衍生工具和风险管理-im05.docVIP

CHAPTER 5: OPTION PRICING MODELS: THE BLACK-SCHOLES MODEL END-OF-CHAPTER QUESTIONS AND PROBLEMS 1. In a discrete time model, the stock price can make a jump to only one of two possible values. The length of time over which the move can be made is finite. In a continuous time model, the stock price can jump to an infinite number of possibilities. The length of time over which the move can be made is infinitesimal (very, very small). The difference between the two models is perhaps best described as in the text as the difference between still photos and a movie. The familiar normal or bell-shaped distribution is a symmetric probability distribution that depends only on the mean and variance. A lognormal distribution is skewed, having more extreme right values. A lognormal distribution, however, is normal in the logarithm. Thus, if x is lognormally distributed, its logarithm, lnx, is normally distributed. With respect to stock prices, the logarithm is of the rate of return. That is, let (S1 – S0)/S0 be defined as the return over period 0 to period 1. Then if it is lognormally distributed, ln(S1/S0) is normally distributed. Note what appears to be slight inconsistency: (S1 – S0)/S0 is a percentage return whereas S1/S0 is 1.0 plus the percentage return. That does not matter as we can always shift a normal distribution by adding a constant such as 1.0 and it does not affect the fact that it is normally distributed. It just slides it over. 3. The Black-Scholes model assumes no dividends, but actually this was just for convenience to allow us to start at the simplest level. If the dividends are appropriately modeled, the Black-Scholes model handles stocks with dividends with only the minor adjustment that we must remove the present value of the dividends from the stock price before using it in the model. We can do this by subtracting the present value of the stream of discrete dividends over the life of the model or by discounting the stock price by the div