This post is for Giles Wilkes at the FT, who wanted a follow-up to my Tweets about the falls in Sterling reported as markets opened. We might guess that this fall is the reaction to the news that Boris Johnson had declared that he would campaign for the UK to leave the European Union. In that respect the news provides quite a clean experiment to identify the effects of Brexit, since not much else has changed since Friday.
The standard macro-finance way to view this would be as follows.
There are two forces pushing down on Sterling.
The first is that the perceived risk premium on Sterling assets increases. This could happen for a number of reasons. But an illustrative line of logic is: Boris campaigning for out increases the chance that the UK will exit the EU. Exit is an as yet ill-defined long-run state and there are many possible outcomes, ranging from not much different to In, to much worse. Aside from the long-run uncertainty, exit also involves a series of event risks, including the referendum itself, and key milestones along the exit negotiation journey. At these points, there will be heightened concerns about many things including self-fulfilling runs on markets, akin to – and perhaps including – old-fashioned bank runs. The effect of this uncertainty is to make holding Sterling denominated assets less attractive, hence the price of them has to fall to leave investors indifferent.
The second force pushing down on Sterling is the effect of a projection of lower long-run GDP per head. It’s possible to conceive that exiting raises long run prosperity, but, in expectation – and, I maintain – in the expectation of markets, GDP per head will fall somewhat. With less income, the money available to compete for non-traded goods in the UK falls, and their price falls. This lowers the real exchange rate, which is some complex weighted average of traded and non-traded prices. A lower real exchange rate, other things equal, lowers the nominal exchange rate.
This is the Balassa-Samuelson effect, or, as I rather unfortunately abbreviated it in the first draft of my specimen international macro exam, ‘the BS effect’.
Three final remarks.
First, even if things don’t actually work like this, it may be enough that people think they will for these things to happen.
Second, my ‘BS effect’ joke should be a reminder that many things in the macroeconomics of exchange rates remain puzzling to macroeconomists. In fact, there is an industry in generating ‘puzzles’ about exchange rates, where a puzzle is defined as a very stark difference between the implications of a macro model that is held dear and the data. The most pertinent is called ‘the exchange rate disconnect puzzle’, which, as the name suggests, documents how the exchange rate is disconnected from things that you’d expect to determine the exchange rate in a standard macro model. So this reasoning has to be taken with a pinch of salt.
Third, it is actually possible to reverse the correlations described above with determined spanner-work on the models’ innards, but I’d say that the correlations I highlight are where most macro people would start.