A Bayesian Approach to Real Options: The Case of Distinguishing Between Temporary and Permanent Shocks

This paper studies the optimal timing of investment in the presence of uncer-tainty about both future and past shocks. Speci\u85cally, in addition to the standard Brownian uncertainty driving traditional real options models, we also allow for Bayesian uncertainty over distinguishing between the temporary or permanent na-ture of past cash ow shocks. As a result, the evolving uncertainty is no longer constant, and is driven by Bayesian updating. We solve for the optimal investment rule and show that the implied investment behavior di¤ers signi\u85cantly from that predicted by standard real options models. For example, in contrast to the standard real options implications, \u85 rms hold both a traditional Brownian option to wait as well as a Bayesian option to learn, investment may occur at a time of stable or decreasing cash ows, investment may respond sluggishly to positive cash ow shocks, and investment timing will critically depend on the maturity structure of the project cash ows