CASH FLOWS AND RISK ANALYSIS

Much has been written about probabilistic methods of risk analysis and about the probability distributions required for its use. This paper demonstrates a method known as the probability of acceptance error approach using simple scenarios in which the only variable tested by a probability distribution is rental growth but in which other variables such as discount rate, capitalisation rate and holding period are also tested. In particular, the sensitivity of the probability distribution chosen for the rental growth values is discussed where both the rental growth values themselves and the probability distributions are normally distributed and also where they have a skewed distribution. It is shown that for current market projections for rental growth great accuracy in the selection of a probability distribution is not required. It is also shown that assumptions about independence or serial correlation of cash flows may be similarly treated and that the probability of acceptance error may be described as a range having independence and serial correlation as the two extremes. The range usually turns out to be fairly narrow. The most sensitive item in the calculations is, as expected, the discount rate. The above findings are demonstrated in a series of appendices.