How Can We Build A Model In Which An Invisible Japanese Bond Vigilante Attack Is Contractionary?

by: Brad DeLong

Let me gnaw on Paul Krugman's observations on Japan and on his and Ken Rogoff's Mundell-Fleming lectures some more:

Paul Krugman:
Shinzo and the Invisibles:

Brad DeLong is puzzled by… Ken Rogoff['s]…

…warning that Japan could face an attack from invisible bond vigilantes if it doesn't quickly tackle long-run fiscal issues. I'm puzzled too… The truth is that I said such things about the US back in 2003. But I was wrong…. Rudi Dornbusch's 'overshooting' model…. Invisible bond vigilantes. Suppose… they suddenly demand that Japanese 10-year bonds offer a rate of return 200 basis points higher than US 10-year bonds. You might be tempted to say that Japanese interest rates will spike-but the Bank of Japan controls short-term rates, and long-term rates are mainly an average of expected short-term rates, so how is this supposed to happen?… Instead, the yen would depreciate now so that investors can expect it to appreciate later. And this yen depreciation would be expansionary…. The invisible vigilantes would be doing Japan a favor if they suddenly materialized and attacked!

I've had many discussions with smart people about this, and have never gotten an explanation of why it's wrong; we usually end up with something like a warning that Japanese deflation might suddenly turn into uncontrolled inflation, which seems unlikely and certainly isn't the way the warnings are usually phrased-we're supposed to worry about turning into Greece 2010, not Weimar 1923. You might think that what we're talking about is the lessons of history-but as far as I can tell, there are no historical examples of countries with debts in their own currency facing a Greek-style crisis…

I confess that I too am completely flummoxed. But it is time to get I'm flummoxed. So let us set up 2 x 2 tables with four boxes: fixed exchange rates and floating exchange rates along the side, sticky prices and flexible prices along the top.

  1. Fixed exchange rates and sticky prices: This is Greece. The bond vigilantes demand a higher real return. Prices and exchange rates are both fixed. So the bond vigilantes get a higher real return in the fixed-value foreign numeraire, a higher real return in domestic currency, and a higher nominal return in domestic currency. This is a deficient-demand depression.

  2. Fixed exchange rates and (largely) flexible prices: The bond vigilantes demand a higher real return, and so go on a capital strike against funding domestic investment projects until they get it. An excess supply of labor and goods drives domestic nominal prices and wages down until the expected real appreciation of the currency pushes returns high enough to satisfy the bond vigilantes and they cease their capital strike against funding domestic investment projects. But in the meantime - while the capital strike against funding domestic investment projects continues - the economy is in a deficient-demand depression.

  3. Floating exchange rates and (largely) flexible prices: The bond vigilantes demand a higher real return. The real value of the currency bounces down so that thereafter real appreciation supplies it.

The problem appears to be on the nominal side. The nominal value of the currency falls as the shock hits. But domestic prices and wages measured in domestic currency rise as import prices rise. In order to reach equilibrium, the value of the currency has to fall farther and faster than domestic prices and wages rise in order to lower the real value of the currency and so set the stage for the real appreciation we need to anticipate in order to be in equilibrium. That process - the rise in domestic nominal prices and wages, and the larger fall in the nominal value of the currency - may derange the price system and so disrupt aggregate supply.

The new equilibrium may be one in which the real depreciation of the currency is expansionary in the sense that it tends to push real aggregate demand above potential output. But the economy may nevertheless be in depression, if the process of getting to the new equilibrium has entailed nominal price swings large enough to have been sufficiently disruptive to the market-mediated division of labor. Weimar 1923.

In order to avoid such supply-side disruption to the market-mediated division of labor, the central bank may respond to such a shock by raising domestic real interest rates to generate a deficient-demand depression in order to moderate the rise in the domestic nominal value of prices and wages, thus moving partway back to case (2). The problem is that case (3) is one of extreme nominal devaluation coupled with domestic inflation, while case (2) is one of nominal fixity in currency values coupled with domestic deflation, and it is not clear how the halfway house fits together.

  1. Floating exchange rates and (largely) flexible prices: This is the Dornbusch base-case: The bond market vigilantes bounce the nominal exchange rate down until expected appreciation is sufficient, and so the attack of the bond market vigilantes is expansionary. France 1923-1927.

Note that these chains of argument do not rest too heavily on the assumption of rational expectations in exchange rate markets. If expectations of exchange rates are static, the real depreciations needed to restore equilibrium when the bond market vigilantes attack are larger: they need to provide higher returns not through expected future capital appreciation but through the fact that profit margins on exports are very large if your currency is substantially undervalued in real terms. But it is still a true fact that in the base, Dornbusch, case, an attack of the bond market vigilantes is expansionary.

Now we may not be in the base, Dornbusch, case. Raghu Rajan, for example, clearly thinks that India is not. He believes that India is in case (3): that if the bond market vigilantes attack he cannot let the nominal value of the currency bounce down until real appreciation satisfies them without unleashing a disastrous domestic wave of inflation. Perhaps he thinks that Indian domestic nominal prices and wages are tightly coupled to import prices, so that the gearing between nominal depreciation and real depreciation is very loose. Perhaps he thinks that international financial speculators-the Gnomes of Zurich… no, that's not right any more, the Djinn of Dubai Jim…-have, over some considerable range, not rational but rather extrapolative expectations of exchange rate changes so that the required real depreciation would itself be so large as to temporarily destroy India's integration into the international division of labor.

The fact that Raghu Rajan does not think that Krugman's (and my) Dornbusch logic applies to India today gives me pause. But, even with pause, I cannot see for what reasons Krugman's (and my) Dornbusch logic would not apply today to Japan.