Helmut Reisen: Green light for danger TUESDAY FEBRUARY 3 1998

Credit rating agencies need radical reform if they are to do their job properly

The Asian financial and currency crisis of 1997-98 and the Mexican crisis of 1994-95 have again demonstrated how vulnerable emerging markets are to sudden and excessive swings in flows of private capital. The problem is a bit like a traffic jam. These build up as the number of cars increases partly because individual drivers do not take into account their personal contributions to congestion (unless there is road pricing). In the same way, emerging countries are vulnerable when the supply of foreign capital soars because private borrowers do not take into account the rise in the marginal social cost of a country's foreign debt.

In this analogy, sovereign credit rating agencies might be thought of like a traffic light: by flashing early warning signs, they help smooth the flow of money. At least, they are supposed to.

In fact, the evidence suggests that sovereign credit ratings are reactive rather than preventive.* As a result, they tend to amplify boom-bust cycles in emerging-market lending. It is rather as if the traffic lights flashed green whenever traffic was building up.

Credit rating agencies were conspicuous among the many who failed to predict the Mexican and Asian currency crises. Having failed to perceive the extent of problems as long as foreign money flowed in, the rating agencies then overreacted by downgrading the affected countries to junk status.

The trouble is that credit rating agencies are not at all separate from the financial markets as a whole. They do not have superior information on emerging-market economies. They have little scope for acquiring advance knowledge of matters that affect sovereign risk. And they share with investors the same views about what determines defaults. This can be seen by looking at the information on which sovereign-risk ratings are based, and the nature of sovereign risk itself.

First, sovereign-risk ratings are primarily based on publicly-available information, such as foreign debt and reserves or political and fiscal constraints. This makes them different from ratings of companies. With companies, credit rating agencies may have access to inside information from domestic corporate borrowers (such as acquisitions, new products and debt issuance plans). Such advance knowledge or better information can then be conveyed to market participants through ratings on private borrowers. This is not the case with sovereign borrowers.

Second, in the absence of a credible international mechanism to sanction a sovereign default the premium charged to reflect the risk of default is determined by a borrower's willingness to pay, rather than by his ability to pay. Borrowers know whether they are willing to pay. Lenders cannot be sure. It is also a problem of enforceability: the authorities cannot give an absolute promise that in future they and their successors will put foreign capital to productive use or that future returns will be used to repay foreign debt.

Moreover, the sovereign rating agencies get most of their revenue from governments to provide a debt rating. Naturally, they are loath to downgrade their clients. This may well introduce 'downgrade rigidity' into ratings especially in periods of large capital inflows.

Unlike with private-sector ratings, then, sovereign ratings can hardly be interpreted as an indication that rating agencies lead the market by conveying new or superior information. Yet sovereign yields tend to rise when ratings worsen. Why are ratings so influential? The answer may well be that herd instinct, often reinforced by poor prudential regulation, give sovereign ratings the power to influence sovereign bond yields even though they add little to the market's information. Many institutional investors may not hold any form of debt security, except investment grade. Hence, sovereign ratings absolve money managers from making independent judgments about sovereign risk.

So reactive sovereign ratings tend to amplify boom-bust cycles. During a boom, improving ratings reinforce euphoric expectations and stimulate excessive capital inflows. During a bust, downgrading adds to panic among investors, driving money out of the country.

So what should be done? The answer is to turn sovereign ratings into proper early warning signals. Since part of the problem is that rating agencies get much of their revenue from borrowers, the industry will have to reorient its fee structure towards investors. Their dependence on borrowers is incompatible with the incentive to come up with timely negative rating. At the same time, prudential regulators should reconsider the role of sovereign ratings that they stipulate when institutional investors hold emerging-market assets. The removal of investment grading requirements for institutional portfolios might attenuate the boom-bust cycle in emerging-market assets. Unless sovereign ratings can be turned into proper early warning systems, they will continue adding to the instability of international capital flows, make returns to investors more volatile than they need be and reduce the benefits of capital markets to emerging countries.

*Merging Market Risk and Sovereign Credit Ratings, by G. Larrain, H. Reisen, and J. von Maltzan, 1997, OECD Development Centre, technical paper No.124, Contact

The author is head of research at the OECD Development Centre