Over the last decade, U.S. firms have used their large cash holdings to buyback shares and pay dividends, instead of investing in R&D activities that can spur future growth. One possible reason for this behavior is that R&D activities — yielding new products and services or cost reductions — are generally believed to be subject to high uncertainty. But what specifically are the uncertainties driven by R&D? We hardly know. In this study, we examine three potential drivers of R&D uncertainty: product uncertainty, process-cost uncertainty and technological disruption uncertainty. We measure product uncertainty using future sales volatility; process-cost uncertainty using future cost of goods sold volatility; and technological disruption uncertainty using future special items volatility. Regarding product and process-cost uncertainties, we find that the association between R&D expenditures and subsequent volatilities of sales and cost of sales, is lower than our benchmark, tangible capital expenditures, generally considered a relatively safe investment. However, the association between R&D and future special items volatility is significantly higher than that of capital expenditures. Collectively, the findings show that the major uncertainty created by R&D is from technological disruptions and not product failures. By implication, firms should be using their large cash holdings to invest in radical R&D, which will enable them to be the leader in disruptive technology
