What the SNB Teaches Us About Off Equilibrium Threats

One explanation for why the SNB went off the exchange rate peg was because its balance sheet was becoming too large. Even though balance sheet losses at central banks should have no theoretical effect on policy, Tyler Cowen argues that they do matter because balance sheet considerations serve as a political limit to how far central banks can go in their interventions.

The SNB’s failure to maintain their exchange rate peg can give us some insight into why unconventional monetary policy at the zero lower bound can be so difficult. In particular, it shows that the “off equilibrium threats” that are required to make monetary policy effective at the zero lower bound might be too large for central banks to stomach.

Apriori, it’s hard to imagine why central bank asset purchases can fail to push prices and output around. In Bernanke’s criticism of the Bank of Japan, he noted that if it were the case that domestic prices didn’t change in response to asset purchases, then the central bank could buy up the entire stock of domestic assets! In the limit, this would mean that the central bank could fund unlimited transfer payments. If that were really the case. This is patently absurd — supply side constraints realize themselves eventually—and so by contradiction it must be that at some point the central bank must be able to affect prices, and that in particular it can realize any arbitrarily high price level.

The international macro version of this argument is that a central bank must always be able to depreciate its own currency because otherwise, the central bank could buy up the entire global stock of assets.

But these results are only asymptotic. For any finite asset purchase, there’s no guarantee that the central bank will have an effect. And therein lies the problem. The frictions that balance sheet policies exploit are small, and the central bank needs to be able to commit to extremely large asset purchases to have a large enough direct effect. To the extent that smaller purchases can have an effect, the only way they do so by operating on the expectations of future policies.

To formalize this, imagine that the central bank and foreign investors are playing a game. The central bank can choose between small interventions and large interventions, whereas the foreign investors can choose between going long the currency (and thus causing it to appreciate) or going short the currency (and thus causing it to depreciate).

If the central bank does a large intervention while foreign investors go long, the central bank can inflict infinite pain by buying up all of their real assets. In the end, all of the world’s assets would have been acquired while the foreign investors sit on piles of useless paper. However, this causes the central bank to suffer any political costs that come along with mass asset purchases.

So the central bank might want to do a small intervention. But if it’s not large enough, and expectations for inflation and exchange rates do not change, then foreign investors can continue going long the currency and thereby limiting the effect of monetary policy.

The central bank is in a dilemma. While it would prefer a small intervention, it needs to be willing to threaten a large intervention, or else foreign investors would not budge. But if political constraints are large enough, then the threat will not be credible and small interventions will have limited effect. For the central bank to ensure that the small intervention — short the currency world to be an equilibrium, it needs to have a credible “off equilibrium” threat that the central bank will buy up the world in the case that foreign investors do not short the currency as the central bank desires.

The conclusion is that when the SNB abandoned its peg, it lost its credibility to enforce these off equilibrium threats. This does not bode well for its future attempts at monetary policy at the zero lower bound.