Solving Regulators' Credit Rating Agency Dilemma

Although Congress was swift in its condemnation of the bond raters, the financial industry can't wean itself off of their advice so easily

Few sub-industries have endured as much criticism and anger over the past couple of years as the credit rating agencies. First, they completely misjudged the size of the housing bubble and had to drastically lower thousands of ratings on mortgage-related securities. Then, sovereign debt crises arose and more downgrades followed. Congress had concluded that the financial industry should no longer rely on these firms in any way. But how does it function without them?

This problem has manifested itself in a number of ways. But arguably, the most serious issue comes into play with capital requirements. The amount of collateral that banks have to hold currently depends in part on how rating agencies judge the bonds they own: the higher the quality of a bank's debt, the less money that the bank must hold to guard against possible losses. That concept is simple enough, but what happens if you no longer have a rating agency to judge the quality of those bonds?

And that's where it gets messy. JPMorgan now expects that Congress' insistence to cut out the rating agencies is going to cause a delay in U.S. banks complying with new capital requirements. Dave Clarke at Reuters reports:

In their note, JPMorgan analysts said the immediate result is likely to be a delay in implementing a rule before the end of the year that tweaks capital requirements based on market risk for banks that do a lot of trading -- dubbed Basel 2.5.

The ban on credit ratings in U.S. regulations will also likely make it more difficult to implement the latest sweeping global capital accord reached as part of the Basel III agreement, the note said.

That agreement does not have to be fully in place until 2019, but U.S. regulators hope to begin issuing draft rules by the end of this year.

Regulators have a few ideas for how to combat this problem, which are explained here. One would use credit default swaps -- derivatives that provide the market's assessment of risk on a given bond. Another suggestion would provide regulators more discretion on defining how much capital must be held against different sorts of debt. Banks' internal risk models might also be more heavily relied upon.

All of these options have flaws and centrally one problem: you're asking someone to predict the future. Nobody can do that particularly well, especially in the context of a highly complex financial system. To be sure, the rating agencies got it wrong a few years ago, but so did almost everybody else. Even the housing bears didn't expect a financial crisis and resulting recession as severe as what occurred. You can't reasonably rely on past experience to provide a realistic estimate of future losses, yet we don't have a clearly better way to assess risk.

Instead, capital requirements need to be chiefly based on concentration risk, that is the amount of exposure a bank has to correlated events. For example, if a bank has a bunch of mortgage-backed securities, lots of collateralized debt securities based on mortgage bonds, and a huge portfolio of mortgages, then it has a lot of risk correlated to the housing market. If a specific domino falls over and deep, cascading losses may result, then more capital needs to be held.

This concept alone won't completely solve the problem either, but it will help create a more stable financial industry. If banks have enough capital on hand to cushion losses if any particularly industry segment or market deteriorates, then the only market shock that could lead to its failure would be a multifaceted, catastrophic panic. And in that situation, few banks would survive anyway.

Image Credit: REUTERS/Shannon Stapleton