Carmen Reinhart and Ken Rogoff have a very peculiar op-ed in the WSJ today, warning of the risk that there will be another wave of emerging-market sovereign bond defaults. Now Rogoff, a former director of the IMF’s research department, is the kind of person who tends to know what he’s talking about. So let’s just say I’m puzzled by his attitude here, and would love him to clear up a couple of things.
Reinhart and Rogoff are certainly provocative: "Already, a good share of Argentina’s debt is in default," they write, and they’re not talking about untendered pre-restructuring debt:
What else do you call it when a government that owes over $30 billion in inflation-indexed debt manipulates its consumer-price statistics? Through a variety of crude measures (such as firing its top statisticians), the government is publishing an understated inflation rate that is used for calculating indexation payments.
Well, I, for one, don’t call it default, I call it manipulation of consumer-price statistics. Someone buying inflation-indexed debt runs both default risk and statistics-manipulation risk; the latter is not a subset of the former. There’s a big debate raging in the US about whether and how the US CPI might understate "actual" inflation, however that might be defined, but no one is suggesting that if CPI is understated then TIPS are somehow in default. And default is a state of affairs, not a state of mind: you can’t say that Argentina is in default just because it is deliberately manipulating its statistics.
But the idea that CPI fudging is a form of default seems to be key to the rest of Reinhart and Rogoff’s argument:
Governments do not usually cheat holders of only one type of debt. In April, we published a National Bureau of Economic Research paper based on centuries of debt data from many countries. We found that most countries default on external debt only a bit less freely than on domestic debt.
So my first question is: does the April paper consider fudged inflation statistics in the presence of index-linked debt to be the same thing as a default? Somehow I doubt it, I can’t see how it would be possible to come up with a rigorous metric of inflation fudging.
I can quite readily believe that outright default on domestic debt is a good indicator that the country in question is likely to default on external debt as well. But I’m not convinced that the same is true of inflation-fudging, which, at the margin, is more of a means of avoiding a default on domestic debt.
Reinhart and Rogoff then make a huge leap from Argentina, with its fudged inflation statistics, to any country with high inflation:
Considering the duress of domestic bond holders across the world as global inflation rises, it is surprising that both private investors and multilateral international financial institutions seem so complacent about the rising risks of defaults on external debts.
I really don’t understand what this means. Yes, inflation is a good way of reducing a government’s real debt burden, especially if a lot of that burden is in the form of bonds carrying a nominal interest rate. Insofar as inflation is higher than investors expected when they first bought the debt, domestic investors especially will get a lower real return than they had hoped for. But again, this is inflation risk, not default risk. Indeed, my intuition is very much that inflation can be used to "inflate away" nominal debts and thereby avoid default. I really don’t understand the logic here; how does rising global inflation mean a rising risk of default on external debt?
But maybe it’s not rising inflation that Reinhart and Rogoff are worried about, so much as rising domestic debt:
There have been many episodes in the past where rising levels of domestic debt have sharply raised risks to external debt holders. There is nothing new about the rise of domestic debt markets. They are simply growing again after a bout of suppression during the high-inflation 1980s and 1990s.
Well, are levels of domestic debt rising? In real terms? I understand that not all emerging-market countries are commodities powerhouses with massive balance-of-payments surpluses which have already achieved net creditor status (Brazil). But I also haven’t seen much evidence that levels of emerging-market sovereign domestic debt are rising particularly quickly. Still, Reinhart and Rogoff do have one example of the kind of thing they’re talking about:
When India effectively defaulted on its domestic debt through massive inflation and financial repression in the early 1970s, external debt holdings suffered payment reschedulings even though they constituted only a tiny fraction of overall debt.
So we’re back to Argentina, it would seem – we’re not worried about "global inflation" rising a point or two, but rather about countries with "massive inflation" – inflation so massive, indeed, that it constitutes an effective default.
If Reinhart and Rogoff were making a narrow argument – that Argentina’s inflation constitutes an effective default on its domestic debt, and that having defaulted on domestic debt the country is unlikely to have a lot of qualms about defaulting on its external debt – then I wouldn’t really have a problem with the op-ed. I’m not sure I’d buy the argument completely, but at least it would make a certain amount of internal sense. And with Argentina’s external debt trading at 570bp over Treasuries, there is indeed quite a lot of default risk priced in already.
But Reinhart and Rogoff seem rather to be making a much broader argument, that inflation globally is a warning sign that emerging-market sovereign debt generally could be facing a new wave of defaults. And the case for that seems to me to be much weaker.