An intelligent decision support system for private equity investment

In this paper we investigate whether statistical learning algorithms can be effectively deployed in an Intelligent Decision Support System (IDSS) to reduce idiosyncratic risk in private equity investment decisions. This is particularly desirable in light of the new Equity Crowd Funding (ECF) private equity investment class created by the JOBS Act, since the Act opens up private equity investment to non-accredited investors who are not necessarily experts in the private equity due diligence process. By comparing confusion matrices of 6 statistical learning classifiers, we show that logistic regression can reasonably discriminate between investments that yield rates of return that are attractive, intermediate, or poor. These results also demonstrate that an Intelligent Decision Support System based on statistical learning can effectively reduce the idiosyncratic risk inherent in Equity Crowd Funding and other private equity investments.