Scott Sumner  

Brexit: What I got wrong, and what is still uncertain

PRINT
The New York Times Overpaid... A Nobel Prediction...

Immediately after the Brexit vote, I used the knee jerk reactions of various markets to try to estimate the impact. After 14 weeks, it looks like one key market reaction was wrong. Here's the FTSE250, a stock index with a somewhat more domestic focus than the FTSE100:

Screen Shot 2016-10-04 at 9.07.45 AM.png
It looks like the knee jerk reaction was wrong, on this occasion. Why might markets have gotten it wrong? Let's start by looking at the most recent data out of the UK:

Recent data have reinforced optimism that the domestic economy is weathering the aftermath of the vote. The construction industry unexpectedly grew in September, while a manufacturing gauge jumped to its highest level in more than two years. A Citigroup Inc. index tracking economic surprises in Britain is near a three-year high.

After four years of underperformance, U.K. shares have become the highlight of Europe, when looked at in local currency terms. The FTSE 100 has surged 14 percent in 2016, while the regional Stoxx Europe 600 Index is still down 5.4 percent.

"It's more than just a sterling weakness story," Higgins said. "Economic data has actually been strong. There is a bit of a catching up here."


The opponents proponents of Brexit have a right to claim that the "Remains" got it wrong, predicting that uncertainty would cause a recession in the UK. It still might, but that seems increasingly unlikely. If I were a Brexit fan who didn't believe in the EMH, I'd claim that the stock market initially fell because investors believed all the dire warnings coming from elite prognosticators, and then rallied when it found out those warnings were false.

The forex markets are one place where the knee jerk reaction does seem correct, in retrospect. The British pound fell sharply after the vote, and has not recovered at all. It seems likely that the forex markets still expect slower long run growth in the UK, even as the economy has held up in the short run. And here's one area where the Brexiteers may have gotten it wrong---they promised that Britain would sign free trade agreements with the EU, and that option seems increasingly unlikely.

As far as my other predictions, it's too soon to say. I thought Brexit would be a near-perfect test of the "uncertainty theory" of business cycles, and still believe that strongly. It created a massive amount of uncertainty, and most of the experts who believe in the uncertainty theory also predicted that Brexit would cause a recession. This should be a very good test.

Of course the usual caveats about monetary offset must be kept in mind. Fortunately the BoE seemed to react to the shock in such a way as to keep NGDP growth reasonably stable (although we won't know for sure until more data comes in). I predicted only a small (0.5%) rise in unemployment, and thought the bigger impact would be on RGDP. I'm sticking with that prediction for now. If forced to revise my views, I'd probably move even further away from the conventional wisdom on the importance of uncertainty, as the early data suggests that uncertainty might be even less important than estimated by skeptics like me.

To summarize, Brexit is a really important natural experiment, which the entire profession should watch closely. The impact of Brexit on the real economy will take years to determine, as it will not actually occur until at least 2019. The impact of Brexit uncertainty will be clear much sooner--probably within another 4 to 8 months. By then we will know whether "uncertainty theories" of the business cycle can be safely tossed into the trash bin, or can live for another day. After all, it's very unlikely that we'll ever again see such a big uncertainty shock from the supply side. (Of course AD uncertainty is an entirely different story, and shocks like 2008 will continue to occur.)

PS. William Hague's proposed negotiating strategy seems sensible to me, but I don't know if the EU would accept it.

PPS. Nick Rowe has an interesting take on Brexit.




COMMENTS (18 to date)
Andrew_FL writes:
The opponents of Brexit have a right to claim that the "Remains" got it wrong,

This should be proponents, yeah?

marcus nunes writes:

"Confidence doesn´t matter"
http://ngdp-advisers.com/wp-content/uploads/2014/12/Tenet-why-confidence-doesnt-matter-1.pdf

marcus nunes writes:

There´s also an alternative view
http://ngdp-advisers.com/usd-bus-driven-off-cliff-gbp-right-get-off/

Matthew Moore writes:

Is it possible that the referendum reduced uncertainty?

That is, the main uncertainty was Brexit Yes/No, not Brexit Hard/Soft.

I remember thinking that investment would surge after the referendum whatever the result, as the option value of postponing strategic market decisions fell.

Lorenzo from Oz writes:

Don't believe changes in uncertainty are much of a cause of economic downturns.

Though surely part of the analytical problem is uncertainty is both an ambit and a direction phenomenon. That is, there is a question of
(1) the range of uncertainty (how large the information gap)
but also
(2) how uncertainty (the information gap) is construed.

It is not clear that uncertainty is always construed negatively. I would have thought that tech surges are cases of positively construed uncertainty and tech collapses could well be a case of uncertainty narrowing (possibly dramatically narrowing) via increased information, but increased information which suggests much lower returns for particular assets.

By "positively construed uncertainty" I mean positive (income/value) expectations which go beyond available information. (So Knightian uncertainty but with guesses to fill the gap.) Or do we think people do not ever have them to any significant degree?

Lorenzo from Oz writes:

So, Keynes' "animal spirits" are about how uncertainty is construed.

Especially as, in the absence of other information, what other folk are doing is going to be a prime indicator.

Scott Sumner writes:

Andrew, Thanks, I fixed it.

Marcus, I'm increasingly skeptical of that too.

Matthew, It's possible, but it seems unlikely. In any case, I go with what the "uncertainty theorists" think is important.

Lorenzo, I think the big issue here is regulatory uncertainty, which seems quite different from uncertainty in technology. Although I admit I have not given this issue as much thought as I should have.

mbka writes:

Scott,

The impact of Brexit on the real economy will take years to determine, as it will not actually occur until at least 2019

That to me is the real pudding, what the real economy does under real, new constraints. Even these constraints will be limited. As of now they just announced that all EU law will become UK law with the stroke of a pen. I suppose this is to cement UK sovereignty, as big an irony as there ever was.

Uncertainty may well postpone some investments too, but this too is not visible with immediate effect. Only the markets work with immediate effect and they readjust quickly from knee-jerk. Why, there is no new information for now.

I'd be really curious if there's a secondary quake if / when (hard?) Brexit is actually announced in 6 months.

Lorenzo from Oz writes:
I think the big issue here is regulatory uncertainty, which seems quite different from uncertainty in technology.
Regulatory uncertainty is about having insufficient information about what the legal rules are going to be.

Technological uncertainty is about having insufficient information about what the technological constraints and possibilities are going to be.

Since both are about lacking information about constraints and possibilities, it is not clear to me why the impact would be different in kind. Especially given both could potentially have very wide ambits.

Particularly as it is not hard to construct a story, if an existing regulatory regime is bad enough, such that an increase in regulatory uncertainty could be felt to reduce constraints against, and increase possibilities for, profitable transactions.

Either way, it is peregrinations about the standard problem with the future--there is no information from it, just expectations about it which can only be derived from existing information (which never comes from there).

Tim Worstall writes:

"somewhat more domestic "

That's doing quite a lot of work there. FTSE250 is still 50% foreign income rather than the 75% of FTSE100.

But otherwise, yes, largely agreed.

rtd writes:

What about this from Lars Christensen?

https://twitter.com/MaMoMVPY/status/783634552531148801

Brian Donohue writes:

Lars' take is straightforward.

Luca1000 writes:

Regarding monetary policy: "Politicisation of monetary policy or pure conference blather?"
http://ftalphaville.ft.com/2016/10/05/2176635/taxi-for-carney/

Plus lots of talking about immigration:
http://ftalphaville.ft.com/2016/10/05/2176612/fessing-up-to-foreign-workers/

Let's see what will happen.

[broken html fixed--Econlib Ed.]

mico writes:

"It looks like the knee jerk reaction was wrong, on this occasion."

There was a massive conspiracy on the part of institutions to generate propaganda to persuade the British to stay in the EU, which was couched in economic terms but ultimately had non-economic motives.

Some market participants believed this propaganda, and this dumb money exited the market very soon after the vote. The vote of most of the most successful market participants, however, did not follow this dumb money.

Scott Sumner writes:

mbka, I'm increasingly of the view that the Brexit impact will be gradual, not cyclical.

Lorenzo, OK, but then I'd claim that technological uncertainty evolves very gradually over time, due to the "law of large numbers". There are many tiny innovations occurring all the time. In contrast, regulatory uncertainty can be more abrupt, at least in certain cases (say the Russian Revolution.)

Having said that, I'd reiterate that my interest here is not my own views on regulatory uncertainty, but rather the views of those who emphasize it more than I do.

Tim, Good point, although the strongly rally in the FTSE since the late June lows is meaningful in either case.

rtd and Brian, Those UK inflation forecasts always seem a bit high to me. What time frame?

And keep in mind that NGDP growth is not likely to accelerate, and that the uncertainty proponents claimed it was a real shock, even holding monetary policy constant.

Luca, Thanks for the links.

Mico, I'm generally skeptical of "dumb money" arguments, but who knows. The reporters at places like the FT seem pretty sincere to me.

Luca1000 writes:

I've had now a bit more of time to look at the data and I've the following question. FTSE250 in GBP is up 7% since the day before the referendum, but down almost 10% in USD and almost 9% in EUR, even if as Tim said 50% is still foreign income: would not it be better to look at FTSE250 and make comparisons with other equity indexes using a common currency? Even the Italian index FTSEMIB is down by almost only 5% in EUR since, even with all the problems of little growth and of an ailing banking sector. GBP is down 16% since, and down 5% in the last week after it was announced that the government wants to trigger article 50 early next year. And last night there was a flash crash in gbp/usd, something a bit worrying for one of the most traded ccy pairs: liquidity is lacking.

Lorenzo from Oz writes:

Well, I did start by saying I did not believe that uncertainty was much of a cause of economic downturns.

My point is that if uncertainty is a multi-dimensional phenomenon -- having both issues of direction and ambit -- that would tend to undermine blaming economic downturns on uncertainty, since it becomes a bit of a "how long is a piece of string?" question unless one further specifies.

Lorenzo from Oz writes:

Also, not sure that ambit distinguishes between regulatory and technological uncertainty in quite the way you are suggesting. The railway booms and busts of the C19th would be counter-examples on scale. The housing bust can also be seen as a case of technological uncertainty -- if we take financing forms to be a form of technology.

Comments for this entry have been closed
Return to top