|  FEB 1995

A Mutual Fund Switch System by Nelson Freeburg

A Mutual Fund Switch System by Nelson Freeburg Testing indicators to confirm viability is always important, and here, two indicators that stand up to historical testing are presented and then used to develop a simple mutual fund switch system. Many successful stock market analysts and money managers rely on quantitative market models - systematic timing methods that incorporate a variety of technical, monetary, valuation and sentiment data. Analysts such as Norman Fosback and John Hussman employ multiple regression to blend their indicators while other analysts such as Gerald Appel, Ned Davis and Martin Zweig typically use less formal techniques to achieve splendid investment returns. However the logic works, all such models face one irreducible minimum requirement. Before the model is built, the analyst must determine which indicators work and which do not. Unfortunately, many popular investment methods turn out to have little actual forecasting value when tested over long periods of market history. Here are two indicators that do stand up to historical testing. After examining their derivation and past performance, we will use these indicators to develop a simple mutual fund switch system. In testing, this hypothetical timing model outpaced the Standard & Poor's 500 index by a wide margin since 1957 to the present while exhibiting far less statistical risk. Though this model is effective, our main focus is on the two component indicators themselves and not any particular application. Consider this study as one step in the long process of isolating which of the hundreds of available stock market indicators actually have forecasting significance, and of those that do, how the indicators are best employed.

by Nelson Freeburg