Collective Learning about Systematic Risk

This study proposes a new perspective on the systematic risk -- firms are uncertain about their risk exposure and instead learn about the parameter from relevant observations of industry peers as well as their own. Our model predicts that the learning causes unique time-series patterns in firms' capital investment and valuation. Both the investment-capital ratio and the market-to-book ratio should decrease with the posterior estimate of the risk exposure and increase with a cumulative number of firm-year observations. The model also predicts that the investment should respond negatively to a part of cash-flow growth that the systematic component accounts for. We find that the empirical data exhibits these learning-related regularities. Furthermore, this learning appears to be collective rather than individual process in corporate practice.