Should We Care What Debt Rating Agencies Say About National Governments?

11 05 2010

The financial turmoil in Greece has focused attention on bond markets and credit rating agencies. The yields on Greek government bonds have shot up, meaning that it costs the Greek government to borrow money because investors think that a default is fairly likely.

The election of a minority parliament in Britain has created fears that that the agencies that rate the creditworthiness of sovereign debtors could downgrade UK bonds from the current AAA status. The Guardian’s Katie Allen reports: “The prospects of a weak coalition government has rattled UK markets, sparking growing fears that a downgrade to Britain’s coveted top-notch credit rating could follow.”

There have been repeated references to bond-rating agencies in the discussion of the various proposed coalitions in Britain. There was relief when one of the bond-rating agencies announced that the hung parliament would not affect the UK’s credit rating. “The fate of the UK’s gold-plated sovereign debt rating will not be decided until the end of 2010 despite a hung Parliament, Standard & Poor’s said.”  See also here.

Political commentators are quite right to pay attention to the reaction of the bond markets to political events. In the last few centuries, capital markets have proven to be a sensitive register of political and even military events.  For instance, during the American Civil War and the Second World War, the prices of the combatant governments’ bond fluctuated as news from the battlefield arrived. Financial markets can be spectacularly wrong when it comes to predicting the future (as the Sub-Prime Mortgage Crisis suggests), but overall they are useful tools. Capital markets exploit what James Surowiecki calls “the wisdom of crowds” because they aggregate the intelligence of large numbers of people. They also rely on the self-interest of investors: because people are putting money on the line, they are more likely to put aside sentiment (what they want to happen) and focus on what will happen.

The price of Confederate government bonds reflected events on the battlefield during the American Civil War. Image Courtesy Vanderbilt University Library.

However, I’m not certain that we should pay any attention at all to the bond-rating agencies, as they have had a spectacularly bad track record of predicting defaults by sovereign and other major borrowers. Just a few weeks ago, executives from a credit rating agency testified before Congress about their role in the sub-prime mortgage crisis. The rating agency is accused of having given unjustifiably high ratings to the mortgage-backed securities that were at the heart of the sub-prime mortgage crisis.  Gullible investors saw the ratings and then purchased the securities. When the dead-beat homeowners who were the other end of this complex financial chain defaulted on their mortgages, people came to realise that the securities were not nearly as safe as had been made out. Just today there was word that Moody’s, one of the bond-rating agencies, is being investigated by the SEC.

Much of the criticism of credit-rating agencies in recent months has revolved around the conflict of interest inherent in the agencies current business model, whereby borrowers rather than lenders pay the agency to issue ratings.  To my mind, this is only one of the problems with ascribing too much credibility to the statements that come out of bond rating agencies. Here are some other big problems with having credit rating agencies assign ratings to the debts of governments in large industrialized countries.

1)      Decisions are credit-rating agencies are made by relatively small teams of people, so you don’t have the benefit of the “wisdom of crowds” effect.

2)      There are currently few penalties for making a rating that turns out to be inaccurate. A bad rating agency is unlikely to lose business since the business of rating sovereign debt is very oligopolistic, meaning it is dominated by only a handful of firms. There is little competition, which means that a rating agency can deliver inaccurate ratings without fear of losing business to other firms. Moreover, the legal penalties for making an inaccurate prediction are unclear. It is true that the California state employees’ pension fund and other investors who lost money in the sub-prime mortgage meltdown are trying to sue the relevant credit rating agency, but it is unclear whether this lawsuit will be successful. The rating agency can always deflect charges that its rating was dishonestly high by pleading that they made the rating in good faith and that no human being is infallible.

3)      The agencies that rate sovereign debt are supposed to compete with each other, but it appears that they cooperate in rating countries.  Martin Weiss, the head of one such agency,  recently published an open letter calling on Standard and Poor’s, Moody’s, and Fitch to downgrade U.S. government debt. See here.

4)      Can political bias influence a rating agency’s statement? I wouldn’t want to question the professional credibility of Martin D. Weiss or anyone else as a rater of bonds, but it would be interesting to know if he is a registered Republican, a registered Democrat, or an Independent.  Being based in the academic world, I am very familiar with the phenomenon of professors who tailor their lectures and research to suit a political agenda. With historians, this is particularly likely when it comes to people dealing with recent periods of history and the histories of their own countries.  Academics can allow ideology to bias their judgement because there are few penalties for doing so.  The uncompetitive nature of the rating-agencies game and the fact there are few clear legal penalties may give a similar sense of licence to the handful of individuals who rate sovereign debtors. The fact that Weiss is an American means that his comments about the US government are likely to be coloured by his personal background. Perhaps ratings of sovereign creditworthiness should only be made by non-citizens of the country in question. Swiss experts should rate the US, and US experts should rate the Swiss government’s chances of defaulting. I must also say that Americans seem to get more emotional about their politics than the Swiss– this needs to be taken into account in thinking about how Americans speak about their own country’s chances of defaulting.

5)      Sovereign bond rating agencies haven’t been around for that long. According to Timothy J. Sinclair, the author of an excellent book on bond-rating agencies, rating agencies only really began rating government debt in the 1970s and 1980s. The fact that sovereign bond ratings were not around before the great defaults of the early 20th century makes it hard to tell whether ratings of sovereign debt are any good.

6)      The social, political, and monetary consequences of a default on government debt by a major industrialized country would be so massive as to destroy the legal and economic foundations of the bond-rating agency. To put things in human terms, I’m not certain what life would be like for a bond-rating agency employee in New York or London in the event of the collapse of the US government’s ability to pay interest on its debt.  Would the legal system still be in operation? Would paper money be worth much? Or would firearms and tinned food be more valuable in the new environment? How would the bond-rater be able to get off Manhattan Island?

This means that the ratings (i.e., predictions) issued by the agency aren’t worth that much. It’s like asking a bookmaker to tell you the odds of a nuclear war that destroys all live on earth. He is free to promise to pay you pretty much any sum of money he can think of in the event of such a war, because you won’t be around to collect it. Rating agencies may do an adequate job of predicting the chances of default on the part of a particular homeowner with a mortgage, company’s bonds, or even a small country like Greece.  I’m not certain rating agencies are equipped to predict the probability of truly catastrophic events of the sort that might wipe out the agency. Similarly, prediction markets can’t predict events that would wipe out the prediction market itself. This is the major problems I have with the prediction-market concept advanced by Robin Hanson. Prediction markets can’t deal with the once-in-a-lifetime Black Swan events.

I have one other point—Timothy Sinclair should complete the promising-looking website on ratings agencies he has started.

Article in Guardian About Ignatieff

2 06 2009

Yesterday’s Guardian carried a piece by Michael White comparing current Canadian and British politics. It is rare to find an article that comments on both the UK MP expenses row and the Tory attack ads.

If only Ignatieff had a moat that needed cleaning, that would make for a great attack ad.