Application of Moore-Penrose inverse in deciding the minimal martingale measure (Q601957)

The authors prove that the Moore-Penrose inverse of any \(\mathbb R^{d\times n}\)-valued, predictable process is still predictable and then obtain an solution of auxiliary unsolved stochastic equation that appears when we one intends to derive the minimal martingale measure. The solution is obtained via the Moore-Penrose inverse method. Moreover, an accurate expression of the minimal martingale measure is obtained in a generalized Black-Scholes model.