In a volatile economic environment, creating a business case for a new project or product which contains financial projections is an extremely difficult task. Looking at historical data and extrapolating the numbers to create future projections simply doesn’t cut it anymore. Creating a business case is still imperative for companies to plan for the future, attract investor funding and gain approval for projects, although most of the time, financial projections are way off the mark. How do we create a financial model for a business case though that is not complete guesswork?
Scenario and Sensitivity Analysis
Thorough stress-testing, along with scenario and sensitivity analysis will provide your business case the rigour and robustness to cope with various fluctuations in economic inputs. At absolute minimum, any business case financial model should have a best, base and worst case scenario. Your scenario analysis may include answers to the following questions, as well as others specific your industry. Obviously, particular attention should be paid to the downside!
- What is our break-even number of customers?
- What happens if we lose our biggest expected customer? What if we gain another customer?
- What if we lose a supplier? Will this impact costs?
- What if interest rates increase or decrease?
- What if we lost a key staff member?
- It’s unlikely, but what if all of the above negative outcomes eventuate?
Of course, if you count on the worst case scenario, you’ll probably end up doing nothing, and you definitely won’t get any funding or approval! However, by creating a well-built, robust financial model, we can know exactly what the possible outcomes are so that we can show that we are prepared for the best or worst eventualities.
Stress-testing a business vs. stress-testing a model
There has been a lot written recently about the importance of stress-testing businesses, and I need to point out that this is a little different to stress-testing a financial model. Stress-testing a business involves putting the business’ financial forecasts through scenario modelling as described above. Stress-testing a financial model is more about testing the technical workings of the model; i.e. varying the inputs to see how much the outputs change. Strategies to stress-test a model include:
- Set inputs to zero and check that the outputs respond as you would expect. For example, by setting price to zero, you would expect revenue to also be zero.
- Double your units sold. Does your revenue double?
- If you are indexing costs, try setting the indexation percentage to zero and see if the costs remain flat.
In conclusion, stress-testing your business case by way of scenario and sensitivity analysis will help you to be prepared for varying outcomes as a result of fluctuating external factors. Once your have completed your financial model you should stress-test the workings of model to ensure that it is robust and accurate.
The point about distinguishing between stress-testing a business vs. stress-testing a financial model is a good one and often overlooked.
People with a commercial hat on will be more interested in stress-testing the business represented by a financial model, and I think it is fair to say that they will assume that the underlying logic of a financial model is all correct. However, from a model developer point of view (which could often be the same person changing hats), it is important to test the underlying logic of a model during the development process as described in the article.
A few more similar tests that could be performed include:
- An NPV calculated with a discount rate of zero must equal the sum of the cash flows.
- An NPV calculated with a discount rate equal to the IRR must equal zero (it is often forgotten that this is the definition of IRR in the first place).
- A weighted average must always be between the minimum and maximum values.
- What happens if interest rates are set to zero?
- What happens if exchange rates are set to zero?
- What happens if depreciation periods are set to one?
- What happens if the tax rate is set to zero or 100%?
From a commercial point of view many of these tests may seem non-sensical, but the intention of these tests is to check the model logic only, not commercial robustness.