The author investigates management’s ability to “learn” from its observations of the firm’s stock price and determines whether such learning occurs by studying management’s adjustments to the level of cash savings. Specifically, he hypothesizes that the larger the amount of private information, the greater the sensitivity of cash savings to the stock price.

First, he demonstrates that cash savings are sensitive to the amount of firm-specific information contained in prices, which is estimated as the stock’s return that is not explained by the market and the industry returns. Next, the author controls for private information that management may possess. When proxies for management’s level of private information are incorporated, the savings-to-price sensitivity based on the level of firm-specific information weakens but does not vanish. Furthermore, the suggestion that mispricing could be contaminating the result (i.e., firms issue stock when the stock price is too high and then accumulate cash) does not appear to be supported by the research.

In general, most researchers studying the information contained within stock prices tend to believe that managers try to control the dissemination of management’s private information to the public. The author reverses this idea by demonstrating that management may actually learn from information contained within stock prices themselves. Consequently, the information contained within stock prices actually flows to both investors and management.

Another interesting aspect of the author’s work is his use of the firm’s level of cash savings to demonstrate the information flow from stock prices to the management team. Thus, the level of cash savings serves as a proxy for the level of information that management receives from the market.

The author applies regression analysis to weekly data from 1970 to 2006 for all firms in Compustat except those firms classified as financial or as a utility, for a total of 11,937 firms. He regresses the firm’s stock returns against a market index and an industry index and uses the residual variance as a proxy for the firm-specific information. Next, he regresses cash savings (annual change in the holdings of cash and other liquid assets divided by lagged assets) against the normalized stock price (market capitalization of the firm divided by the book value of the assets) from the previous period, the measure of firm-specific information (residual variance from the previous regression), the cross-product (interaction term) of these two independent variables, and a set of control variables.

The regression results demonstrate that the change in cash savings is sensitive to both the normalized stock price and the cross-product term, which implies that sensitivity to stock price is greater when a firm displays a higher level of firm-specific information. Both relationships are positive and statistically significant, which is the main finding of the research and which indicates that managers tend to save more when they experience positive shocks to their stock price and when the signal from the market contains more private information.

To determine whether management is “learning” from changes in the stock price, the author uses proxies to measure the level of management’s information (analyst coverage, insider-trading activity, and earnings surprises). Although his main finding weakens as management becomes more informed, it does not vanish. The author obtains similar results when he reestimates the regression by including changes in net working capital and short-term debt, which are viewed as substitutes for cash. Results remain essentially the same when the author reestimates the regression equation with two alternative measures of price informativeness (illiquidity and gamma).

A further robustness check determines whether the main finding is sensitive to potential mispricing. If prices are above fundamental value, a firm may choose to issue stock that will increase the savings level. When the author controls for equity issuance, his main finding does not change.

I cannot dispute the empirical findings of the paper, but I am hesitant to fully accept them. Cash is an expensive form of capital that certainly provides a hedge for an uncertain future. But I wonder whether management is setting the cash savings level based on information implied from stock prices, information from the broader economy, or future credit availability. The use of more control variables within this investigation would lead to a more refined result.