Consequences of group-think
For forecasters, what goes up will come down
by Brian Reading in London
Fri 13 Jan 2017
Speaking to the Institute for Government on 5 January, Andy Haldane, chief economist at the Bank of England, admitted that economic forecasting was in crisis. He called the failure to predict the 2008 financial crisis a ‘Michael Fish’ moment, referring to the 1987 instance when BBC weather forecaster Michael Fish famously declared that a hurricane would not hit Britain. It hit the next day, Monday 19 October, and closed the City.
Before the UK-EU referendum in June of last year, economists predicted a sharp downturn in the second half of 2016 following a vote for Brexit. Instead, the economy remained buoyant, possibly a timing error. Computer screens and silicon chips have replaced crystal balls and tea leaves, but something is clearly wrong with economic forecasting.
The International Monetary Fund conducts post-mortems on its forecasts and commissions independent studies. These suggest that its failure to predict the 2008 financial crisis was partly due to group-think – IMF forecasts rarely depart far from the consensus. There is an unwillingness to stand out from the crowd and be subject to scrutiny. Being alone and wrong is a sin. Being together and wrong is excused. This is aggravated by the inputs of state department country desks, and governments chastise forecasters for adverse predictions. Groupism is on display especially in consensus forecasts from Emerging Markets Economic Data – among nearly 40 forecasters, standard deviation for 2017 growth and inflation are 0.3%. Chinese inflation is the exception, with a 0.5% deviation.
Forecasters rely on the same data and short-term models. Largely based upon Keynesian multipliers and investment accelerators, these models suppose the data will continue to follow the same path as before. Policy apart, what went up goes up, what went down goes down, and turning points don’t exist. Financial shocks are known unknowns.
But forecasting errors are most pronounced when economies turn to a greater degree than the consensus standard deviation. EMED collects 39 forecasts for the US, of which 33 are from banks and financial institutions and only six from independent forecasters. Few have large teams using bespoke models. Many are small economist teams expressing no more than marginal disagreements with the Organisation for Economic Co-operation and Development, IMF, European Union or official forecasts. Financial institutions are of course reluctant to predict crises and spread alarm.
Looking to 2017, last month’s EMED average forecasts for nine major economies predicted slower growth for all except the US and Canada and faster inflation across the board. High frequency consensus forecasts are based on lower frequency forecast revisions, and few forecasters fine tune their predictions every month. Consensus forecasts are lagging and some are already stale.
The consensus is now busy writing-up 2017 growth and inflation. Few may predict market turmoil. The danger of a more abrupt rise in bond rates is clear in the light of faster growth and inflation. Euphoria around Donald Trump’s proposed and supposedly business-friendly plans may subside if protectionist policies and retaliation take the centre stage. Forecasts which go up go up faster, and then come down faster.
This may prove unlikely in 2017, and short-term optimism need not be misplaced. Thereafter, however, a turning point is predictable, including possibly another financial crisis. We do not know what shock, among numerous potential threats, might trigger it. To quote the English poet Robert Herrick: ‘Gather ye rosebuds while ye may, Old Time is still a-flying; And this same flower that smiles today, tomorrow will be dying.’
Brian Reading was an Economic Adviser to Prime Minister Edward Heath and is a Member of the OMFIF Advisory Board.
Tell a friend