There is more to risk than credit ratings
SLOWLY but surely, regulators are turning their fire on the credit rating agencies. These hapless firms are rightly being accused of having failed to assess risk correctly during the bubble years. Ludicrously risky financial products were routinely awarded the safest ratings, with ultimately disastrous consequences.
It is important to point out that the overwhelming majority of ratings agency employees were honest. Unfortunately, many were psychologically contaminated by the folly of the past few years (the same was true of most financiers, investors, central bankers and regulators). Especially insidious was the widespread belief that as long as there was a 95 per cent chance of nothing going wrong, the remaining 5 per cent probability of Armageddon could safely be ignored.
But it is also true that incentives in the industry are not helpful; in some cases they probably encouraged a handful of staff to grant lenient ratings to products they might have been privately suspicious about. It is those who issue the debt who pay the agencies, not those who buy it – and who therefore truly care about its risk.
Institutions sponsoring securitisation programmes are a case in point: it is in their interest that their products are deemed low-risk to find more buyers, such as pensions funds. But during the bubble years, something even more damaging happened. Because ratings are used for regulatory purposes – for example, a bank that holds AAA securities needs to put less capital aside that if it holds securities that are deemed riskier – the buyers too had an incentive to see the agencies underestimate risk. So neither sellers nor buyers had a real interest to look out for rating inflation or to query optimistic grades awarded to property-based derivatives; everybody had an incentive to turn a blind eye.
So what should be done about this? And how can the riskiness of stocks, bonds or derivatives be better assessed? Regulators should ideally strip the agencies of their moral authority and rely more on market mechanisms. The agencies’ views should be seen merely as one opinion. More attention should be paid instead to credit default swaps: while they too fell foul of bubblenomics, the CDS market proved that traders’ views on risk and shifts in prices are better guides to looming problems than movements in credit ratings.
Several other ideas are explored in a paper by Charles W Calimoris of Columbia university, published by the Cato Institute. The best way to assess the risk of a loan is to use the interest rate paid as an index of its risk, he says. Calimoris also argues that banks should be forced to issue uninsured bonds; interest rates paid on these would be a good guide to what the market thinks of their riskiness.
It is likely, however, that governments will want to continue to rely on ratings agencies to measure the riskiness of assets held in bank portfolios. If so, agencies should produce precise numerical estimates of the probability of default – as well as loss given default – for all rated instruments. At the moment, the agencies produce such stats only retrospectively, looking at past failure rates, which is pretty useless. There could even be penalties for consistently underestimating risk.
Witch-hunts serve little useful purpose. But we must find better ways of estimating risk if we are to avoid another global boom and bust. [email protected]