Chapter 39
Trends, Data, Correlations, and a Few Other Things—Again
A trader may be able to exploit an intra-day trend in a stock and have it play out, but economic trends tend to last longer. They can be driven by bureaucratic or politically controlled “regimes” or from unplanned forces that dominate the economy. An example of a controlled regime would be the “easy money” low interest rate policies put in place after the great recession by many central banks around the globe. An example of a more unbridled trend was the digitization of media, which started in the early 1980s with music CDs106 and has since included all forms of media and data. While the intra-day trade is valuable, shifting capital to invest in a major trend is very powerful for long term investors. Medium- and long-term economic trends tend to transition rather than collapse and often include a long unwinding tail that can give ample opportunity for an exit without too much degradation of asset value from the peaks. These trends do not follow straight lines; there will always be some divergence along the way. If you have a long-term time frame, these undulating waves that occur during a long-term trend can cause entry and exit points around a core position. Tsunamis are much less common than waves.
Economic trends can be very narrow or broad in their impact. For example, a shortage of cubist paintings and a growing demand among collectors may prove to be a long-term trend, but is unlikely to impact the broad economy. Whereas, a massive new oil discovery will have broader impact on the region in which it was found but also possibly across the globe. Major broad reaching trends typically are signaled by multiple data points and impact values across several types of assets. For example, the structural long-term decline in interest rates that started in the early 1980s in the United States was signaled by actions by the Federal Reserves, moves in the dollar and measures of inflation. The decline in rates impacted home values, bonds, currencies, equities, and the investment decisions of businesses. The global economic impact of an increase in the prices for cubist paintings would probably not transmit as broadly through the economy.
You want to monitor data and look for trends, but the data you are using is always old. The components of the economy may be very different now in critical ways than in past cycles, which can limit the value of using historical data blindly. Analyzing which historical data is relevant and putting it into context for the current economic situation will lead to better decisions. No single piece of data is “always” the best to use. Various environments will require the use of different data, especially as change happens more rapidly. Analysis that is aware of the forces of change and how they are impacting the value of the data being used is what is valuable.
Markets often lead the economy; they can react quicker than the economic data does. However, markets are reactionary and often give false signals. A market price on an asset, despite some false signals, is still valuable data. It is fleeting but it gives you the best real-time data point on value. It adds to your knowledge on how others view the value of the asset. Long-term patterns of prices can be more critical data than one data point in time and you can always overweight the latest price point relative to the one from a year ago. Focusing less on individual prices or variances and more on the broad changes in the direction of asset prices can help spot trends.
It seems like almost every investor is drawn to try to find correlations between data and asset prices, hoping to find the magic formula. Investors often grow dependent on correlations between some specific data set and a specific asset valuation. Sometimes the strength of these correlations breakdown and create an opportunity because they quickly revert. However, sometimes things have changed and the correlations no longer work. This is happening more rapidly in the economies of today. Correlations do not usually last forever, and analytical logic can be critical in recognizing when a change in correlation makes sense.
Simplicity has been popular through the ages. William of Occam, a medieval logician, is still remembered for teaching to avoid complexity when simplicity will do.107 Albert Einstein, the legendary physicist is best known for a simple formula E=MC2 that changed the world. Simplicity can often be the best path for good decision making. With constant changes in the economy the value of using any data point is going to vary considerably over time. The more different data sets you are utilizing in a model the more likely some of them have become irrelevant. Keep it simple, fewer moving parts, fewer things break. Despite timeless advice about simplicity, there are some people that take delight in building complex models or doing things the hard way.
Models work well when the data inputs are easy to track and monitor. Trying to find investment themes through highly complex models with a multitude of variables can confound an investor with many false signals. Simple models allow more careful tracking of how the data interacts and affects the economy. This does not mean advanced concepts and ideas are not valuable. However, investment themes work well when they are based on inputs that can easily be analyzed to see how they are changing over time. Models also should not be binary; several scenarios with probabilities can drive better decisions than a single view of the future.
No investment trend will move in a straight line. At some point prices will move against a thesis and if the data and the thesis have not changed, the investor needs fortitude. Alternative viewpoints should be examined and ancillary data looked at, but investors should not be afraid to stick with their convictions. For example, just because something has been going up for some time is not enough of a reason for it to go down.
Asset values often move when results vary from expectations. Missing or beating expectations is very different than a change in a long-term trend. However, for very short-term trades, results relative to expectations are more important than the actual result. An over-reaction in the market to miscalculated expectations can create an opportunity for investors with long-term convictions. Understanding what is driving the expectations of others is important because that will be part of how they set the price at which they will be willing to sell you an asset and where they will be willing to buy something when you are ready to sell.
Changes in the level of uncertainty cause valuations to reset. If uncertainty increases enough, shocks can happen. Like a championship fighter takes some punches but does not get knocked down, shocks can happen but don’t always lead to a crash. Valuations can adjust to the uncertainty over time and just as uncertainty can increase it can also decline. A few shocks, or sometimes even a single good one, can change people’s tolerance for risk. When this happens people will pay more for risk aversion. Monitoring a market’s tolerance for risk can be quite useful, and following relative movements in data points such as credit spreads and prices of safe haven investments can help.
The time horizon that you have for your investments needs to be matched with the types of investment themes that you choose to pursue. The decisions you would make managing a pool of money that is not going to be touched for fifteen years will likely be different than a pool of money that might get taken out after twelve months. The time periods you choose to measure investment performance can also greatly impact how you invest. It is human nature that you will think and act slightly differently if your performance is questioned twice a day, quarterly or annually. It is odd that everyone seems to measure performance off the same calendar year; a study of how performance varies if someone is measured off-cycle, such as every five, or every fifteen months, would be a good project for a behavioral economist.
It is not worth changing investments every time the wind blows. However, any investment thesis should be questioned constantly and examined using various sources of data because everything is always changing rapidly.