Discussion of the 'Role of Volatility in Forecasting'

Minton, Schrand and Walther (2002) (MSW) investigate whether cash flow (earnings) volatility helps predict subsequent levels of cash flow (earnings). Price is the present value of expected future cash flows, so if cash flow volatility forecasts future cash flows (the numerator in the present value calculation), it should have valuation implications. A similar motivation applies to earnings, which may be viewed as a proxy for cash flow.

To my knowledge, MSW is the first study that directly examines the relation between cash flow volatility and subsequent cash flow levels. The authors hypothesize that cash flow volatility is negatively related to future cash flows because market imperfections induce a wedge between the costs of internal and external funds, and hence cash flow volatility increases the likelihood of underinvestment (relative to the cost of internal finds). Indeed, it appears that firms? investment decisions are sensitive to the source of funding.

Therefore, I believe that the research question is interesting and relevant. In fact, as MSW point out (in their Section 2.1), there are additional explanations for a relation between cash flow volatility and future cash flows, which make the research question even more appealing. However, by focusing on the underinvestment effect, MSW provide only limited evidence on the research question. In particular, while their results suggest that cash flow volatility helps predict future cash flows, the evidence that the relation is due underinvestment is weak.

My discussion is organized as follows: Section 1 reviews alternative explanations for a relation between cash flow volatility and future cash flows, and how they may affect the interpretation of MSW?s results (in particular, the conclusion that the predictive ability of cash flow volatility is due to underinvestment). Section 2 discusses additional concerns regarding the analysis in MSW, and Section 3 presents an empirical analysis that demonstrates some of the arguments in the prior sections. Section 4 concludes.