Likelihood-based specification analysis of continuous-time models of the short-term interest rate ☆

摘要: An extensive collection of continuous-time models the short-term interest rate is evaluated over data sets that have appeared previously in literature. The analysis, which uses simulated maximum likelihood procedure proposed by Durham and Gallant (2002), provides new insights regarding several unresolved questions. For single factor models, I find volatility, not drift, critical component model specification. Allowing for additional flexibility beyond a constant term drift negligible benefit. While would appear to imply short nonstationary, fact, stationarity volatility-induced. simple elasticity volatility fits weekly observations three-month Treasury bill remarkably well but easily rejected when compared with more flexible specifications daily data. methodology can also be used estimate stochastic models. adding latent large improvement physical process, it does little improve bond-pricing performance.