Is the issuer-pays compensation model still viable for credit rating agencies?
It's a fair question, considering the market turmoil we have experienced, and many market participants — investors, bond issuers and regulators — have asked it. Given the disappointing performance of the ratings of some mortgage securities in recent years, it is natural for people to wonder what can be done to improve ratings, and even to ask if there might be a better compensation model than one in which issuers pay the rating agencies.
I should start by noting that we are quick to acknowledge that our ratings of mortgage securities, particularly those issued between 2005 and 2007, did not perform well, and we profoundly regret that. Although we stress-tested these ratings to withstand as much as a 20% decline in housing prices, we did not anticipate the speed and severity of the downturn, which in led to housing declines of as much as 60% in some regions. We were not alone in estimating a less severe decline.
Indeed, few in the marketplace were farsighted enough to foresee just how far housing prices would fall.
But it is also important to remember that during this same period, our ratings — of corporate and sovereign debt; securities backed by credit cards, auto loans and equipment leases, and even mortgage-backed securities issued in Europe — have all performed as we expected despite using the same analytical approach as U.S. mortgage securities and, of course, within the same business model.
Nevertheless, we recognize that confidence in ratings has been shaken and must be restored if the credit markets are to function efficiently. Over the past few years, we have taken a hard look at our ratings, and we have worked closely with market participants to understand their concerns and answer their questions. They include:
Wouldn't conflicts be eliminated if investors paid for credit ratings?
The question assumes that subscriber-paid compensation is free from conflicts of interest, and this is just not true. While an issuer of debt prefers to receive the highest possible rating for his bond, an investor would rather have a lower rating because it decreases the price and offers the possibility of arbitrage should the rating be upgraded in the future.
But if investors benefit the most from ratings, why shouldn't they pay for them?
Most investors are not in a position to pay subscriber fees to a rating firm. The issuer-pays model allows us to publish our ratings for free on our website, whereas the subscriber-based model creates information haves and have-nots in the marketplace. Market coverage is another significant factor. A subscriber-based firm has no incentive to develop ratings that investors have not agreed to pay for, making it potentially difficult for new issuers to get ratings and for investors to learn about new opportunities.
What about a third-party board, such as what Sen. Al Franken has proposed, that would separate rating agencies from issuers?
The policy of the U.S. government is to reduce reliance on ratings by eliminating references to nationally recognized statistical rating organizations in certain regulations. Standard & Poor's emphatically supports that policy, and the Franken amendment would run counter to it. For example, having the rating agency assigned by a third party, whether the government or its designee, could lead investors to believe the resulting ratings were endorsed by the government, thereby encouraging over-reliance on the ratings — the opposite of what the government has said it wants to do.
Moreover, if rating engagements were assigned either by a random or a next-in-line basis, credit rating firms would have less incentive to compete with one another, pursue innovation and improve their models, criteria and methodologies. This could lead to more homogenized rating opinions and, ultimately, deprive investors of valuable, differentiated opinions on credit risk.
Then what can be done to improve ratings quality and manage conflicts?
Our approach has been to take a top-to-bottom look at our ratings and implement some transformative changes in our policies, procedures and criteria. For instance, we have made changes to our criteria for rating mortgage securities so that it will be much more difficult for such a security to receive a triple-A rating. We have greatly enhanced our already robust system of checks and balances to maintain analytical independence. In addition, our credit analysts, who generally have earned advanced degrees before joining S&P, must pass a rigorous certification program designed by New York University's Stern School of Business. We have further strengthened our firewalls that separate analysts from commercial activities. Our compliance function is one of the strongest and most comprehensive in the financial services industry. And we have embraced new levels of transparency that allow investors to see how we arrive at our ratings and to agree or disagree with our opinions and act accordingly.
We believe that issuers and investors should have a choice of firms using different compensation models, so long as potential conflicts of interest are disclosed along with the steps the firm has taken to mitigate those conflicts. We also believe that investors large and small will always have a need for independent assessments of credit risk as part of a complete examination of investment suitability.
Even as we have experienced traumatic dislocations in the global economy, we have also seen a very positive development. Investors now have a better understanding of the need to analyze all facets of risk — not just creditworthiness — in making investment decisions. This should help bring greater stability and fewer shocks to the system.