Page 24 - PowerPoint Presentation
P. 24

LOS 34.k: Describe modern term structure                                                  READING 34: THE TERM STRUCTURE AND
    models and how they are used.                                                                              INTEREST RATE DYNAMICS
                                                               dr = a(b − r)dt + σdz             MODULE 34.6: INTEREST RATE MODELS


     Assume that the current short-term rate is r = 3% and the long-run value for the short-term rate is b = 8%. Also assume that the
     speed of the adjustment factor is a = 0.40 and the annual volatility is σ = 2%. The Vasicek model provides the following formula
     for the change in short-term interest rates, dr:


     dr = 0.40(8% – r)dt + (2%)dz

     The stochastic term, dz, is typically drawn from a standard normal distribution with a mean of zero and a standard
     deviation of 1. Assume that a random number generator produced standard normal random error terms of 0.45, 0.18,
     −0.30, and 0.25. The Vasicek model would produce the evolution of interest rates shown below:

     Note that the simulation of interest rates                                                                        Note that because
     starts with an interest rate of 3%, which                                                                         both the Vasicek
     is well below the long-run value of 8%.                                                                           model and the CIR
                                                                                                                       model require the
     Interest rates generated by the model                                                                             short-term rate to
     move quickly toward this long-run                                                                                 follow a certain
     value despite declining in the third                                                                              process, the
     time period, which reflects the mean                                                                              estimated yield curve
     reversion built into the model via the                                                                            may not match the
     drift term a(b – r)dt.                                                                                            observed yield curve.
                                                                                                                       But if the parameters
     This example is stylized and intended                                                                             of the models are
     for illustrative purposes only. The                                                                               believed to be
     parameters used in practice typically                                                                             correct, then
     varysignificantly from those used here.                                                                           investors can use
                                                                                                                       these models to
                                                                                                                       determine
                                                                                                                       mispricings.
   19   20   21   22   23   24   25   26   27   28   29