Picking profitable investments: The success of equal weighting in simulated venture capitalist decision making

Using computer simulation, we investigate the impact of different strategies on the financial performance of VCs. We compare simple heuristics such as equal weighting and fast and frugal trees with more complex machine learning and regression models and analyze the impact of three factors: VC learning, the statistical properties of the investment environment, and the amount of information available in a business plan. We demonstrate that the performance of decision strategies and the relative quality of decision outcomes change critically between environments in which different statistical relationships hold between information contained in business plans and the likelihood of financial success. The Equal Weighting strategy is competitive with more complex investment decision strategies and its performance is robust across environments. Learning only from those plans that the simulated VC invested in, drastically reduces the VC's potential to learn from experience. Lastly, the results confirm that decision strategies differ in respect to the impact of added information on the outcomes of decisions. Finally, we discuss real-world implications for the practice of VCs and research on VC decision making.