Investment decisions as choice under uncertainty: Micro, market, and macro analysis

This dissertation contains an analysis of investment decisions as choice under uncertainty. Investment is modeled as a two-stage decision process undertaken by managers with bounded rationality. First, the strategic decision is made under Knightian uncertainty in which agents have difficulty calculating probabilities across outcomes. This is modeled as a satisficing decision that incorporates reliability weights to represent the company's behavioral reliability in different activities. This strategic decision allocates the company's scarce decision-making capacities among different activities. After searching for and developing investment opportunities involving acceptable activities management will then select specific investment projects by maximizing expected value. The strategic decision therefore determines the set of activities from which projects can be selected. This strategic set will expand and contract with changing conditions, sometimes restricting the company's response to investment opportunities. Under simple, stable, and familiar conditions an immediate response to change will occur. Under complex, unstable, and unfamiliar conditions response to changing investment opportunities may occur only with a lag, or may never occur.;At the microeconomic level the basic behavioral prediction of this analysis is that it is hard to enter new activities and hard to leave familiar ones. The descriptions of innovative and entrepreneurial behavior are consistent with empirical evidence concerning those activities. At the market level it is shown that a competitive equilibrium can be ensured only under simple, stable, and familiar conditions. Under complex, unstable, and unfamiliar conditions the tendency toward competitive equilibrium may never be fulfilled. This is consistent with theoretical results that have shown that the presence of some non-maximizing behavior in a system will cause a general equilibrium allocation of resources that is significantly different from that achieved with all maximizing agents. And, finally, at the macroeconomic level this investment analysis could contribute to an explanation of business cycles in which the level of investment is sometimes a fast-adjusting, "rational" variable and sometimes a slow-adjusting, adaptive variable, depending on the role that the satisficing strategic rule plays in restricting or allowing response to change.