Page 34 - PowerPoint Presentation
P. 34

LOS 35.c: Describe a binomial interest rate                        READING 35: THE ARBITRAGE-FREE VALUATION FRAMEWORK
   tree framework (BIRT).
   Assumes rates have an equal probability of taking one of two value the               MODULE 35.1: BINOMIAL TREES, PART 1
   next period. Over multiple periods, the set of possible interest rate paths
   used to value bonds is called a binomial interest rate tree.


                        Adjacent is e 2σ  Away:  i 1,U  = i e 2σ or  i 1,L  = i 1,U e- 2σ
                                                    1,L
                                                                                                   = i 2,LL e 4σ           = i 3,LLL e 8σ
                                              where:
                                              e ≈ 2.7183 (base of natural log)
                                              σ = standard deviation (volatility)
                                                of interest rates
                                                                                                                       The relationship
                                                                                                                       among the set of
                                                                                                                       rates (branches)
                                                                                                                       associated with
                                                                                                                       each individual
                                                                                   i 2,UL  =                = i 2, LL e 2σ  nodal period
                                                                                                                       (stem) is a
                                                                                                                       function of the
                                                                                                                       interest rate
                                                                                                                       (roots) volatility
                                                                                                                       assumed to
                                                                                                                       generate the tree.
        e = BIRT Framework is a lognormal random walk
        model with 2 desirable properties:
        (1) higher volatility at higher rates; and
        (2) non-negative interest rates.


        σ = estimated from historical data or implied from            Implications? For exams, you might only need 1 to solve the rest
            interest rate derivatives.
   29   30   31   32   33   34   35   36   37   38   39