Dealing with Uncertainty

In their way of trying to deal with uncertainty, I see some portfolio managers are looking closely at the 2008-2009 period to see how their indicators responded at different phases. They want to make themselves feel less uncertain and more in control of the outcome, so they spend their energy trying to figure out what’s going to happen next. They do this by studying the last period they wish they had avoided and try to find the perfect signal. It may be self sabotage- they create an expectation about how it should play out. Then, they worry that it may not. When it doesn't, they are caught in a trap.

A decade ago I completed a vast study of market cycles, so I have a strong understanding of how big trends can unfold. While that probably makes me feel much more comfortable, it’s probably not what you think. Having a strong frame of reference about different market conditions doesn't mean we know what’s going to happen next. Instead, I am aware there is a wide range of outcomes that could happen.  For example, I can tell you that after the stock market has declined about 20%, as it has since May, it may stall and have a period of indecision. That period of indecision is evident by the prices oscillating up and down for a while, a non-trending “trading range” state. During this period, the people (market) are indecisive about whether to stay in, get out, or enter new positions. Many of them are trying to figure out what’s going to happen next and they trade what they believe.

Actually, they trade what they expect. People hate uncertainty.  People tend to prefer knowing over not knowing. We are taught that it’s good to know, bad if you don’t know. You can probably see how people have such a hard time with not knowing what’s going to happen next. Human nature is to try to figure it out because that’s all they know to help them feel more secure, certain, and in control. But in fact, none of us really know.

How we do differ is in our frame of reference about the possibilities. I accept and even embrace that anything is possible because every new moment is unique - it hasn’t existed before. Having closely studied all the historical trends that have occurred in the past, I am aware of the wide range of what is possible. But, I’m also aware we haven’t yet seen some conditions we’ll see in the future that haven’t ever occurred before. I know this because I’ve simulated trends that haven’t ever occurred before. If we only have 100 years of actual data to study, realize that in no way includes all the possibilities. Knowing that price trends can go down much farther than people expect was motivation to develop my active risk management tactics. Many portfolio managers have scrambled to learn how to limit their losses after they had large losses during the 2008 and 2009 period. I have developed my tactics and systems for nearly two decades because I have always known the edge that can be gained from understanding that losses compound exponentially.

During the summer of 2008 the stock market had declined about -20% and entered a period of prices oscillating up and down.  Buying demand and selling pressure weren't out of balance enough to sustain a directional trend.  The stock market conditions are similar today. Based on history, I can tell you it’s when the stock market declines more than -20% that the world markets become serially correlated. That is, at some point the stock market loses enough value that investors start to sell just because it’s gone down a lot. They hit their “uncle point” and tap-out. A waterfall decline becomes a cascade when panic selling occurs. That’s when other markets like bonds, gold, etc. may start to participate. In other words, a -20% decline may not be enough to incite panic selling, but as their losses mount they may eventually tap-out. When they do, other markets become serially correlated to the downside. During the late 2008 cascade, for example, bonds, currencies, and even commodities like gold declined. I knew that was possible before it happened, so my systems were designed with the reality that anything is possible. Today, everyone should be aware: diversification alone isn't sufficient to control or limit your downside. There is exposure to the possibility of loss in any position. 

So, you can probably see why studying just the last crash and how your indicators responded to it may not make you any more certain about what to do next. Yet, maybe the bigger issue is how they may set their expectation based on it. When your decisions are based on your expectations of what’s going to happen next I think you set yourself up to be even less able to deal with uncertainty. I’m a big fan of uncertainty. I know exactly what I’ll do next, no matter how it all unfolds. Rather than an expectation about what’s going to happen next, mine is derived from a mathematical expectation over thousands of different periods and my self worth (ego) isn't determined by how well it plays out any one time. I flip a lot of coins over time and I'm not betting our existence on any single outcome. 

Expected Value and Errors in its Calculation

This is an outstanding presentation by Dan Gilbert about expected value and  potential errors in calculating it. Gilbert first explains the expected value equation:

Expected Value = (Odds of a Gain) x (Value of a Gain)

If we can estimate and multiply these two things we will always know what decision to make - how to make the "good bet". Of course, these things are probabilistic, never certain.Many people don't think their decisions are a bet or gamble at all - they assume they can be right, so they don't consider what may happen if they are wrong. They are the people who have unexpected bad outcomes and they're unprepared for them.

There are errors in estimating the odds of an outcome and errors of estimating the value of the outcome. Gilbert goes on to explain the behavioral errors in these estimations.

Expected value is the little-known foundation of tactical decision-making in portfolio management. To be a great portfolio manager (among the few with a real edge) one must have a system of decision making that uses the probabilistic nature of outcomes to create a positive expectancy portfolio management process. That is, a method for entry, exit, and bet size that results in greater average profits than average losses. It isn't about stock picking or perfectly timng the market - it's about generating larger average profits than average losses that leads to a positive asymmetric imbalance between the two. The pursuit is one of magnitude (size of the profit) not frequency (probability). Those two make up the expected value equation.Once a person grasps what this means to portfolio management, then you have to figure out all the ways you'll play games with youself in creating the expectation. If you get beyond those errors and illustions then you may find an edge: average profits > average osses that looks like the graphs on the top of this page.

A hat tip to John Kopp for sending me this video.

http://www.ted.com/talks/dan_gilbert_researches_happiness.html