The uncertainties of uncertainties

We always face all sorts of uncertainties but some of these uncertainties are certain and some are uncertain. The certain uncertainties have known probabilities, for example, when gambling, some games have certain winning probabilities. However, many events have unknown probabilities that why many companies hire actuaries to estimate the risks, in other words, make these uncertain uncertainties less uncertain.

By actuarial methods, individuals and firms can lower their uncertainties when facing significant individual and environmental risks. However, sometimes uncertainties are uncertainties, not because their probabilities are unknown, in many cases, we cannot cover all risk sources and the risks and the risks that we fail to notice become much more uncertain uncertainties. When such uncertainties exist, the system will fail to observe the overall risk in the system and underestimate the systematic risk.

Of course, people could give an estimated value about how much unobservable risk exists in the system, but it is really difficult to identify the credibility of such estimation as there are many cases that we fail to estimate the unobservable risk and get ourselves into great troubles.

Sometimes when we try to get very accurate results, we face greater risks. For example, in a stock market, many institutions give target prices about stocks. Instead of simply telling the price could be higher or lower in a year's time, giving a detailed price has more uncertainties. Knowing whether the price will rise or fall, traders could only choose one strategy from the two strategies in total, buy or sell. When the actions taken by the traders in the market is simple, the uncertainties during the one year period are relatively low. However, when we know a specific price, there are countless strategies that could be chosen, this will increase the uncertainties and the previous predictions are known to be less credible. Moreover, the difficulties of getting a detailed result are much greater, meaning more errors could be made. Therefore, sometimes it is better to give a simple estimation rather than a detailed and specific result, as the results are definitely less credible and could cause more uncertainties.