Brexit shows the value of scenarios

When the result of the UK referendum on exiting the EU (Brexit) was announced, the FTSE 100 immediately fell by 8.7% and sterling fell around 8% relative to the dollar, and down 5% against the Euro. Clearly, markets were surprised. But why? Even though opinion polls and betting markets indicated that "Remain" would win, it was hardly like Leicester winning the Premier League. I think the problem comes down to mindset.

The debate between opinion polls and prediction markets is essentially a debate about forecast techniques. Although useful, they drive attention to narrow outcomes. Billions of pounds were clearly hinging on a a few percentage points spread. This implies that in future we need better quality opinion polls and better functioning prediction markets, and lots of money can be made from better probability values.

However I'm pleased that the failure of forecasting has generated increased attention and utilisation of the scenario method. A scenario planner doesn't care what the probability of Brexit is. We simply didn't know. What we did know, was that one of two possible outcomes could happen - a Leave vote, or Remain. Therefore plans should be made around both. As an economist, I was routinely asked in the buld up to the vote "do you think we'll leave?" This is based on a forecasting mindset. I tried to say "I don't know", but it's hard not to weigh in with a (flawed) prediction. The better question would be "what should we do if we leave, and what should we do if we remain?"

Now that the result is in, the uncertainty is not over. The manner in which Brexit will occur, if at all, is the topic for discussion. And thus far no one has asked me "what do you think the most likely arrangement will be?" At heightened uncertainty, you only really have scenarios. Hence newspapers are discussing "The Norway option" or "Article 50 isn't triggered" or "Scotland has another referendum". Scenarios are the go to framework.

I don't meen to disaparage polls or markets. Both tell us different things, and are useful. But it's clear that too much money was riding on their predictive power, and we should be humbled by that. I am encouraged that people are thinking in terms of scenarios, and hope the economics professsion does the same.