Managing treasury targets with Cashflow-at-Risk
by Kylene Casanova
There are two well-known ‘at-risk’ methodologies. Value-at-Risk (VaR) which quantifies risk with a present value approach and Cashflow-at-Risk (CFaR) which focuses on projected future cash flows. Both try to statistically answer the same question: How bad can things get? VaR is probably better known since it is widely used by financial institutions but does not extend easily to corporate risk management where cash flow-based approaches like CFaR are more appropriate.
In this article we explain the concept of CFaR and how it can be used to manage treasury risk and make effective hedging decisions. We then introduce an innovative new concept called Target@Risk®. This concept enhances the utility of CFaR by providing treasurers with a practical guide to assist their hedging strategy so they can work towards clear targets and successfully deliver on prescribed financial objectives.
Cashflow-at-Risk
CFaR attempts to model the behaviour of underlying risk factors and their effect on the cash flows of a portfolio. It is a statistical measure which attempts to quantify how a company’s cash flows may vary over a period of time.
We will illustrate CFaR with an example of an Australian exporter who is expected to earn GBP10M per month for the next 12 months. The current AUD/GBP spot rate is 0.5300 and treasury fears this may rise and result in a significant reduction in AUD revenue. A CFaR based approach enables them to understand the range of potential outcomes and quantify the effect of any hedging strategy.
Figure 1: AUD/GBP Spot rate for last 10 years
To provide perspective, the above chart shows the path of AUD/GBP spot for the last 10yrs. It has moved substantially in that time from 0.3800 to 0.7000. So based on key statistics determined from this known historical behaviour (such as volatility and long term means), CFaR seeks to simulate possible but credible future paths of AUD/GBP spot over the next twelve months. This is normally done with Monte Carlo simulations which introduce a ‘controlled randomness’ to the projected outcomes. Often, the simulation range will capture extreme events which might otherwise be overlooked.
Figure 2 shows 1000 simulations of the AUD/GBP exchange rate over twelve months. They commence from a spot of 0.5300 and generate forecast outcomes ranging from a worst case 0.6900 to a best case 0.4500.
Figure 2: Monte Carlo simulations showing a set of possible outcomes of the AUD/USD exchange rate over 12 months
Cash flows are then generated from each simulated path and the resulting 1000 sets of cash flows are collated as a distribution, shown in Figure 3. The distribution aids the following steps in the calculation:
- The mean of the distribution is used to represent the expected outcome.
- The cash flow at a chosen confidence interval (95%) represents the worst case outcome.
CFaR is then calculated as the variance between the mean and worst case cash flow outcomes.
Figure 3: A distribution of Net AUD cashflows from a stream of GBP earnings
From here we can see that on an average (or mean) we would expect to earn approx. AUD 224M over the next 12 months i.e. the mean of the distribution. We can then turn our attention to the extremes on either side and in particular, the worst case outcome. In statistical models such as the one above, the worst case outcomes are generally denoted by a confidence interval. If we consider a 95% confidence interval then we can state that the worst-case AUD earnings should not be less than AUD 210M. This gives us a CFaR range of AUD 14M or 6.3% of the expected mean income.
This traditional form of CFaR as described above enables us to quantify the difference between a 95% worst case outcome and a mean expected outcome. However, one of the key problems for this type of traditional CFaR analysis is that both the baseline mean and 95% outlier outcomes can change from month to month depending on the movement in markets and the simulations run. As such, the approach cannot easily deliver management any clearly defined or specific outcomes that help them actually manage their risk on a day to day basis. It can also be difficult to explain the merit of predicting a 95% tail event when the market is volatile and that 95% level so changeable.
This is where the Target@Risk metric can step in to add value. Since treasuries generally manage risk from the perspective of a Target or a Budget, a more pertinent and pragmatic measure of risk would be the variance between the treasury Target and the worst case outcome, rather than the variance between variable mean and the 95% levels.
By Dr Alankar Karol, General Manager, Visual Risk , Europe
Like this item? Get our Weekly Update newsletter. Subscribe today