Liquidity Traps, Local and Global (Somewhat Wonkish)
APRIL 1, 2015 6:12 PM April 1, 2015 6:12 pm
There’s been a really interesting back and forth between Ben Bernanke and Larry Summers over secular stagnation. I agree with most of what both have to say. But there’s a substantive difference in views, in which Bernanke correctly, I’d argue, criticizes Summers for insufficient attention to international capital flows – but then argues that once you do allow for international capital movement it obviates many of the secular stagnation concerns, which I believe is wrong.
As it happens, the role of capital flows in the logic of liquidity traps is an issue I tackled right at the beginning, back in 1998; and I’ve been trying to work out how it plays into the discussion of secular stagnation, which is basically the claim that countries can face very persistent, quasi-permanent liquidity traps. So I think I may have something useful to add here.
Start with Bernanke’s critique of Summers. The most persuasive evidence that the US may face secular stagnation comes from the lackluster recovery of 2001-2007. We experienced the mother of all housing bubbles, fueled by a huge, unsustainable rise in household debt – yet all we got was a fairly unimpressive expansion by historical standards, and little if any inflationary overheating. This would seem to point to fundamental weakness in private demand. But one reason for the sluggish growth in demand for U.S.-produced goods and services was a huge trade deficit, the counterpart of huge reserve accumulation in China and other emerging markets. So Bernanke argues that what Summers sees as evidence of secular stagnation actually reflects the global savings glut.
That’s a good point. But Bernanke then goes on, as I understand him, to argue that international capital flows should solve the problem of secular stagnation unless if affects the world as a whole, because capital can seek higher returns abroad; he also argues that the global savings glut is mainly a thing of the past, and that in any case such problems can be addressed mainly by putting pressure on foreign governments to open their capital accounts and stop pursuing policies that promote excessive current account surpluses. And in these assertions, I’d suggest, he goes somewhat astray.
Let’s first ask whether the possibility of investing abroad obviates the problem of liquidity traps in general (as opposed to secular stagnation.) To do this, consider the analysis I laid out a couple of months ago when I was trying to think about the dollar and U.S. recovery, but run it in reverse. Suppose that a country or currency area – let’s call it Europe — suffers a decline in aggregate demand. And suppose that this threatens to push the Wicksellian natural rate of interest – the rate of interest at which desired savings and investment would be equal at full employment — below zero. Can this happen if there are positive-return investments outside of Europe?
You might think not: as long as there are positive-return investments abroad, capital will flow out. This will drive down the value of the euro, increasing net exports, and raising the Wicksellian natural rate. So you might think that you can’t have a liquidity trap in just one country, as long as capital is mobile.
But this isn’t right if the weakness in European demand is perceived as temporary (where that could mean a number of years). For in that case the weakness of the euro will also be seen as temporary: the further it falls, the faster investors will expect it to rise back to a “normal” level in the future. And this expected appreciation back toward normality will equalize expected returns after a decline in the euro that is well short of being enough to raise the natural rate of interest all the way to its level abroad. International capital mobility makes a liquidity trap in just one country less likely, but it by no means rules that possibility out.
Still, secular stagnation – as opposed to liquidity-trap analysis in general – is concerned with excess saving that lasts a very long time, that’s quasi-permanent. So in that case wouldn’t we expect capital mobility to be decisive? Shouldn’t it be impossible to have secular stagnation in just one country?
When I first approached this issue, that’s what I thought. But I immediately ran up against a big real-world counterexample: Japan. Japan has effectively been at the zero lower bound since the 1990s, and it wasn’t until the end of 2008 that the rest of the advanced world joined it there. So why didn’t capital flood out of Japan in search of higher returns, driving the yen down and boosting Japan out of its trap?
The answer is that real returns in Japan weren’t exceptionally low – they were, in fact, more or less equal to those abroad. But this equalization of real rates didn’t occur through an equalization of Wicksellian natural rates. Instead, what happened was that persistent deflation in Japan, combined with the zero lower bound, kept the actual real interest rate well above the Wicksellian rate. Here’s the data from 1996 to 2008, with inflation measured by the GDP deflator and interest rates measured by 6-month Libor:
OK, it’s not full equalization. but not too far off — despite being at the zero lower bound for many years, Japan ended up offering more or less competitive real returns.
The moral of the Japanese example is that if other countries are managing to achieve a moderately positive rate of inflation, but you have let yourself slip into deflation or even into “lowflation”, you can indeed manage to find yourself in secular stagnation even if the rest of the world offers positive-return investment opportunities.
Which brings me to the future of the global savings glut.
As Bernanke notes, as far as big current account surpluses go, Germany is the new China. However, he argues that the large current account surplus of the euro area as a whole is a temporary phenomenon driven by cyclical weakness in the euro periphery, and therefore not likely to be a source of persistent trouble.
But look at what bond markets are saying! German interest rates – presumably an indicator of perceived euro safe rates – are negative out to seven years; the 10-year rate is only 16 basis points. This is telling us that markets expect the euro area economy to be depressed, and ECB rates very low, for many years to come. In effect, European bond markets are flashing a secular stagnation warning.
And this makes sense: the case for secular stagnation in Europe is considerably stronger than it is for the U.S.. Working-age population is declining, Japan-style; the euro system, with a shared currency but no fiscal integration, arguably imparts a strong contractionary bias to fiscal policy; and core inflation is already down to just 0.6 percent.
By the logic I’ve already laid out, this should imply a persistently very weak euro and a persistently large European current account surplus, as Europe in effect tries to export its secular stagnation – a process limited only by the way low inflation or deflation interact with the zero lower bound to keep interest rates from falling to their Wicksellian equilibrium rate.
Sure enough, if we try to figure out the market’s implied prediction for the euro, it seems to imply persistent weakness. The euro/dollar rate is down around 30 percent from its level before Europe began running such large surpluses. Meanwhile, German 10-year real interest rates are around -1, while U.S. real rates are slightly positive; this implies that markets expect the euro to recover only a third or so of its recent decline over the next decade.
What this in turn implies is that even if you downplay domestic U.S. weakness and focus on the export of capital from other countries, especially Germany, there’s no good reason to believe that this new version of the global savings glut will end any time soon.
Which brings me to the policy debate. Bernanke seems to be saying that if there is a problem, it can be solved by cracking down on currency manipulation:
[T]he US and the international community should continue to oppose national policies that promote large, persistent current account surpluses and to work toward an international system that delivers better balance in trade and capital flows.
If my analysis of the European problem is right, however, this is pretty much irrelevant: Europe’s trade and capital imbalances are the result of fundamental weakness of domestic demand, which is then exported to the rest of us, who aren’t that strong either. If true, this says that we have a problem that must be solved with policies that boost demand. So on the policy debate, I come down firmly on the Summers side.
krugman