The future is one thing that is truly unpredictable. Unfortunately, most people have a problem handling uncertainty and investors are no exception. In fact, extreme events with large impacts happen more often than we think. Are you tricking yourself when it comes to random events? Find out how you could, and should, protect your portfolio against these extreme situations.
Predicting the Future
Ancient kings, Eastern emperors and tribal leaders all had one thing in common: the desire to have knowledge that revealed the future. For this reason, they usually took on some kind of astrologer or omen interpreter. Today's leaders no longer rely on astrologers, but employ armies of statisticians from various sectors of government to give them advice of what might happen next. One problem, however, is that most statistics are rooted in the Gaussian curve (or normal distribution), while most events and people in the world are not. The problem also has its grip on the world of finance.
Randomness in the World and the Financial Markets
Look back at the greatest groundbreaking discoveries in history. The Americas were discovered while looking for a new route to India. The discovery of penicillin occurred by accident. These events changed human history in very large ways, and we can learn a valuable lesson from them: large extreme impact events, similar to the ones above, are not usually found because we are searching for them, but more often because we are exposed to them.
So how does this principle apply to the markets? Looking at the price changes in the financial markets as a random walk, we can see just how this exposure can be beneficial. Let's take a look at a fairly volatile industry, such as the biotechnology industry. Many of these firms rely on very concentrated sources of cash flows to operate. They are all looking for the new blockbuster drug. Some will get lucky and get their drugs approved and go on to blockbuster status; others are not so lucky and will go bankrupt.
Our Models: Triumphs and Failures
Having a complex system of mathematical models spit out a few numbers about what we can expect gives investors peace of mind. Things like intervals with "95% certainty" make us even more assured of our future. The underlying problem, however, is the use of the Gaussian curve. Gaussian-based models, which have, say, a theoretical error rate of 5%, often seem to produce much larger empirical error rates than theoretically predicted - in this case, larger than 5%. The problem is that in real life most distributions of probabilities seem to fit a Gaussian-like curve with "fat tails", which generate larger probabilities for extreme events. This is not to discredit about the use of statistics, but rather their implementation in situations where they do not apply.
Almost systemic use of Gaussian-based models in finance leaves us with opportunities to exploit what can now be viewed as the mispricing of securities with respect to large extreme events. Such opportunities would include setting up long volatility positions across a very wide spectrum of securities (remember it's not about prediction, it's about exposure) in a fundamentally volatile industry to capture a few idiosyncratic extremes. Using a strategy like this, you're risking pennies to win dollars.
Thinking of implementing this idea? You could run your portfolio as follows: Invest 95% of your capital in U.S. Treasury Bills and use the remaining 5% to set up positions across a very wide spectrum of securities similar to what we talked about above. Of course, the percentages can vary to adjust what you find acceptable as a maximum yearly net loss after accounting for interest earned on your Treasury Bill position.
Handling Uncertainty: Investing Like A Cow?
You still may not be convinced about our inability to predict correctly, so let's continue our discussion. When extrapolating historical data into trends or simply extrapolating trends themselves, just think of a cow raised for meat. This may not seem like an apt comparison, but the cow's belief is actually very similar to that of an investor who is having a good run. It all starts out with the cow being well fed and generally well taken care of. This happens day after day, and as each day passes the cow becomes more confident that this will continue and that its best interests are being looked out for (unbeknownst to the cow, the objective is to create food for humans). Unfortunately, this confidence doesn't save the cow from a trip to the slaughterhouse.
This example reflects investors who merely extrapolate current numbers into future periods. Historical predictions and confirmations, no matter how correct, have nothing to do with the validity of the model. As you can see, the mental model the cow made in our example failed horribly the day he was brought to the slaughterhouse. The point is that no matter how many previous confirmations you get from a model or method, it proves absolutely nothing about the validity of the model or method itself, as one extreme event can disprove everything. As a result, you need to be prepared for extreme events, regardless of how well things seem to be going.
Conclusion
Our inability as humans to handle the uncertainty associated with the future puts us at a natural disadvantage, which can affect our portfolios. However, we can acknowledge this uncertainty (although it takes effort) and prepare ourselves for the various outcomes in order to protect or benefit our portfolios. Through a combination of hedging and designing portfolio strategies to benefit from limited exposure to extreme events, we can triumph over our natural flaws as humans.
No comments:
Post a Comment