PD Modeling of Middle Market Firms with Macroeconomic and Industry Covariates
The credit risk behavior of large public firms is well understood. In recessions, risk neutral and physical default probabilities as well as loss given default rates increase. Intuitively, this effect is driven by a worsening of business conditions and a rise in market risk aversion. Estimating both through-the-cycle (TTC) and point-in-time (PIT) PDs for public firms is relatively easy because of the wealth of financial and accounting data.
The literature on credit risk modeling for private firms is sparse. The main objective of this project is to gain a better understanding of the determinants and dynamics of middle market defaults. The key idea is to enhance the PD model with industry, macroeconomic, and financial covariates to obtain PIT PD estimates. Macroeconomic variables include industrial production, unemployment rate, monetary policy variables, etc. Industry variable include capacity utilization, book-to-market ratio, leverage, cash flow volatility, etc. The challenge is to correlate the dynamics of public firms, which are observable, with the one of private firms. This set can be achieved with industry specific Z-scores.
