Avoid a Binary View of Uncertainty
Balancing risk and reward is the essence of financial management. Although uncertainty doesn’t define risk, it’s one of its key components:
- Uncertainty – the odds of being rewarded for your investment of effort and money
- Potential payoff – how much you can win
- Skin in the game – what you risk losing
So, to understand financing – and to get the most appropriate financing – one must understand uncertainty.
Too often, though, we base our decisions on a binary view of the world. That is, we act as if the future is absolutely known or hopelessly unknowable. That translates into our propensity to create point forecasts. Alternatively, we fly by the seat of our pants without the aid of any model.
What is a Point Forecast?
We dodge the topic because we don’t want to get bogged down in an intimidating blizzard of probability calculations and arcane statistics. There is, however, a useful middle ground between abdicating responsibility to address uncertainty and becoming statisticians.
Four Levels of Uncertainty
Hugh Courtney offers a useful framework for assessing uncertainty in his book, 20/20 Foresight: Crafting Strategy in an Uncertain World. Courtney presents four levels of residual uncertainty:
Residual uncertainty is what is left over after you have taken all available information into account. Residual uncertainty isn’t what you don’t know – it’s what you can’t know.
By characterizing the kind of uncertainty you face, you’ll gain insight into the type of financing best suited for your business. You’ll be better equipped to maintain the flexibility to adapt to changing circumstances while giving investors and lenders what they need.20/20 Foresight
A Clear Enough Future
Level I uncertainty exists when there is a lot of valid data, and the business environment changes slowly. In such circumstances, the range of possible outcomes is narrow enough that you can make a point forecast with confidence.
For example, when an established fast food chain such as McDonald’s® estimates first-year revenue for its umpteenth store, it can make do with a point forecast. The forecast won’t be perfect but it will be good enough to make informed decisions. In general, established products and services in stable markets are the most likely to face Level I uncertainty.
In Level II uncertainty, you can be sure that one of a limited set of outcomes will occur. That’s often the case when an industry is facing regulatory change or when there is a “war” over technology standards or formats. You don’t know in advance which format will win, but you can identify the contenders.
A generation ago, when analog video tape was state-of-the-art, it wasn’t clear whether the video cassette standard would be the VHS or Betamax format. However, you could know that it would be one of the two.
More recently, there was a similar shoot-out among competing memory card formats. Ultimately, the SD format emerged as the accepted standard.Format War
A Range of Futures
Level III uncertainty is a bit like Level II in that you can identify the range of possible outcomes. However, in Level III you can’t be sure where on the range you’ll fall.
Level III uncertainty is increasingly common as the velocity of business increases. When you launch a substantially new product or service, you’re likely to be facing Level III uncertainty.
Consider the case of the A380 super-jumbo jet. Back in 2000, when Airbus was deciding whether or not to move forward with the new airliner, it had the benefit of years of experience in the commercial airline industry. Furthermore, Airbus could derive insight from copious data related to its competitor, Boeing’s 747 series. Nevertheless, there was still a lot of residual uncertainty regarding potential demand. After all, the A380 had a price tag of over $400 million and required modifications to most existing airports.Airbus A380
A point forecast is good enough for Level I uncertainty. Level III uncertainty calls for more dynamic and qualitative tools. These include scenario planning and system dynamics models.System Dynamics
In rare occasions, we find ourselves lost in the fog of Level IV uncertainty. Here the future is both unknown and unknowable. Analysis cannot determine a plausible range of outcomes, let alone a meaningful point forecast. Only action and time will allow for the generation of valid data that will decompose Level IV uncertainty into a lower level.
A current example of Level IV uncertainty might be the prospects of next-generation nuclear power technologies such as the molten salt reactor. The technological, political, legal, and environmental hurdles appear overwhelming. However, a successful implementation might yield monumental rewards. Although aspects of such technologies can probably be characterized by lower levels of uncertainty, there is a lot of residual uncertainty.Molten Salt Reactor
Introduction to Confidence Intervals
Level IV uncertainty is rare and—with time—will resolve into a more manageable level. Level II uncertainty is a function of specific conditions such as regulatory change or technology format wars. While it’s important to be able to recognize Level II and Level IV uncertainty, most of us won’t deal with them very often. So, let’s focus on Levels I and III.
A characteristic of uncertainty is the range of plausible outcomes for a key metric – future sales for instance. Level I uncertainty is expressed as a narrow range around a point forecast. (A point forecast is also known as the “expected value”):
The range of possible values in which we can have a high degree of confidence is called the “confidence interval.” In other words, Level I uncertainty is characterized by a narrow confidence interval.
For example, when faced with Level I uncertainty you might conclude, “My analysis indicates sales next year will be $1 million plus or minus 4%.” That implies a confidence interval ranging from $960,000 to $1,040,000 of sales next year.
In contrast, Level III uncertainty reflects a much wider confidence interval – whether or not you make a point forecast:
In this case, you might say, “Our best guess is that first-year sales of our new product will range from $40,000 to $200,000.” Instead of basing your related decisions on a point forecast, you might model and examine a host of plausible scenarios in order to help you recognize and adapt to emerging realities.
Horses for Courses
“Horses for courses” is a British idiom that means that it’s important to choose suitable people for particular activities. Different providers of financing respond very differently to broad or narrow confidence intervals. That is, they behave very differently when confronted with Level I or Level III uncertainty.
In other lessons, we’ll explore how – and why – investors react to uncertainty in different ways. Knowing the differences will help you focus your energies on the right kind of investor for your business.
- Financing decisions involve balancing risk and reward, and risk is – in part – a function of uncertainty.
- We can avoid taking a binary view of uncertainty.
- In business, we most often confront Level I and Level III uncertainty.
- Confidence intervals describe the range of likely outcomes for a key variable such as sales. A narrow confidence interval characterizes Level I uncertainty. A broad confidence interval characterizes Level III uncertainty.
- Different types of investors respond to uncertainty differently.
- Understanding the circumstances that give rise to different levels of uncertainty will help you manage your business better. It will also help you evaluate the fit of different types of financing.
Questions to Ask
- What are the key drivers of the performance of my business?
- How much information is available about those drivers?
- Are my products or services new to the world, or do I participate in an established, stable industry?
- What external factors beyond my control can influence the performance of my business over time?
- Given the preceding, what level of uncertainty best characterizes my business?