The complexity of economies is comparable to the complexity of weather systems, but meteorologists consider themselves brave if they forecast more than a few days ahead.
New Year; new forecasts
In December each year, most economists revise their forecasts for the following year. Investors often pay more attention to the consensus of economic forecasts in January as they consider their portfolio allocations for the coming year.
Just like the forecasts a year ago, everyone seems to have a similar view as all the forecasts are more optimistic, although there appears to be a bit more uncertainty in the UK about the strength of the recovery.
The past errors in forecasting are also evident. In the US 2011 and 2013 (so far) turned out stronger than predicted, while 2012 was worse. In the UK, forecasters were too optimistic about 2011 and repeated the mistake for 2012 before being surprised again by 2013.
Despite the crisis, forecasters did a better job on the eurozone in recent years. Perhaps practitioners of the ‘dismal science’ are just better at forecasting bad news than good news. Or perhaps the uncertain effects of deleveraging and unorthodox monetary policy in the US and UK were harder to forecast.
Spectacular misses
So perhaps economists were not that bad at forecasting growth after all but, for example, how well do economists (and the financial analysts they work with) forecast government bond yields?
Forecasts for bond yields in 2011 and 2012 were spectacularly wrong across all three economies, and 2013 was generally off target (other than for German bunds). The errors in forecasting growth were not big enough to explain these misses. A more important explanation is that economists underestimated the actions of the central banks who embarked on aggressive unconventional measures such as QE in order to push bond yields down.
Once economists started to understand how this worked, forecasts improved in 2013 (but tapering in the US still provided an upside surprise).
Economists are now so certain that they understand how central bank policy works that their forecasts for 10-year yields at the end of 2014 are very tightly packed together for the US and UK. This level of agreement on the path of rates is part of what central banks have been trying to achieve through forward guidance.
There is a somewhat interesting pattern which holds true in most cases: when the distribution of forecasts for either growth or inflation has been quite wide, economists have tended to get it wrong. But in 2013 when the distribution was quite narrow, the forecasts for both growth and yields look to be more accurate.
Follow the herd
The wider distributions in 2011 and 2012 may reflect the realisation that there were many tail risks that could affect the economy, and the consensus is constructed from economists’ central estimates rather than their risk scenarios.
In 2013, economists may have seen the upside and downside risks as more balanced. So should we read anything into the narrow range of forecasts for 2014?
While a return of confidence in the recovery is certainly good news, so much agreement amongst analysts should make investors a bit wary. Many of the political tail risks have indeed been reduced and the pick-up in data is certainly consistent with a recovery that should lead to a gradual rate normalisation.
However, with consensus all positioned the same way, even a small negative surprise could cause sharp and potentially painful market movements. Just think of how markets reacted when the Fed surprised them by delaying tapering in September.
Given the challenge of forecasting the consensus has not performed too badly, but following it blindly would be senseless.
