Multi-factor models and the risk premiums: a simulation study

Abstract

The estimation of commodity spot price models often involves the estimation of risk premiums. We show in a simulation study that the market prices of risk cannot be accurately estimated using two popular estimation techniques; the Kalman filter and an iterative routine. Risk premium parameters may be dependent on the starting value for the iterative routine, and cannot be accurately estimated using the Kalman filter technique. We conclude with a short analysis of results from the spot price model literature by examining the implied volatility term structure from other published research papers.