Alpha-beta trend following system
Version II, tuning for performance
by Anthony W. Warren Ph.D.
Several years ago, I published the alpha-beta trend following method in Technical Analysis of Stocks &
Commodities. Although similar in concept to the classic moving average crossing systems, this method
utilized an uncertainty band or trend channel to eliminate most of the false signals associated with
moving average crossings. In addition, it graphically represented the preferred times to go long, go short
and to stay out of a particular stock, bond or commodity market.
One of the problems, however, with the original alpha-beta code published in Stocks & Commodities in
1985 was that it did not easily lend itself to profitability analysis or tuning the input parameters. The
original code chose the trend channel's width based on all the data in the file, rather than just the data
prior to each trading date.
The revised computer code listed at the end of this article is more responsive to the current market
environment and is more easily tuned to eliminate spurious trading signals and to exit with maximum
profit! Let moving averages rest in peace. Try alpha-beta!
The alpha-beta trend-following method plots four time series:
the current price time history which, typically, is the daily or weekly closing price,
the upper and lower trend channels which define the uncertainty band for trade decisions, and