For the development or continuity of a business lots of investment decisions are required. Making the right choices is key for the future success of the organization. Good templates for decision making processes exist that will allow arriving at such a choice in an orderly and structured manner. Although decisions can even be challenging in the context of limited uncertainty, for example because of delicate trade-offs between attributes of alternatives, the dimension that I would like to focus on in this article is uncertainty.
Many significant investment decisions are usually complicated by the fact that various parameters are (still) uncertain at the time the decision needs to be made. How sure are we about the cost and timing assumptions? Will the technology work? What is the outlook for the sales volumes? Is there a risk that some unforeseen development sticks a spanner in the wheel? Will the wider context for the investment (regulations, market issues, political stability, etc) continue to be favorable? These are the types of issues that require addressing in some way before the decision can sensibly be made.
To compare costs and benefits, typically a cash flow model will be constructed. This will yield some key financial metrics such as the net present value (NPV). Many assumptions need to be made for the inputs. Hence the number that comes out may look solid, in reality it will not be. A sensible decision maker will want to understand the uncertainty range around the numbers on the basis of which the decision will be made. This calls for probabilistic analysis where the inputs are not regarded as fixed, single numbers, but as having a range of uncertainty. Then, also ranges can be presented for the key decision metrics, providing insight into the credibility of the investment proposition.
This may seem like a no-brainer, in many companies this concept is absent or poorly implemented. For a start, amongst decision makers there are still quite a few ?just give me the number? characters around. If it is valid to present a value proposition by a single number, then the decision making process could be automated. In fact, uncertainty is the very reason why decision makers have a role. Interpreting the output of a probabilistic analysis does require a little basic understanding, but this should not be a major barrier. On the other hand, analysts in project teams preparing the proposition can get carried away with it and may be mechanically running one Monte Carlo simulation after the other. Such complicated simulations can be useful but are not always needed. There are other ways to structure, analyze and present the uncertainties, for example by using decision trees or analytic methods. As little time as possible should be spent on the mechanics, to have enough time to do what really matters: frame the problem, think about the uncertainties and risks, quantify where possible and gain rich insights through the process.
This does not mean that the quantified probabilistic analysis is the sole and sacred foundation of a good investment decision. It is just one element. You still need a deterministic analysis of one or more possible outcomes of the investment project. These cases need to be tangible and imaginable. But decision makers should have a sense where they approximately sit on the probability curve.
Yet, this is not the end of the story. There can be risks and uncertainties that cannot be credibly quantified. Quantitative (probabilistic or deterministic) analysis is not a goal in itself. It is just a means to make some sense of the full array of options, data and uncertainties associated with a possible investment project. Sometimes we cannot put a number on a particular uncertainty. Think about political developments, regulatory matters, market trends, technology. Such an issue needs to be captured by a narrative, perhaps in the form of a what-if discussion. If there are several issues like that, in particular when these are linked, then a scenario approach can be useful. Scenario thinking can even be applied in a limited local context, around a few linked trends that might affect the investment project in the future. Or it can be required to put the investment project in the context of extensive global scenarios with many different moving parts. Whatever makes sense, it would be very useful to link the qualitative scenario concepts to the quantifications, even if just indicatively. Otherwise the scenario concepts, aiming to better understand the key contextual uncertainties, remain too abstract and remote for the investment decision at hand.
Decision makers should love ambiguity. It makes their role challenging and interesting. Probabilistic analysis, quantification, scenario analysis will all contribute and shed light on a complicated uncertainty-ridden investment decision. Decision makers should skillfully use the analyses presented to them, combining that with their experience and intuition, making the final call with confidence.
In a next article I will discuss ways in which scenario thinking can be incorporated in the decision making processes of an organization.
