Evan Soltas
May 31, 2012

Switzerland, Take 1.2

Bern Bundesplatz
Considering the response I got to my previous post about Switzerland -- thanks, Scott Sumner and Timothy Lee -- I just want to add a few things to this point about Switzerland, because I believe it's really important that people who read this blog understand what the Swiss central bank's (SNB) currency floor tells us about the strength of the expectational channel in monetary policy.

I did some more research into this, and here's what I found in a May 18 story in the Financial Times:

The SNB has said since September it stands ready to buy as many euros as needed in order to defend the Swiss franc at SFr1.20 against the euro. But it hasn’t been clear that it’s done anything as yet. Big bids have sat in the system – the EBS platform used by FX traders – at the floor, and those have been assumed to be from the SNB, which usually comes into the market directly. Because people generally believe the SNB will defend the floor, no one has really tried to test it...
Woah. Just think about that for a moment. The SNB said in early September, as it solidified its commitment to the currency floor, that "it will no longer tolerate" an overvalued currency and that it will "enforce" its desired exchange rate "with the utmost determination." When all the SNB would have to do is print and sell Swiss franc if it faced a challenge to its currency floor, there's simply no good reason to bet against it. And thus expectations are made. It hasn't had to do anything.

This is not to say that I endorse managed exchange rates in general or in the abstract; but given the fact that Switzerland is a small open economy which has an unusual status as a producer of safe haven assets, I think that this approach probably makes sense.

Prices in Switzerland have gone deflationary, even core CPI, and given that the overnight rates are at the floor and that 10-year Swiss bonds go for a 0.61 yield -- see here -- this is exactly what the SNB should be doing. Their currency is the most overvalued in the world on a purchasing power parity basis, which implies that eventually the Swiss franc will fall in value against other currencies and thus the floor will be made irrelevant and can be scrapped, but until then, conditions in capital markets would have determined the exchange rate of the Swiss franc to the detriment of that country.

Now, in the previous post on this topic I made a comparison between the SNB's exchange rate floor and the long-standing discussion I've had on this blog about setting a path for nominal GDP. I wrote that a bit too quickly, so I want to go back through and explain why that final connection is so important to understanding how the Fed, our central bank here in the United States, does not use appropriately the expectational channel of monetary policy.

I recognize that the feedback loop between central bank action and NGDP is considerably longer than for exchange rates -- when the SNB sells Swiss franc on the foreign exchange market, that immediately pushes down the value of the Swiss franc back where the central bank wants it to be; if the Fed was to commit to an NGDP rule, there is no NGDP market per se -- although Sumner's idea of NGDP futures would go a long way toward that end. So I recognize there could be a less exact maintenance of an NGDP target on a term-to-term basis, although the path targeting, as opposed to rate, will make the small volatility irrelevant because there will be no base drift, but they are comparable because it is eminently in the power of the central bank to print money or to stop printing it. All that is required is credible commitment.

Sometimes this commitment takes a little bit of proving at first, and Nick Rowe over at the "Worthwhile Canadian Initiative" blog memorably described it as central banks having to go Chuck Norris on the doubters. Switzerland has tested and proved Rowe-Norris theory of central bank credibility.

If the Fed was to stay, we are going to buy Treasury bonds, MBS, other currencies, whatever we can get our hands on, until the annualized quarterly change in NGDP hit 5 percent -- it's currently at 3.5 percent and falling -- would, after long, anyone bet against them? What the Swiss experiment tells us is a conclusive no; that is a resounding victory for expectations-based monetary policy.