Tuesday, March 15, 2011

Backward-Looking Forecasts

We spend a lot of time reading forecasts. The financial news media is rife with new articles every day that take a position on the near-term future of inflation, interest rates, and stock prices. But are these forecasts of any use? James Montier, an author on the topic of behavioural finance, says no.

He has compiled and aggregated past forecasts for a number of popular financial metrics. The following chart illustrates how well forecasts of inflation have approximated actual inflation over the last several decades:

Note that the forecasts of the deflator actually lag the actual deflator! This forecast is telling you what happened, rather than what is about to happen!

Next, consider forecasts of the US government 10-year bond yield:

Once again, the same thing seems to be occurring, where the forecast is simply telling us what has already happened! Finally, a similar phenomenon occurs when forecasts of the S&P 500 are considered:

Of course, it should be noted that these forecasts are considered in the aggregate, whereas some forecasters may be able to consistently outperform others. However, there does appear to be a lack of supporting evidence that some forecasters are clearly superior.

Sixth-century BC poet Lao Tzu remarked that "those who have knowledge don't predict" while "those who predict don't have knowledge". So if forecasts are so useless, why do we spend so much time making them and reading about them?

1 comment:

Эдуард Лавров said...

And we have everything in Russia falls.