Sovereign ratings aren’t very important
Bloomberg has another one of its meandering monsters today, this time under the headline “Austerity Doesnâ€™t Pay as Debt Markets Ignore Rating Cuts”. The 3,300 words can be reduced quite easily to one sentence: Countries are implementing austerity measures in order to bolster their credit ratings and keep their borrowing costs low, but it turns out that credit ratings have no effect on borrowing costs after all.
Bloomberg put a lot of work into this piece: it looked at “314 upgrades, downgrades and outlook changes going back as far as 38 years”, and compared government bond yields on the day of the ratings action to the same government’s bond yields one month later — to allow “time for markets to adjust to assessment changes while minimizing the effects of subsequent unrelated events.” The results show that there’s no real correlation there: 53% of the time rates followed the ratings move, and 47% of the time they didn’t.
This comes as no surprise, for three reasons. Firstly, the news from ratings agencies which causes markets to move is normally when countries get put on watch for a possible ratings change, not when the change actually happens. Markets specialize in pricing in future events, and few if any ratings actions actually come as a surprise.
Secondly, the ratings agencies are, famously, lagging indicators when it comes to bond spreads and yields: a rise in spreads does a much better job of predicting a ratings downgrade than the other way around. The Bloomberg study would have been much more interesting if it looked at the change in spreads before the ratings action, rather than the change in spreads after it. After all, the whole thesis of the Bloomberg article is that governments are using the ratings agencies as a proxy for the bond vigilantes. And in that sense it doesn’t make any difference at all whether the ratings agencies or the bond vigilantes get there first.
And finally, the Bloomberg findings come as no surprise because they were already well known:
In a January analysis of Moodyâ€™s rating changes, researchers at the IMF used credit derivatives to show that prices moved in the expected direction 45 percent of the time for developed countries and 51 percent for emerging economies. For outlook changes, the ratios were 67 percent and 63 percent.
Why did Bloomberg feel the need to replicate the IMF study, which it cites, but doesn’t link to? That’s not clear. And it’s also unclear why Bloomberg doesn’t publish its results in a tractable form: for instance, they don’t give separate figures for the effect of outlook changes on bond spreads. Indeed, while bits and pieces of the methodology are sprinkled through the article, the actual results are confined to a single number — 47% — in the seventh paragraph; no more numbers can be found in the associated graphics, which give anecdotal examples of yields falling after downgrades but no aggregated numbers from the dataset as a whole.
More invidiously, Bloomberg talks about “the austerity policies prized by the rating companies”, without ever mentioning that to a significant degree the ratings agencies actually want less austerity, not more. For instance, here’s S&P in January, announcing a mass downgrade of European sovereigns:
We believe that a reform process based on a pillar of fiscal austerity alone risks becoming self-defeating, as domestic demand falls in line with consumers’ rising concerns about job security and disposable incomes, eroding national tax revenues.
And here’s S&P on Spain, in April:
We believe front-loaded fiscal austerity in Spain will likely exacerbate the numerous risks to growth over the medium term, highlighting the importance of offsetting stimulus through labour market and structural reform.
The fact is that austerity is a political decision more than it is an economic one: fears of bond vigilantes are ex-post justifications for political decisions, rather than genuine reasons for implementing those decisions.
Ultimately, the whole thesis here is a little silly. As former Moody’s sovereign chief Vincent Truglia says at the end of the article, ratings agencies are supposed to pay a lot of attention to governments, but governments really shouldn’t pay much attention to the ratings agencies. And neither, frankly, should market reporters, lest they massively oversensationalize the impact that downgrades can have. This, for instance, is pretty ludicrous:
S&Pâ€™s downgrade of the U.S. last year contributed to a global stock-market rout that erased $6.1 trillion in value between July 26 and Aug. 12.
The fact is that the press tends to care much more about ratings actions than the markets do. The bond markets were undoubtedly lulled into complacency by the dodgy triple-A ratings on structured bonds in the run-up to the financial crisis, but that had nothing to do with upgrades or downgrades. And frankly the idea that Europe’s austerity policies come in response to imprecations from Moody’s and S&P is pretty ridiculous as well.