CORMSIS Seminar - "Psychological Heuristics for Portfolio Decisions", by Ian Durbach (St Andrews, Cape Town) Event

Event details

In portfolio decisions the goal is to select a best subset of projects using limited resources. We develop and test a number of fast and frugal heuristics - psychologically plausible models which use few pieces of information and combine these in computationally simple ways - for portfolio decisions. The proposed "add-the-best'' family of heuristics constructs a portfolio by iteratively adding a project that is best in a greedy sense, with various definitions of what is meant by "best''. Simulation and analytical results show that portfolios selected by some heuristics can be close to optimal under a range of reasonable conditions, and offer very substantial reductions in information requirements. Results from a behavioral laboratory experiment show that choices are often consistent with the use of heuristics. A main result is that a heuristic that begins by identifying projects involved in positive synergies with other projects, and then selects from these with an extremely simple value-to-cost ratio, performs excellently on both the quality of selected portfolios and on the information savings, relative to an optimal solution.