A Decision Analysis Approach to Cash Flow Management

We propose a method of applying the Hespos and Strassmann "stochastic decision tree" framework, originally intended for investment decisions, to cash flow management. Sequences of uncertain events, such as a strike, affecting forecast cash flows are represented by a probability tree. Forecasts of constituent cash flows such as sales and costs are represented by Beta distributions dependent on paths through the tree. Monte Carlo simulations sample these distributions, and equations provided in the model convert the sampled cash flows to cash balances in each period. Frequencies of cash balances weighted by probabilities along paths through the tree yield a combined relative frequency distribution of cash balances for each period. These and related results may be used by management to plan financing arrangements to meet cash requirements in the future.