Probability, Exposure, and Consequences
Most of the randomness we see outside of financial markets fits nicely in a bell curve. Things like coin flips, dice rolls, the height of humans, and the movement of atoms, all fall neatly under bell curves. The variation is neat and regular. Individuals are difficult to predict, but large groups will display consistent characteristics. If we roll a six-sided die 100,000 times; we'll come very close to observing 16,666.67 of each number. Of course, it won't be exact, but we can calculate the probability of any outcome before we start rolling.
Even still, randomness is a strange animal because exposure plays a part. If we flip a coin four times, there's a 1-in-8 chance we'll see all heads or all tails. Interestingly, if we repeat our set of four flips five more times, we'd be more-likely-than-not to see at least one set of four with all heads or all tails. If we continue repeating our sets of four flips, we'll be increasingly likely to see a set of all heads or tails. Of course, we'll never quite reach absolute certainty; we will, however, get closer and closer to it. One of my high school math teachers used a colorful example to demonstrate this concept. Assume a teenage boy and teenage girl are sitting at each end of a park bench. Every few minutes, they close the distance between them by half. He'd quip, "Eventually, they'll be close enough for practical purposes."
If we lengthen the number of consecutive heads or tails for which we're looking, the probabilities drop quickly. If we're looking for ten-out-of-ten flips to be all heads or all tails, the probability drops to 1-in-512. It's doubtful any of us has the time or desire to repeat sets of ten coin flips long enough to find a collection of all heads or all tails. Therefore, low probability events lie outside of, or on the edge of our experience. We generally don't see them - until we do.
Take, for example, the Thanksgiving turkey. He shows up at the farm as a baby chick in July or August. Every morning for the next 90-120 days, the farmer arrives with a bucket of grain. For the turkey, his morning observations are always positive. Every morning his breakfast appears. The turkey grows comfortable with mornings, even looks forward to them. Then, one November morning, the farmer arrives, and things are very different. Instead of a bucket of grain, the farmer is holding a butcher knife. The turkey is about to experience a low-probability, high-consequence event.
Just because we haven't yet experienced a particular event, does not mean that event is impossible or even unlikely. Like the coin flip example in paragraph two, if we run enough repetitions, we become likely, even nearly certain, to see low-probability outcomes. Standard risk models take into account probability, exposure, and consequences. Quality risk management should address each of these factors. Ignoring high-consequence events simply because they seem unlikely is a recipe for disaster.
Market returns don't always fit nicely in a bell curve, and we can't know the distribution of those returns until after-the-fact. Markets are much more difficult to predict than sets of coin flips. That's the knowledge on which we build our risk management strategies. Quartzite Risk Management LLC's Dynamic Grain Price Risk Management (DGPRM) program doesn't try to predict markets. Since we can't know the probabilities ahead of time, we work with our DGPRM clients to manage exposure and consequences. We're experts in measuring and managing risk. Contact us today for more information.
