Money Supply and Stock Prices: A Probabilistic Approach

A relationship between money supply and stock prices is fairly recognized in the literature; Sprinkel and Palmer have tried to determine whether the money supply can predict stock prices.  If this could be done, an investor could allocate his capital to and from the market portfolio, or in and out of high and low beta stocks, in an attempt to time the market.

The authors use a probability function to predict when a turning point in the money supply yields a turning point in stock prices; and they boil this down to the combination (i.e., an "efficiency index") of a "reliability index" (i.e., what ratio of predicted turning points were true turning points) and a "opportunity loss index" (i.e., what ratio of true turning points were predicted by the system).

They then used M1 (currency held + demand deposits) and M2 (M1 + time deposits) components of the money supply and the Standard and Poor's 425 over the period January 1948 to December 1970 in their analysis.  Their results show that using a 3% filter for M2 seems to provide the best reliability index; wherein, the trough signals using M2 are 65% accurate and the peak signals using M2 are 59% accurate.  However, using the higher filter of 3% yielded a lower opportunity loss index, meaning there were several true turning points undetected by the system.

Overall, their efficiency index finds that the 2% M1/5% Stocks filter works best for peaks, and the 1% M2/5% Stocks filter works best at troughs; although the 1% M1/5% Stocks is close to the 1% M2/5% Stocks for troughs.  As such, using M1 with relatively sensitive filters seems to provide the best balance of reliability and opportunity loss for predicting turning points in stocks.

Gupta, M. C. (1974). Money Supply and Stock Prices: A Probabilistic Approach. Journal of Financial & Quantitative Analysis, 9(1), 57–68.