When British voters decided in July 2016 to leave the European Union, virtually every governmental economic unit, bank, and private forecast predicted a disaster for the UK. The reality has been much different.

BREXIT REALITIES

Since the vote, the British economy has hardly been in freefall.

• The number of people working in the UK stands at a record high.
• New orders for manufactures are at the highest level in a generation.
• Employers are struggling to find enough staff.
• Equity investments are at record highs.
• Housing prices are at record highs.
• Foreign investment is at an all-time high.

“For a country supposedly crawling out of the ruins of the Brexit vote, the UK has been having a strikingly good year so far,” Fraser Nelson wrote in The Wall Street Journal.

Added Brexit minister David Davis, “Every forecasting model on the performance on the British economy post the referendum by every major organization, the banks, the government organizations and, indeed, international organizations has proven wrong.”

The question, then, is this: Why were so many forecasts about Brexit so wrong?

For finance professionals, market forecasts are a key starting point in developing our own business models, so a miss translates into a miss in our own forecasts. Here are a few theories on why the forecasts were so off; it may be the interaction of these as well as others are in play.

Theory #1: Macro forecasting is hard

Forecasting something as big and complicated as an economy is hard; market forces are the byproduct of countless small decisions by individuals and are difficult to forecast. However, the Congressional Budget Office noted in 2015 that on average, its two-year forecast of GDP growth missed the actual rate by 1.4 percentage points. It also noted that this was significantly better than Wall Street consensus projections and other governmental outlooks. A 2000 IMF paper noted that, globally, forecasters predicted only two of 60 recessions.

On this theme, there is also a theory that economists were aiming for a redemption after largely missing that last great macro-economic event, the 2008 financial crisis. Here was a visible black swan, an opportunity to predict a significant new trend and redeem the profession, and so they trained their analysis on the big impact to come. Except that it did not arrive.

Theory #2: Fear and bias behind the numbers

We must consider whether there was a bias in the forecast based on a fear of such a major change. Many of the forecasts came from institutions with an integrative world view, including the International Monetary Fund and the UK government in power at the time. What looks like such a significant change in outlook must have severe ramifications, right? The economists predicted a panic-based reduction in citizen spending—which did not happen. Cynics and supporters of Brexit dub the Remain in the EU campaign “Project Fear” and claim the forecasts were the triumph of political motivations rather than well-executed economics. Jacob Rees-Mogg, British MP, said, “Many forecasts are honestly wrong, but unfortunately many of the forecasts around Brexit were politically motivated and have undermined trust in the probity and impartiality of those making them.”

Theory #3: Trend-breaks are hard to predict

Why did so many miss the recession of 2008? Why do people think trends will continue? Forecasting in calm waters is easy, but identifying the black swan event, or the turn in the market, is notoriously difficult. Prognosticators of many stripes look back on the “mega trends” that have changed the world in the past fifty years and want to get ahead of the next one. Brexit had the appearance of being the harbinger of a mega trend, a rare case when a potential turning point was visible to all, but perhaps it was just an event. Not all indicators are right, as Nobel Prize-winning economist Paul Samuelson quipped when he noted that “the stock market has predicted nine of the past five recessions.”

Theory #4: Limitations of models

Models are driven by inherent assumptions to model what we think we know, but there may be other factors under the surface that are not visible in a calm environment. Major shifts suddenly “awaken” these other factors and their changes may be outside of the model

Steven Baker, the Brexit minister, said the civil service analysis “does not yet take account of the opportunities of leaving the EU.” Perhaps there are other items gaining strength, or perhaps the correlations or strength of certain factors are incorrect at this time.

A Cambridge University review of the UK Treasury models came to several conclusions about variables that did not behave as predicted. Interest rates went down rather than remain unchanged. The value of the British pound dropped initially, which boosted exports (it has since recovered as the economy remained healthy), and several other assumptions that simply were “wrong.”

Theory #5: Sticky qualitative factors

Most forecasts were created by models and simulations in which various quantitative inputs become fodder for calculations. But what if other factors predominate? London was predicted to lose 100,000 sophisticated, high paying financial jobs. But it seems that only 3,000 jobs have been lost. Duplicating the most powerful financial center in Europe is hard—it has advantageous time zone, language, fintech, and is a desirable place for the financiers to live. Deutsche Bank reversed itself and instead of moving 4,000 employees out of London, it expanded its footprint at a new headquarters in The City.

The details matter

The terms of the Brexit can make a big difference in the ultimate impact. Nelson notes several places where the separation anxiety may be overstated: Britain could negotiate a free-trade agreement with the continent. If negotiations lead to an impasse, they could be hit with a 4.2 percent tariff under World Trade Organization rules. Some jobs may move as subsidiaries to London headquarters, but information flows freely through borders. The Europeans have an incentive to negotiate a punitive agreement with the UK to dissuade others from leaving, but they also may have an economic incentive to negotiate better terms that facilitate trade.

AFP 2018 has multiple sessions on forecasting in the FP&A track.