Market Easier to Predict Over Long Time Horizons

By on May 29, 2010

So say statisticians according to Mark Hulbert.

Given how difficult it is to forecast the stock market’s short-term direction at any time, much less now, you might imagine that it is even more difficult to forecast the market’s direction over longer periods of time.

Ironically, however, that may not be the case. Statisticians tell us that it is less hard (note I didn’t say easy) to predict the market’s return over the next six to 12 months than it is to forecast its return over the next day. And it’s even less difficult still when the forecast horizon extends to longer periods of time.

Consider an econometric model maintained by Sam Eisenstadt, the former research chairman at Value Line, Inc. , the author of the famed Value Line stock-ranking system, and a rigorous statistical student of the stock market for over 60 years. He reports that his model sports an impressive track record back to 1952 in forecasting six-month returns. (Its r-squared, for the statisticians among you, is 0.3).

Or consider another econometric model with similar statistical success devised by Norman Fosback, editor of Fosback’s Fund Forecaster and formerly head of the Institute for Econometric Research. His primary trend model focuses on the market’s returns one to five years into the future, but makes no predictions about the market’s shorter-term movements.

According to Hulbert both models are now bullish; Eisenstadt’s model predicting a 20% return over the next 6 months while Fosback’s model is forecasting a 26% total return for the stock market over the next year and a 75% five-year return (equivalent to around 12% annualized).

Source: Marketwatch
***

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>