Guess who won first prize in the ERC economic forecast contest
There are myriad variables for the professionals to get spot-on – but this time my luck was in
What was the most difficult major event to predict this year? The historic winter general election, perhaps, which saw a swathe of northern seats swing to the Tories? If you thought that was surprising, it was child’s play compared to economic forecasting.
Instead of estimating simply which way every adult will vote, most economic variables mean you also have to consider those adults’ likely working, spending and saving patterns, plus their interactions with each other and with people across the rest of the world.
Even working out what happened in the past is difficult – GDP figures are revised for many years down the line, erasing events such as the “double dip recession” of 2012.
That is no reason to stop trying to get the figures right, though – which is why, this year, I decided to have a go myself. After all, forecasting the economy helps the rest of us plan for the year, in a slightly self-reinforcing way. If economic predictions are chirpy, we might feel more able to spend or invest or dish out pay rises – encouraging just such an economic lift. By contrast if everyone anticipates a recession, they cut spending and risk causing a recession.
An entertaining, if nerve-wracking, way to do this is to put on a public competition and hold all participants accountable by publishing errors over a year. The Economic Research Council (ERC) does just that. Back in January it challenged the public to submit four quarters’ worth of GDP, inflation, unemployment, interest rates and average earnings forecasts, plus the year’s low point for the pound versus the euro.
More than 800 entrants tried, putting their good names on the line for glory, and a £500 prize. The reason you are hearing about this competition is because a certain economics reporter managed to bag the top spot in the rankings.
Not that my forecasts were bang on. For each tenth of a percentage point miss in any given data point, the ERC gives entrants one point. The smaller the final number, the better.
In my case, this amounted to 24 points over the year.
That might not sound particularly accurate, but the median score among the forecasting enthusiasts was 53 and the worst an implausibly bad 7070.
Compare it with the professionals who entered. The competition was accompanied by a “clash of the titans” – big names from Goldman Sachs, Royal London and the National Institute of Economic and
Social Research. Respectively, their scores were 29, 30 and 60.
So how did I pip them at the post? In fairness, I had the advantage of seeing the titans’ entries in advance, allowing me to pick a lower number for second-quarter growth – anticipating a politically induced slump effect from the first Brexit deadline. I also bet that most professional forecasts of the jobs market are too gloomy.
If any investment banks would like to hire me on a consulting basis, I can offer very reasonable fees. That said, the purpose of this report is to say that a big chunk is lucky guesswork.
Try as we might – and using all of the supercomputing power the City of London can harness – it is difficult to correctly calculate all of the myriad possibilities affecting a vast, complex and ever-changing economy.
Do not pin all of your financial hopes and plans on any one forecast – and whenever you read “forecast”, maybe substitute in the word “guess”.
And what about next year’s competition? A repeat performance seems unlikely.