Modelling Information Flow along the Yield Curve
In this paper we model the US yield curve using high frequency data drawn from the GovPX database. In particular we analyze the interaction between trading at different maturities through univariate and multivari- ate Hawkes processes to model trade and the price intensities. These price intensities are then used to estimate the instantaneous volatilities which in turn can be used in the HJM framework to directly model the entire yield curve and price related derivatives- caps and caplets. We have explic- ity exploited a time deformation framework in which the subordinating process is determined by the stochastic trade intensity process. We have found from the multivariate Hawkes process that there is no simple single market time that drives the yield curve but that different maturities ap- pear to be driven by different time scales. Activity at the short end of the yield curve drives the medium and long end of the yield curve but there is a clear structure to the interaction between trading at the different ma- turities both in terms of impact effect of news at one maturity and the shape of the decay function of that news.
Mark Salmon Wing Wah Tham
Financial Econometrics Research Centre Warwick Business School
国际会议
成都
英文
1-37
2007-07-09（万方平台首次上网日期，不代表论文的发表时间）