Getting a Handle on Uncertainty
Frank Knight looked for the reason why firms are able to make a profit (in perfect competition situations that is) and he ultimately decided that firms were paid for UNCERTAINTY. He then went on to distinguish uncertainty from risk. Risk is the toss of the dice. With risk, the frequency & severity distribution of possible outcomes is known. Uncertainty differs fundamentally from risk because with uncertainty, the future likelihoods are unknown.
You are uncertain, to varying degrees, about everything in the future; much of the past is hidden from you; and there is a lot of the present about which you do not have full information. Uncertainty is everywhere and you cannot escape from it. Dennis Lindley
In risk management, we tend to treat everything as if it were a Knightian RISK and totally ignore UNCERTAINTY. We do our best job of estimating the frequency distribution of gains and losses and treat every best estimate the same. See Sins of Risk Measurement.
But we can and should make an effort to identify the uncertainty that lurks, to vastly differing degrees within our risk measures. A simple start to such an effort would be to develop a classification system for UNCERTAINTY.
- Almost Totally Certain – like a prediction of time of sunrise. No experience contrary to predictions and good reason to believe that there will not be a regime change in the event. Highly unlikely that any human activity will fall into this category. Humans are just not this predictable.
- Highly certain – like a prediction of the Cubs not winning the World Series. Never happened, but it is possible, but highly unlikely that there will be a regime change. Things in this category will be things that there is a long amount of historical evidence. The possibility of a fall in home prices were felt to fall into this category, but the historical evidence turned out to be from one single cycle. To put something in this category, a firm should have direct experience with the activity in question so that there is insight within the firm about the reasons for the historical drivers of the seemingly highly certain event.
- Conditionally certain – Apple will stay successful as long as Jobs stays healthy (oops). For these sorts of uncertain events, the firm should have a that clear idea of the drivers of a string of predictable experience and an understanding that the driver(s) are not themself highly certain events.
- Somewhat uncertain – “Bill says that it takes him 20 minutes to get to the airport” or “it usually takes me 20 minutes to get to the airport but sometimes it is an hour.” Here the firm either has only moderate amounts of experience to judge the actual uncertainty and the event seems to be fairly certain or else the firm has experience and knows that the event is somewhat uncertain.
- Unknown uncertainty – “this is the first time I am parachute jumping and I plan to land in my backyard lawn chair.” Something new. With only limited knowledge of other people’s experiences and not enough experience to know whether there are significant differences in the drivers.
The first time a firm does an economic capital model, they might classify the result as having Level 5 uncertainty. Over time, some calculations might move up to Level 4 or Level 3. In a few areas, the firm might have been doing risk calculations for a particular risk over much longer time and could move up to Level 2 uncertainty there.
But change the question from an estimation of a 1-in-200 risk to a “will this project make money or not” question and is is quite possible that many of the answers might have Level 2 or Level 3 uncertainty.
But firms should try assigning Uncertainty ratings to their efforts. And track over time the degree to which the firm is devoting resources to projects with Level 5 Uncertainty.
Riskviews has worked for several firms that were over 100 years old at the time and those firms usually were very uncomfortable taking on any Level 5 Uncertainty. Most often they kept those activities small until they gained experience. When they went for long periods of time with no Level 5 Uncertainty, however, they tended to shrink relative to the rest of the industry.
On the other hand, the financial crisis was touched off by Banks and other institutions who committed to enough Level 4 and Level 5 uncertainty to send them over the edge. Investors would certainly be interested to know how much Level 5 Uncertainty that a firm is taking at any point in time.
Using an Uncertainty scale like this and discussing the reasons for changes to the level of commitment to higher uncertainty projects will be a healthy and productive exercise for many firms.
Tags: RiskYou can comment below, or link to this permanent URL from your own site.