-backed ABS has a rather telling precursor in the collapse in the Manufactured Housing Loan (MHL) sector of the ABS market in 2002. At the time, Manufactured Housing-backed issues were numerous (indeed there were more MHL issues than Home Equity Loan (HEL) issues.

Through 2002 and 2003, prices in MHL issues behaved very like what we have seen in the HEL sector this year. For a given rating, the range of values ballooned. Most tellingly, it took years for the prices to converge to a more reasonable spread and by then the volumes in that market had fallen by 80%. So anyone who wishes to see a revival in ABS-market new issues has to be prepared to consider radical measures. The most effective step would be for issuers to publish widely all the details of their deals including frequently updated versions of the information necessary for outsiders to analyse valuation and risk in detail. Secrecy and discretion are no longer an option if the structured productmarkets are to recover anything approaching their former level of dynamism.

Source for graphics: “Determinants of Asset Backed Security Prices in Crisis Periods,” William Perraudin & Shi Wu (Mar 08).