Why is it not suitable to use Geometric Brownian Motion to model bond price?

Consider a bond P(t, T) observed at time t that matures at time T. Why is it not suitable to use geometric Brownian motion to model the bond price P(t, T), where:

dP(t, T) = mu*P(t, T)dt + sigma*P(t, T)dz(t)

when mu and sigma are constant.Why is it not suitable to use Geometric Brownian Motion to model bond price?This was the sort of mumbo jumbo used to bamboozle regulators,impress the credulous, and part the naive from their money while at the same time and raise grasping avarice to the operation of spurious 'physical laws'. And what a fine mess this modelling/guessing got us all into.Why is it not suitable to use Geometric Brownian Motion to model bond price?I ought to know this. I'll look it up and get back. However I suspect it is because changes in the underlying interset rate are the main cause of variability in bond prices. In this model, the interest rate is reflected in mu, which is constant.

Report Abuse