Research Publication

We develop and apply bootstrap methods for diffusion models whenfitted to the long run as characterized by the stationarydistribution of the data. To obtain bootstrap refinements tostatistical inference, we simulate candidate diffusion processes. Weuse these bootstrap methods to assess measurements of local meanreversion or pull to the center of the distribution for short-terminterest rates. We also use them to evaluate the fit of the model to the empirical density.

@article{conley1997bootstrapping,
  title={Bootstrapping the Long Run},
  author={Conley, Timothy G and Hansen, Lars Peter and Liu, Wen-Fang},
  journal={Macroeconomic Dynamics},
  volume={1},
  number={02},
  pages={279--311},
  year={1997},
  publisher={Cambridge University Press}
}