Forecasting future cash flows is a standard practice in corporate finance; it is widely used for valuing firms and investments, and frequently used by managers and analysts around the world for forecasting their firms’ future generation of cash. Future cash flows are usually referred to as expected cash flows, reflecting their stochastic nature, i.e. the fact that their occurrence is not granted; rather, they have some level of uncertainty. Notice, however, that the usual forecasted future cash flows are, in reality, a mix of expected and promised cash flows, and both have different levels of uncertainty and not reaching their expected values have different consequences for the firm. This post explains why this constitutes a problem and proposes a better estimation of future cash flows to help measuring risks.
An expected cash flow is, by definition, a cash flow occurring in the future with some sort of probability distribution associated to it. In other words, it is usually built from an expected level of sales and variables costs that are affected by volatility and whose occurrence might depend on several future outcomes. Some of the cash flows that we usually forecast, are really expected, meaning that they have an associated volatility that is related to normal business conditions, for example volatility of prices and sales volumes, inflation, interest rates, currencies, etc… Some other portions of the future cash flows, however, have been promised to some counterparty; debt payments to banks or bondholders and fixed costs such as salaries to employees are good examples of these promised cash flows. These payments have been promised and cannot be postponed or cancelled without a default or a profound renegotiation with the counterparty. I am not stating that this second group of cash flows is fixed and cannot be modified, but its modification has serious consequences on normal business conditions. My question here is how reasonable is to mix both estimations in the expected cash flows, and, more importantly, to use the same discount rate to discount them when estimating the present value of that cash flow… using the same level of expected return –a measure of risk- for both cash flows seems not reasonable.
In this article I do not want to stress the previous point; I am not discussing valuation and discount rates. In this article I am proposing an improved exposition methodology for cash flow forecast that will help us measuring risk. What I propose is to separate expected cash flows from promised cash flows. For doing this we need to identify the expected cash flows affected by standard business volatility, i.e. the volatility of the variables that we have (or should have) in our risk map, and the promised cash flows, those cash flows we should meet in order to avoid some sort of default. Another way of seeing this is: firms generate expected cash flows and use them to meet the promised cash flows. In some sense, expected cash flows are some sort of expected input, and promised cash flows some sort of required output.
Expected cash flows, usually shown as a number, are, in reality, a distribution of probabilities, and, as such, can be presented as a confidence interval, say for example the 5%-95% confidence interval. This means that we have a 90% of probabilities of falling into that range of generation of operational cash flows. The level and range of the cash flow will depend mostly on the business model and its associated risks. This expected operational cash flow has to be distributed among the counterparties of our promised cash flows; among them we have, (i) debt repayments to debt holders (banks, investors, etc.), (ii) salary payments to employees, (iii) bill payments to any other “fixed costs providers”, and (iv) payments required by the investment process. This last point might be a bit puzzling and should justify some additional explanation. Most of the times, once an investment is started finishing it constitute a major challenge (called Completion Risk). This is usually considered in the case of fixed investments (plants, infrastructure, etc.) but we should also learn to consider it in operational investment as well. Firms entering a new market and not being able to complete its expansion or installation are also a case of completion risk.
After separately estimating both, expected and promised cash flows, we need to combine them in a graph as shown in Figure 1 below.
Figure 1. Expected Cash Flows, Promised Cash Flows and Potential Default Area
Expected and Promised Cash Flows
As we can see in the graph, the operational expected cash flow has an upper and lower bound confidence interval. In this case, the promised cash flow level is above the lower bound of the expected cash flow, generating an area of overlap; the potential default area. As we can see in the graph, if expected cash flows are below a certain level, the company will default on a portion of its promised cash flows, generating some problem: it might default on its debt, not pay salaries or other fixed costs or fail to meet investment cash requirements.
Using simple Montecarlo simulation software it is fairly straightforward to generate this graph. The expected values of the inputs in the business plan should be replaced by their probability distributions, and the expected operational cash flow should be set as the output of the simulation model. The next step is generating a graph showing the expected cash flow and the promised cash flows as shown in Figure 1 above, and generating the overlapping potential default area. This area is easily measured by the Montecarlo simulation software packages, resulting in a measure of the risk of the model; the one related to the risk of defaulting on the promised cash flows.
The methodology described in this post is not difficult to perform; most of the information needed is already in the firm, and what is left is just organizing it correctly in the financial planning models. The advantages of reorganizing cash flows in this manner are that managers can now measure the level of risk embedded in the firm’s business plan and risk map, and assess the reduction in risk of any hedging strategy that might be modeled in the business plan.
