Investing requires a healthy dose of decision (tree) making. Good decisions themselves are based on how to think critically, how to estimate probabilities, and how to estimate expected values. We also must be aware of our own intuitive thinking that can unknowingly introduce errors in judgement. Despite getting everything right and avoiding errors, the outcome of any investment decision will still reflect a mix of luck and skill. Therefore education is a big part of making sure the decision making process is sound. The following are several topics that I find important in shaping my investing process.
History of the Word “Invest”
Before we begin, a quick history on the word invest. Invest and investing comes from Latin investire, which means to clothe and figuratively to give new form. One can think of giving one's cash a new form, equity. Or even more literally, the investor takes a new form as a stakeholder or better yet a shareholder. The possibility or even the idea of investing as a shareholder took off in the 1600's with the East India Company, established in 1600, and the Dutch East India Company, established in 1602, both of which were joint-stock companies. These companies were operating with significant profit potential but with much risk and required large amounts of capital to operate. Investing incorporated risk and return from the beginning. (One of the first insurance companies was also a joint-stock company the Fire Office, which was founded in 1680 to insure against fires in London.)
History of Market Efficiency
Prior to the 1960's, the idea that markets were hard to beat was common but there was no theory tying economics to market prices. Academic finance which took form in the 1950's and 1960's was focused on discovering the “why” of market price movements. A central pillar is Eugene Fama's Efficient Market Hypothesis (EMH). EMH along with rational equilibrium economics provided the economic firepower behind the random walk of market prices.
The random walk was explained by virtue that all existing information was included in an efficient market so that the next move wasn't known and thus the best estimate is the current estimate. Formally this is what is called a martingale process where the next value is expected to be the current value regardless of any prior values. It is competition between investors that balanced supply and demand to an equilibrium where the expected risk of an investment should receive the same expected return. The risk of an investment is what drives the return, not past performance. The intuition gained was that a good stock does not unilaterally mean it will be a good investment.
Note that in a truly random system, whether there is independence between prices today and prices tomorrow (i.e. no pattern or memory), there should be no investing strategy that should consistently win. Variations of this argument are what are often made in support of market efficiency.
The Arbitrage Paradox
The idea that markets are hard to beat is rarely contested. However the notion that the (stock) market is efficient and prices are random has been a bit more contentious. A major stumbling block is that we can't know all investors expectations for tomorrow's price, which means we can't test for market efficiency directly. The theory is really arguing that the markets are efficient because all new information is rapidly incorporated into prices.
The information efficiency argument itself was challenged in 1980 by Sanford Grossman and Joseph Stiglitz who published On the Impossibility of Informationally Efficient Markets. The paper starts with a question:
"If competitive equilibrium is defined as a situation in which prices are such that all arbitrage profits are eliminated, is it possible that a competitive economy always be in equilibrium?”
Arbitrage is an activity where you find situations when you can buy a cheap asset and sell the same amount of a like or very similar asset that is expensive. The argument goes that there is a cost to obtaining information and therefore prices cannot reflect all existing information otherwise there would no incentive to obtain costly information.
It's like using calculus to find the slope of a curve at a given point despite not being able to calculate the slope at that point. The trick is to start from the left or right and inch your way towards the point using a line as an approximation. In highly efficient markets we can get the approximation line pretty close to the curve at the point. In markets that don't have as many investors price searching, then we won't be able to get as close so there can be more error in the estimate.
Skill vs Luck
Do some investors have skill (i.e. market beating returns) or is their success based on luck? To be consistent with EMH and that information bears a cost, one must not expect to find skill beyond what a statistical distribution of results might allow. Specifically the argument against skill is based on the lack of persistence of an individual's returns (above the market). If there isn't persistence, then it's more likely that it came down to luck.
However if we follow the logic of EMH, we should find that competition for asset managers is the same as for assets. Consider an investment manager who has shown skill (i.e. persistence of above market returns) in the past. In a competitive market, investors will invest capital with this manager and the excess demand will drive down returns as the risk and return normalize. Otherwise everyone can buy the fund and enjoy beating the market, which isn't possible. Therefore persistence in it's own right isn't evidence that some managers don't have skill or that profitable trading systems don't exist but rather that capital markets are competitive.
A Day at the Horse Races
Consider a day at the race track. The track offers pari-mutuel betting, where a group of bettors place their bets in a pool. Bettors who win split the pool after the race and after the race track takes its cut. The more bets there are on a particular horse, the lower the payout given the even split among the winners.
Betting offers a chance to calculate expected values based on the odds and the payoffs. Sports betting is efficient when payoffs are appropriate given the expected odds. As the intuition learned from EMH, the best bet isn't always on the likely winner given the associated slim payout. Skill is in recognizing where a bet (if any) is undervalued based on the circumstances and acting on the insight. It's important to remember that when your placing your bet, you are betting on the actual race as well as against how everyone else bets. This is just as true in deciding what security/fund to purchase in financial market.
You are asked to play a game with a group of people. Everyone is to pick a number between 0 and 100. The winner is the person who picks the closest to two thirds of the average of all the numbers picked. The results provide insight into what Howard Marks would call second level thinking. The average number should be 50 and two thirds of that number is 33. However if everyone had that same reasoning then the bias would be to 33 and therefore you might consider choosing 22 or two thirds of 33. If everyone knew the game and played rationally, everyone should bet 0.
Results from when groups actually play the game are spread across from an impossibility of winning 100 to 0. There is usually a sizeable lump of guesses in the low 20's, with more results in the lower third. It doesn't matter if you have played the game before, you don't know how many others have played and how they will influence the average. This next level thinking is the same kind of thinking required in pari-mutuel betting to properly analyze how everyone else is betting as well as how to evaluate a publicly traded security.
The Knapsack Problem
So you are taking a trip and are limited to bringing one bag. You can't fit everything so you must choose which items you keep and what items you leave behind. Each item has a value and a size associated with it. It's your job to figure out how to maximize the utility of your bag. This requires combinatorial optimization and is literally known as the knapsack problem.
Experiments involving knapsack problems have shown that as the complexity of an optimization problem increases, the ability to optimize degrades. This is the same for people as well as computers. Evidence from experimental studies indicate that indeed, we have bounded rationality as argued back in the 1950's by Herbert Simon. We are limited in our ability to solve complex problems and we end up using our intuition and heuristics or mental short cuts. Intuition however can lead to biases in our decision making as we don't frame the situation properly. Experimental results also show a diverse sample of solutions by individuals, which fits most people's intuition that some will be better at solving puzzles than others.
A short coming in my opinion with the economic rational supporting EMH is that it assumes people are rational and can all solve complex optimization problems the same way. We know this isn't true in the real world. However in market trading experiments, participants who showed skill in optimizing problems had no way of knowing if their optimization solution was correct so they ended up continually searching. This searching kept prices in some disequilibrium.
What does John do?
A friend tells you about their neighbor John. John is shy and withdrawn, but more often than not has been a helpful neighbor. He shows very little interest in people or engaging in the real world. A meek and tidy soul, John seems to need order and structure and has shown a passion for detail.
Do you think John is more likely to be a farmer or a librarian?
The question is adapted from an experiment by Amos Tversky and Daniel Kahneman whose research has shown that we humans think with intuition, which often can mislead us. Consider that there are many more farmers than there are librarians. Second consider the percentage of male farmers is much higher than the percentage of male librarians. Therefore regardless of the color of John's parachute, which we were more likely thinking about, it's more likely he is a farmer than a librarian based purely on statistics.
Tversky and Kahneman have unearthed many heuristics that can lead to errors of estimating probabilities and ultimately to making popor decisions. Charles Munger highlighted even more in a speech when he listed 25 misjudgements we risk when making decisions. It is important to be mindful of any heuristics being used and their ultimate usefulness.
After behavioral finance established itself in the 1980's, there was a shift away from asset pricing theories and towards empirical results that indicated risk factors and associated return premiums. Since the early 1990's, we talk about exposures to different risk factors like small company risk or value (i.e. low market to book) risk.
Because there isn't a universal asset pricing model, it's impossible to know “why” certain factors exist. They could be driven by fundamental business risk or they could be driven by investor behavior. The Royal Swedish Academy of Sciences didn't try and figure it out and in 2013 awarded the Nobel Prize in Economic Sciences to Eugene Fama (academic finance) and Robert Shiller (behavioral finance).
In 1997, Michael Carhart wrote a paper adding momentum as a fourth factor to the existing market, size, and value factors established by Fama and French in 1991. Unlike the first three, the momentum factor has no direct ties to economic theory. Since those original factors, what we have seen over the past several years is an explosion in investment factors.
There are now so many it's hard to know which ones are significant and which ones are just artifacts in the data. Fischer Black was skeptical back in 1993 when he wrote Beta and Return. He argued that many of the anomalies found were likely due to data mining especially when there are no reasons given for the effect in the first place.
I'll leave with Black's quip when he wrote:
“strangely, the factor that seems most likely to be priced they don't discuss at all: the beta factor.”
The beta factor would be long low-beta stocks and short high-beta stocks. This is now commonly referred to as the low volatility factor.
Or Just Risk
Risk is the simple fact there are more than one outcome we have to make provisions for. That's why most decisions we make are risky, based on probabilities and inductive reasoning. The problem with risk is that we often forget about it. The longer it remains invisible the less we concern ourselves with it and we the more we focus on the risk of missing out. This is how bubbles happen when the focus is on generating returns with little concern for risk. This is exactly when risk grows. As Warren Buffett said:
"Only when the tide goes out do you discover who's been swimming naked.”
I'll leave with a really good synopsis worth reading more than once by Howard Marks from his book The Most Important Thing on page 45.
”… most people view risk taking primarily as a way to make money. Bearing higher risk generally produces higher returns. The market has to set things up to look like that'll be the case; if it didn't, people wouldn't make risky investments. But it can't always work that way, or else risky investments wouldn't be risky. And when risk bearing doesn't work, it really doesn't work, and people are reminded what risk's all about.”