The collapse in CMBS was followed by a loss of confidence in the organisations that ‘rated' these products. Colin Lizieri asks, how do we eradicate the moral hazard and restore confidence in ratings?

In the current financial turmoil, as the sub-prime crisis gave way to the credit crunch and fears of world recession, it was inevitable that there would be a search for guilty parties. Attention turned to banking practices and the reliance on ‘complex financial products'. Mortgage-backed securitisation (MBS) is portrayed as inherently wrong, promoting demands for regulation or prohibition and calls for a return to traditional, conservative, approaches to property investment. But is the MBS model really fundamentally flawed?

The CMBS structure has many benefits in ‘normal' markets. By pooling loans and selling securities into capital markets, risk is spread widely, providing investors with diversified exposure to real estate debt. The tranche structure separates repayment cash flows, providing appropriate securities for investors with different appetites for risk. Banks access new sources of capital and obtain direct market signals of risk - yields in the CMBS market provide real-time information on property sentiment. These represent substantial efficiency gains, creating competition, limiting non-price rationing and unreasonable lending terms across the cycle.

Efficiency gains rest critically on supply of information and risk assessment. Loan originators must assess the risk of individual loans, setting conditions and interest rates reflecting the probability of loss given default. Securitisation works precisely because investors obtain diversification and do not face high information and monitoring costs. This relies on the securitisation process being transparent and on readily available risk-return signals. It is here that the key role of credit rating agencies becomes apparent. By providing accessible, understandable indicators of the characteristics of MBS tranches, rating agencies enable investors to select and manage portfolios of securities at low cost.

Now the flaws in this system are obvious. If rating agencies systematically mis-identify risk, and if investors make portfolio decisions relying on those ratings, then the market is extremely vulnerable to shocks. The volume of issuance of structured products and, critically, growing product complexity created major pressures for the agencies. With no performance data on the new securities and provision of inaccurate (even fraudulent) information on underlying assets, the complexities and time pressures contributed to ratings models that failed to account for inherent downside risks and high correlations between assets that might default if - when - extreme shocks occurred. This was compounded by modelling error (as with constant proportion debt obligations) but mispricing of risk seems endemic.

Issues of rating methodology are compounded by structural problems. First, there is pressure on agencies to provide more favourable ratings for clients to capture market share. Second, agencies simultaneously advise clients on the structuring of deals and rate those deals. There are voluntary codes of practice - Chinese walls - intended to keep processes separate, but moral hazard exists. In 2008, the SEC concluded that rating agencies had failed to manage conflicts of interest adequately. Many post-credit-crunch reform proposals have focused on this aspect. Thus, EU proposals seek compulsory registration, improved corporate governance, external oversight, inspection and transparency reviews.

Ratings are hard-wired into the architecture of global finance, embedded in the regulatory system, in Basel regulations that were intended to prevent global meltdown. ‘Institutionally acceptable' financial products are defined by ratings. Issuers must create assets that achieve credit ratings to meet regulatory and investment hurdles. CMBS structuring and rating have been locked together - structuring is about rating. Financial institutions bought CMBS portfolios because of their rating - ratings underscored their use as collateral in inter-bank borrowing. The importance of rating in regulation and investment helped preserve the dominant position of the big-three agencies. The EU proposals do nothing to counteract this, although the SEC seems to be moving away from regulatory reliance.

What will restore confidence? Greater oversight, improved governance and transparency address problems of conflicts of interest but do not address the structural issues of past failures. Over-rigid regulation stifles financial innovation and hardly engenders efficiency in credit markets however seductive a back-to-basics approach to financial structures might be. Will regulatory bodies have the skill base to audit rating? Enforcing greater competition and reducing dependence on the big-three agencies seems appealing: but part of the cause of mispricing of risk was competitive pressures among the main players.

Conflict of interest and client pressure arising from issuers paying for ratings could be solved by returning to investor subscription services. New problems arise, however. With so many investors in credit products, free-rider problems emerge - why pay when the rating will quickly become public? Second, issuers have less incentive to provide full, accurate information. A better alternative might be to create a pool of independent agencies. Listing authorities could require issuers to supply data and pay levies on issues to fund the pool. The listing authority independently selects agencies, creating arms' length ratings while maintaining valuable public information. Agencies could still provide consultancy services to issuers. To be effective such a system demands more than three ratings agencies.

Reducing or removing rating from regulatory structures will reintroduce caveat emptor for investors, who will no longer hide behind ratings-based rules. That would force rating agencies to demonstrate added value by producing price-sensitive information. Investors will demand ratings, issuers will commission ratings only if they are of benefit, not simply because they fit convenient Basel or investment mandate boxes.

To demonstrate added value, independent research is needed. Do agencies' initial ratings accurately predict losses given default? Does this result hold for different types of security, different market environments? The paucity of rigorous research on CMBS rating partly relates to data availability and openness. There is a need to encourage research and performance monitoring of the outcomes and effectiveness of rating.

Independent rating of debt vehicles is vital, but the current system has been inefficient at pricing risk. Any new structure must deal with structural flaws in the existing system without blocking information flows, becoming bureaucratically inflexible or stifling innovation.

Colin Lizieri is professor of real estate finance at the University of Reading