Use of orthogonal arrays and design of experiment via Taguchi L9 method in probability of default

Abstract

The Taguchi’s orthogonal array is based on a mathematical model of factorial designs. This paper investigates the effects of four parameters in Probability of Default (PD) using Black-Scholes model (BSM) for call option at one period by considering asset value V, firm’s debt X, expected growth r and the volatility σ. The main aim is to determine which parameters affect mostly on PD of a firm. The experiment is based on the orthogonal array L9 in which the four parameters are varied at three levels. Finally, the ANOM is used to describe the best combination and ANOVA is implemented to measure the contribution of the given independent variables.

Keywords

• PD
• BSM
• Taguchi method
• Regression line
• ANOM
• ANOVA