The rating agencies have been widely criticized for their role in the financial crisis. It is said that they wrongly assessed the risks on trillions of dollars worth of bonds backed by residential mortgages. And indeed they did. But this is hardly surprising.
Rating agencies employ quite ordinary mortals to analyze the credit risk of bonds, just as firms like Goldman Sachs and Merrill Lynch employ quite ordinary mortals to analyze the outlook for stocks. No one is shocked when equity analysts' recommendations don't pan out. Why should we expect any more of the rating agencies?
We should not, but the regulators have, and that is the problem. Regulators of banks, insurance companies and broker dealers have all incorporated the work of the ratings agencies into their regulations in myriad ways. Most importantly, bond ratings determine -- as a matter of law -- how much capital regulated institutions need in order to own the bonds.
For every dollar of equity that insurance companies are required to hold for bonds rated AAA, $3 is needed for bonds rated BBB, and $11 is needed for bonds rated just below investment grade (BB). For banks, the sensitivity of capital requirements to ratings is generally even more extreme.
The Bank for International Settlements also uses ratings to drive capital requirements, so the rating agencies have the same role in global capital markets that they have in the U.S.
For money market funds, ratings are equally critical: They are typically barred altogether from investments rated lower than AAA. In short, the ratings agencies are like a Consumer Reports for financial instruments -- but with the force of law behind their ratings. It is as if you were forbidden by law from buying an iron or a toaster unless it is rated "Excellent."
Since the ratings determine required capital, they have a profound influence on how financial institutions invest their assets -- in effect, the regulatory reliance on ratings makes the rating agencies the de facto allocators of capital in our system. And every actor in the financial system has every incentive to group and slice assets in ways that maximize not their fundamental soundness but their rating.
Indeed, that is the entire raison d'être of the $6 trillion structured-finance business, which serves little economic function other than as a rating-agency arbitrage. Subprime mortgages (and all manner of other risky loans) held directly by financial institutions are questionable assets with high associated capital charges. Each one alone would deserve a "junk" rating. Structured finance simply piles such risky assets into bundles and slices the bundles into tranches. The rating agencies deemed some 85% of the tranches by value as AAA, and nearly 99% as investment grade -- thus turning dross into gold by a sort of ratings alchemy.
This ratings alchemy created enormous demand for dross -- in this case, dodgy mortgages. Credit was extended to countless dubiously qualified purchasers of homes, which in turn drove dramatic increases in house prices. The housing bubble has now burst, with average house prices in America down some 20% to 25% from the peak. This led to the current crisis.
The problem was not the erroneous ratings per se; everyone misgauges risk and ratings agencies are no different. The problem is that these erroneous ratings were incorporated into law. Regulators should not have relied on ratings agencies to asses the risk of bond holdings. Instead, they should have relied on markets.
When they come to design new regulations to govern capital requirements and investment parameters for financial institutions, Barack Obama and Congress should write ratings agencies out of the regulation of financial institutions. The market is a far better judge of risk and value than any individual analyst, team or firm.
The amount of capital required to hold a fixed-income security should be determined not by a rating but by its yield, expressed as a spread over Treasurys. The higher the spread, the riskier the market has determined the asset to be, and more capital should be required to hold it.
Similarly, financial institutions should be required to set aside a percentage of their interest income every year as reserves for credit losses; the higher the spread, the higher the reserve percentage. Should spreads widen, the share of the return set aside for reserves should increase, thus gradually increasing reserves commensurate with the market's perception of increased risk.
A requirement that financial institutions set aside reserves for ultimate credit losses sounds obvious, but does not currently apply to holdings of bonds. Using market data to drive reserves, as well as to drive capital requirements, ensures that market perceptions of risk are continuously, but gradually, taken into account. In contrast, rating-agency downgrades are discreet events that often lag the data reflected in market prices.
Rating agencies would continue to have a useful role. Their assessment of credit risk, and their marshalling of pertinent data, would help investors make their decisions. And whether issuers or bond buyers pay for their services, rating agencies will want their ratings to be as predictive as possible. They will still be the "Consumer Reports" of the bond markets, simply without the force of law behind their recommendations. Buyers of toasters can decide for themselves, and so can buyers of bonds. Our economy would be stronger for it.