As a Professor of Operations Management I am often reminded of the huge contribution of Andray Markov. Markov was a Russian mathematician who dedicated most of his professional life to the study of stochastic processes. The study of what? A stochastic process is one that moves from one position (or state) to another in a random way. In fact, the timing of the move, the direction of the move, and the size of the move can all be random variables. Such processes are notoriously difficult to understand and are typically impossible to predict.
However, Markov specialized in a small subset of these processes that were “memoryless.” This means that the path taken to the current state or position tells us nothing about the next step. In such cases, the best that can be done is to use history to estimate the distributions that define the step timing, direction, and size. These processes ultimately came to be known as Markov processes. The analysis of Markov processes have found applications in a wide array of problem settings including queueing theory, inventory theory, and Yes – stock prices. In fact, much of the research on stock prices in the first half of the 20th century focused on estimating the parameters needed to state a Markov process that described price movements.
The most widely heard representation of such processes is probably the “random walk.” This is the idea that if we look at price movements over some longer span of time and estimate the share of movements that are up or down, that is the best that we can do, in terms of prediction. Just as a point of reference a commonly used data set in analytics courses shows that the S&P 500 went up roughly 56% of the time and knowing what happened the previous day, or the two previous days, etc. adds no predictive power. If stocks follow such a process, then the answer to the question “what direction will prices move tomorrow” is always the same. “I don’t really know but there is about a 56% chance that it will go up.”
Such a statement enrages the stock picker who is typing right now to call me every synonym for idiot that he can think of. “Movements are not random, dumbass. There is always a reason!!! You are a moron.” Fortunately, multiple things can easily be true. I probably am a moron, and there is always an explanation for the price movement – AFTER THE FACT!! Turn on the financial news television or radio after the closing bell and you will hear a beautiful narrative about why “the market” went up or down that day. “The market rose on expectations of rate cuts,” or “The market fell due to tensions in the middle east,” or “the market went up and then fell because Mars was in retrograde until 11 AM and Taylor Swift sneezed in the afternoon.” OK, I made that last one up. The point is that social, pattern seeking animals like humans will universally grab hold of a good narrative that “explains” what just happened. Furthermore, we will almost always conclude that we saw that coming – it was just common sense, we will all say.
Perhaps you recall that fine spring day in 2008 when Ford stock sold for $0.97 per share, or that day in 1997 when Apple sold for roughly $3.75. We all “KNEW” that was a great deal – but you didn’t buy it on those days did you? You know why? Because it wasn’t at all obvious at the time that this was a good buy. In fact, common sense screamed the exact opposite. Writing the story after the data is in is always easy, but irrelevant. The trick is to see the story before the data writes it. If you can consistently do that (and you can’t) then you are a great stock picker. If you cannot – then go back to work, let the market do its thing and check on it a little later – like a decade or so and see what you have.
The fundamental value of a share of stock is the present value of future related cash flows, including dividends and profits. This does not change because the stock rose from where it was last week, or last month, or shows some pattern relative to its 200-day moving average. (Many readers just lost their minds. Hold on. It gets worse.) Note that this is not to say that history doesn’t matter. McDonalds historically has done better when the economy slumps because it will attract bargain shoppers in tight economic times. This isn’t changing any time soon. In that sense, history certainly does matter. But that history has already been “baked in” to the current price because it sets up expectations about the future. The exact pattern of price movements over the last 30, 60, 90, 180, or 360 days has nothing to do with any future price changes.
The simple explanation as to why I believe that this is the wrong way to look at it is this. You don’t invest in stocks. You invest in companies. Companies organize the ways that people manage assets and products while other people (called customers) interact with those products and assets. The real people involved did not change their behavior this morning because they saw some perfectly random pattern in the price of the stock.
The worker, manager, customer, or competitor is not managing the chart. They are managing their work, employees, household budget or business. This is why I cannot buy into the fundamental argument of the technical analysts. Its not that history doesn’t matter. It is that the type of history that does matter is already accounted for, and the next business or price move is a result of the actions of thousands or even millions of people who don’t know or care what the chart shows.
Now, let me also deal with the other 800-pound gorilla in the room. Many stock pickers and technical analysts will try to claim that I am saying that markets are perfectly efficient. I am not, and for very good reason. Just as an example let’s consider a famous paper on the topic that is both simple enough to explain, and sufficiently powerful to make a useful point. In 1988, Andrew Lo and A. Craig MacKinlay published a paper that touched upon 2 critical and simple questions. If stock prices follow a random walk, then a number of things must be true. Among them, the weekly returns must be uncorrelated, and the variance of total return should grow linearly with time. In other words, the return from week i should bear no relationship to the return from week i + 1. Looking at data from 1962 to 1985 they find that the weekly returns have a correlation of about 30% and the variance grows faster than a random walk would suggest. They argued (correctly so) that if this is true, then one could make millions trading short term options on small company stocks. They then demonstrate this using historical data. Many professors were outraged over this finding and argued that there had to be some mistake. They were wrong. There was no mistake. If a stock price rises this week, there is a better than even money bet that it will rise next week as well. You could leverage this fact using options or by active trading and make a lot of money.
But, Mr. Smart-ass, you said that markets are highly efficient, and that means that this can’t work – what gives? Nothing. The authors were correct, and a few people made a killing. This worked until lots of people figured out the game. When that happened, they looked to buy the options needed to use the strategy. Once the demand for such options rose, their prices rose, and the easy profit went away. As Lo and MacKinley put it, markets are not perfectly efficient, but they are Adaptive. Some information is absorbed almost instantly, while other information takes longer to filter through the system. However, it will happen after some lag. Once that happens, you have to come up with a more esoteric strategy that will work for a while, until the market adapts, and you have to move on to the next trick in your bag, and so on. Unfortunately, for the 99.9% of people with lives who don’t have 100 hours a week to look for the latest trick, this is a nice story, and nothing more. If you are smart enough to pull this off, and you have to time to do it, knock yourself out.
But here is the dirty little secret – you’re probably not that smart. Most of the people who thought that they were, lost their shirts. Literally thousands of schemes that sound a lot like this one have been tried. A few people really did make a killing, but the vast majority made nothing, or at best, made what the market averaged. I can buy an outcome that beats the overwhelming majority of the speculators with no effort, almost no cost, and the advantage over that crowd will last as long as you or I may live. My only claim is that, that’s the best deal out there, and you should take it NOW, ALWAYS, and FOREVER.
