This is the first part of a series in which I will attempt to lay down a foundation in investing before eventually delving into more complicated topics.
I read articles every day discussing whether a particular asset class (like U.S. stocks, or junk bonds, for example) is overvalued, undervalued, is on a negative trend or a positive trend. The overvalued and undervalued terms basically mean what the author thinks is an asset’s price (objective) compared to its value (subjective). The positive and negative “trend” or “momentum” terms are types of technical analysis and are grounded in behavioral finance. Valuation and technical analysis are the topics of endless courses, seminars, books, scholarly articles and blogs, so the prior two sentences do not to those terms justice, but we have to start somewhere.
Here are some of the major problems with using valuation and technical analysis to predict the future price of individual assets, such as stocks, groups of stocks and even asset classes:
- There are small sample sizes for these predictions. The New York Stock Exchange has been around for about 200 years. There are a few other stock market exchanges, most of which have a similar or shorter history, but even in the aggregate, we only have about 1,000 years worth of stock market behavior to analyze. While this seems like a long time, it is still a small number. As a result, when we try to look at the past to predict the future, in many circumstances there are only a few times, if any, that a particular combination of events have happened in the past, which is hardly enough to have conclusive proof of what will happen next. Oftentimes, it’s like saying that because the Denver Broncos have beaten the Kansas City Chiefs in Arrowhead Stadium 6 out of the past 7 years, that the Broncos will beat the Chiefs next time. In those 7 years, the coaching staff and rosters of both teams have almost completely turned over, such that those prior games have essentially no bearing on the next one.
- The world’s economies and markets are constantly evolving. In the past hundred years, we’ve had two world wars, the widespread adoption of cars, massive urbanization, a drastic increase in life expectancy, significant medical advances, the creation of computers, cell phones and the internet, among many other things. Now we are dealing with below zero interest rates, world-wide terrorism, Brexit and many other events. So many inputs into the market have changed so drastically in the past 100 years that it diminishes how much we can rely on the past to predict the future. To use medicine as an example, it’s like trying to predict how humans are going to evolve next, when in the the past 100 years humans have grown a third arm, stopped using their spleen, grown another foot in height on average. Oh, and instead of having billions of humans to analyze to make those predictions, there are just a few alive at any given time and significant evolutionary changes occur with each new birth.
- Valuation and technical analysis are not completely art or science. Markets are publicly traded, manipulated and constantly evolving. They are governed by an unpredictable combination of people, computers and nature; and as such we will never be able to figure them out exactly. As soon as someone has found a trend, or valuation technique, that appears to be consistently correct, then people will attempt to arbitrage away that advantage by getting in front of it, or predicting that trend or technique before it occurs. People can get lucky, and for long periods of time, thereby making investing appear as an art. People can also analyze massive amounts of data and come up with sophisticated formulas that appear to predict a particular outcome with meaningful success. Unfortunately, what has worked it the past does not mean it will work in the future.
- As a result of the problems described above, people in the investment world perform “back tests“, which involve simulating a trading strategy on relevant past data to see if it would have been successful. Sometimes the researchers have a goal in mind for their research, and so they design a back test that will support the conclusion, making the outcome unreliable. On the other hand, many conclusions reached by investing research are very logical, seem well-supported and the studies themselves appear to be well-conducted. These tests, however, often cannot resolve the above problems of a small sample size, a constantly evolving market, and an immeasurable number of inputs that we cannot fully comprehend, much less predict. While some valuation and technical trading theories likely add some value, there is no way to prove, in advance, that any will add value going forward. In fact, many will appear to destroy value in the short, medium and long run. We don’t have 50-100 years to wait to see if something pans out, and too much is at stake to take too much risk with most strategies.
Over the next few articles, I will go into some valuation and technical trading concepts as well as the pros and cons of the different concepts. In the meantime, please let me know if you have any questions or want me to write about a particular trading concept.
Photo credit sheelamohan at FreeDigitalPhotos.net