Magni, Carlo Alberto
University of Modena and Reggio Emilia

Source of information: http://mpra.ub.uni-muenchen.de/13539/1/MPRA_paper_13539.pdf

  The heuristics-and-biases program draws attention to logical fallacies of human reasoning which cloud human minds (Tversky & Kahneman, 1974; Kahneman, Slovic & Tversky, 1982; Kahneman & Tversky, 1996): choice behaviors are biased because they deviate from accepted standards of rationality. By contrast, advocates of bounded rationality are concerned with showing that heuristics may often prove useful in decision-making (Gigerenzer, Todd & the ABC Research Group, 1999). The latter aver that cognitive illusions disappear once heuristics are seen as adaptive tools rather than logical devices for solving decision-making processes (Gigerenzer, 1996; Gigerenzer, 2000). Gigerenzer and the ABC group endorse the idea that human beings’ rationality is ecological rather than logical, and their fast-and-frugalheuristics program aims at presenting a number of heuristics that are successfully applied in real-life in specific environments. This work deals with the NPV maximization model, which is a keystone in economics and finance. The main results may be summarized as follows:

1. real-life decision makers often use the NPV criterion, but discount cash flows with a hurdle rate which differs from the cost of capital normatively suggested (e.g. CAPM, arbitrage pricing, multifactor models)
2. the hurdle-rate rule may be interpreted as a boundedly rational approach to investment decisions: decision makers rest on aspiration levels subjectively determined
3. some empirical evidence shows that the bounded-rationality approach is ecologically rational, domainspecific, psychological plausible. The hurdle-rate rule leads to close-to-optimal solutions when confronted with the expanded NPV
4. several factors affect the hurdle rate, among which: decision flexibility, future opportunities, rationing of managerial skills, strategic considerations, agency costs, costs of external financing. Risk is another factor as well, but the notion of risk is not equivalent to the risk measures employed in analytical models
5. consistency with several theories and models is highlighted, among which real options approach, resource-based theory, Top Management Team literature, agency theory, strategic management literature. The hurdle rate is set at a base level but is not rigidly applied: possible fluctuations around the base level depend on various domain-specific and project-specific factors (such as the drivers just mentioned)
6. the NPV maximizing model is not univocal: it may take on several different dressings. The use a particular version depends on the circumstances: for example, the arbitrage-pricing NPV may not be used if the actual market is not complete. Some of the versions are often inapplicable and/or reveal inconsistencies with accepted standards of rationality (additivity, no-arbitrage, description invariance)
7. an interaction between the two systems of rationality may improve performance. Actually, the very distinction between the two systems of reasoning is rather artificial and only useful as a metaphor (see Gigerenzer & Regier, 1996): as attested by recent historical researches, the origins of the NPV just lie in the common efforts of various categories of people facing the need of both logical and eco-logical rationality.

This paper may suggest a direction for research and hopefully will act as a stimulus for further inquiries. The scientific niche that might be disclosed could offer unexpected views about bounded and unbounded rationality and their interrelations. A possible payoff of such a view is that bounded rationality and unbounded rationality are not necessarily rivals. After all, unbounded rationality itself, as derived from logic and mathematics, may not abstain to consider itself as a derivation of ecological rationality: logic and mathematics should be intended as the most advanced step of human simulation, and simulation is a tool for anticipating; as such, it is indispensable for discovering and creating. The impressive degree we have achieved in such an ability may be seen as the result of an evolutionary process, in which its surviving value has been adaptively tested. Therefore logic and mathematics, as symbolic tools, assemble the experience of our ancestors (Monod, 1970, pp. 171-172). To put it in a nutshell, even logic is ecological. More “fluid theories” and more mixed strategies will perhaps help us understand decision makers and help decision makers cope with complex decision problems, with the pleasant by-product of conciliating the two rival parties.