Modelling Portfolio Risks with Time-Dependent Default Rates in Venture Capital

Abstract

Previous risk management in the venture capital industry has focused mainly on qualitative risk management, such as team selection and due diligence. As investment volume has increased during the past decade, and as venture capital becomes more important as an asset class for institutional investors, rules of thumb do not apply any more. Furthermore, high-risk management standards, which are common and established for other assets, are demanded for this asset class. In his study, Kemmerer (2005) introduced a risk model for venture capital portfolios by adjusting the CreditRisk+ model to fit the characteristics of venture capital. The input parameter, rate, is entered as the calculated long-term average of the companies' sector. Based on the initial idea of this approach, the current study's aim is to develop a risk model which considers time-dependent default rates as input parameters which are adjusted yearly. By using time-dependent default rates, instead of long-term average default rates, it is expected that the predictability of the model will increase. This assumption is plausible, because historical regression results, with the default rate as the dependent variable, are highly significant, and demonstrate an outstanding explanation of the coefficient of determination. The empirical results strongly support the assumption that the introduced model measures risks more accurately than the original model. By using time-dependent default rates, instead of long-term average default rates, the predictability of losses increases significantly.