Economic forecasts are rarely correct, whether it’s the consensus forecast or forecasts by individuals. Moreover, they’re often the product of simple arithmetic calculations.
Take the non-farm payrolls (NFP) report, for example.
According to the Credit Agricole chart below, it’s remarkably similar to the average figure for the prior three months. So if the figure arrived by tallying up the forecasts of well-paid economist forecasters can be closely replicated by simple arithmetic, aren’t all these forecasters then worthless?
The answer is ‘no,’ as they’re not entirely useless.
According to calculations by IBTimes, the correlation between actual NFP figures and forecasts is consistently over 10 percentage points higher than the correlation between actual NFP figures and the average for the prior three months.
Economic forecasters really shine over figures arrived using simple arithmetic in periods of massive distortion, such as the 2010 census hirings. For example, in the month of May, NFP rose by 431,000 on 411,000 temporary workers hired by the census.
Forecasters predicted 521,000 while the average of the three prior months was only 139,000.
However, aside from the big distortions, economic forecasting is simply the extrapolations of historic data with a few minor adjustments.
“It might be worth questioning how much value we should put on the ‘consensus’,” said the Credit Agricole report.
Whenever an important economic data is released, “the market will spend most of today captivated by whether this consensus is right or wrong, whereas in reality all we are wondering is whether the number will be better or worse than we’ve seen recently,” stated the report.