Pricing Convertible Bonds with Monte Carlo Simulation

Abstract

This paper proposes a pricing model that values convertible bonds with Monte Carlo simulation. The optimal exercise boundaries for the embedded American-style conversion, call, and put options are inferred from the conditional expected value of continuation which is obtained by least-squares regressions in combination with a backward-induction procedure. The simulation-based pricing method is more flexible than traditional valuation approaches based on finite differences and binomial trees. It allows to better model the dynamics of the underlying state variables and to account for the specifications of the instrument, such as the path dependencies inherent in many callable convertible bonds. Credit risk is accounted for by directly modelling the possibility of default.