Macro-based Adjustment Factors for Valuations in Venture Capital

Abstract: This paper focuses on creating macro-based adjustment factors which can be applied to the valuation of start-up companies. Since start-up companies are private, market valuations are limited to events such as funding rounds or M&A transactions, which happen only at discrete points in time. This is a drawback when compared to the continuous valuations produced in public markets, with current literature focusing on accounting for the timing differences in the valuation of private firms through the application of market indexes. Alternatively, research into the interactions between performance in the venture capital sector and economic conditions suggests that macroeconomic indicators might also be useful in estimating timing adjustments for start-up valuations. The two proposed approaches are compared through a pooled ordinary least squares (OLS) estimation of a panel data sample of relative change in valuations in the United States' venture capital sector between 2007 and 2021 using both models based on market indexes and macroeconomic indicators. The adequacy of the adjustment factors is evaluated based on the comparison of prediction errors in an out of sample forecast between the market-based and macro-based models. Overall, a parsimonious macro-based model using per capita GDP as a sole explanatory variable is found to outperform both the complete macro-based model and the remaining market-based models. However, improvement relative to the market approach defined in Damodaran (2009) is slight. Between the market-based approaches studied, a small cap market index is found to be, in most scenarios, a more accurate predictor of timing adjustments than a synthetic index based on sectorial market performance. Furthermore, the introduction of time fixed effects in the regression generally worsens results, only presenting general improvement relative to the naive prediction when using either the complete macro-based model and a small cap market index as explanatory variables.