摘要

The Moore-Penrose inverse is an important tool in algebra. This paper shows that the Moore-Penrose inverse is also an efficient technique in determining the minimal martingale measure if a security price follows a semi-martingale which satisfies some structure condition. We extend a result of Dzhaparidze and Spreij concerning the Moore-Penrose inverse to the case that the Moore-Penrose inverse of any matrix-valued predictable process is still predictable. Furthermore, we obtain an explicit formula of the minimal martingale measure by employing the Moore-Penrose inverse. Specifically, the minimal martingale measure in a generalized Black-Scholes model is found.

全文