In recent months, I have written about common financial planning mistakes I have seen people make. As is often the case, I received several emails from readers with suggestions for more errors I could highlight. Today’s topic comes from a combination of those reader comments.
Given all the information out there about finances, why do we practitioners keep seeing some mistakes so frequently?
Today, the sheer volume of information makes it a challenge to discern what information is “good” and what to do with that information. Almost every time an error is made, I see a misunderstanding of probability, a logical fallacy, a bias, or a combination of the three.
Most people are taught little about probability and statistics in school. I won’t attempt to make up the gap here, but I will attempt to describe how failing to fully grasp basic concepts can lead to issues.
We’ll start by flipping coins. Grab a coin, a slip of paper and a pencil. The mathematics predicts that half your flips should be heads and half should be tails. Flip the coin and write down an “H” for heads and a “T” for tails. Do this twice more for a total of three flips.
I guarantee that no one engaging in this exercise had half the flips come up heads. There is nothing wrong with the mathematics of probability, yet the expected outcome is impossible.You can have zero or one or two or three but not 1½ heads.
Granted, this criticism that the expected result is impossible to achieve is a bit silly. Any odd number of flips results in the same condition. Nonetheless, I frequently see real world errors made based on similar misunderstandings. Most days you can find a story based upon the premise that because a certain result did not occur, the expectation was wrong and therefore relying on that expectation is dangerous.
I see this often with asset class comparisons. A recent example would be something like, “Academics say value stocks have higher expected returns than growth stocks, but value has lagged growth so including value in a portfolio risks falling hopelessly behind.”
In addition to the odd number of flips, another obvious flaw in our exercise is that three flips is a small sample size. Relying on a small sample seems to be a favorite of pundits, often in combination with a post hoc fallacy—because one thing happened before another, that first thing must be the cause. In the financial press, it looks something like this, “We like tech stocks here because the last five times that (fill in economic statistic here), tech stocks soared.”
We also see stories that use a reasonable description of an expectation to justify a biased conclusion. This happens when someone has an agenda. One we have been hearing for a decade or so goes something like this, “With rates low and stock valuations high, we shouldn’t expect as high a return, so clients won’t reach their goals unless they put XYZ in portfolios.”
XYZ is whatever the speaker is selling, commonly an alternative investment. It is true, of course, that rates are low and valuations high but that does not mean clients won’t reach their goals nor does it mean that XYZ would help anyway.
Another common problem with a small sample size is that all trends start with a few steps in a certin direction. The trouble is just because a few steps are taken does not mean a trend has begun.
We see a lot of investment articles that quote someone calling a change in a trend based on small sample sizes. The source will say something like, “With the dollar declining each of the last three months, XYZ is a good play for further weakness in the dollar during the balance of this year”
Sadly, we see people act on these opinions about short-term trends or streaks continuing to their detriment. For instance, they move assets to money managers profiled in the media for recent good results, ignoring weak, longer term track records, tax ramifications, or the riskiness of the manager’s strategy.
While not as directly a result of a small sample size, the “if this happens, we are doomed” piece also does not rely on a lot of data in making its conjecture. It is an example of the “slippery slope” fallacy. Here, the suggestion is if we take a step on the slippery slope, we will inevitably slide toward some dire consequence.
The most common version says if some economic number comes in at some level or a particular event occurs, a recession or market drop will surely ensue. In recent years, this included tax code changes, election results, debt levels and trade wars.
Because economies and markets are far more complex than the narratives suggest and people adapt to changing circumstances, slippery slope narratives rarely pan out as described.
When it comes to extrapolating data into trends, I suspect chasing performance based on recent market moves may be the most common error. People buy after the market, or parts of it, rises and they wish to sell after a decline. It seems likely that many of the people trading crypto aren’t doing it because of a reasoned assessment of the future of crypto assets. They see prices soar and think its easy money because in their minds the probability of success is high.
While sample size is an issue, increasing the number of flips actually decreases the odds of meeting the 50/50 expectation. Flipping a coin twice has four possible outcomes, two of which are half heads and half tails. Flipping four times has 16 possible outcomes but only six of them involve 2 heads and 2 tails. Think of it this way: would you say the odds are high or low that after flipping a coin a million times, exactly 500,000 were heads and 500,000 were tails?
What does become more reliable is the range of results. The expected result is 50% heads. In a three-flip scenario, the range runs from 0% heads to 100% heads. After a million flips, you expect the resulting number of heads to be fairly close to 50% even if you don’t expect exactly 50%.
This is why when a player wins money at the craps table, the casino does not shut down the game. They know the more people play, the more the house will make over time even if they can’t be precisely certain how much they will make. Losing now and then, even losing big is an expectation of the casino, not a fear. The casino focusses on the long term and what they can control.
What if you flipped heads five times in a row? Would you say that was luck or skill? You would know it is luck because you did the flipping. When someone else does it though, you may be inclined to think they possess a special skill (or a rigged coin). This can be particularly true if the person said they would flip heads five times in a row before they started. Your perception of the event is influenced by the person’s declaration.
Differentiating between skill and luck is not always easy. Given the thousands upon thousands of investment managers in the marketplace, there are going to be quite a few that will appear skilled but are just lucky.
You are more likely to believe someone has skill if they tell you they have skill even if their results are random or manufactured, especially if you want to believe it. Probably the most pervasive bias we have encountered is confirmation bias. That’s a fancy term for the ancient wisdom, people hear what they want to hear.
In part, it is this psychology that enabled Bernie Madoff to pull off his scam. People wanted to believe Madoff’s exceptional results were the product of skill. When they saw a good number on their statements, instead of doubting, they were happy with him and with themselves for investing with him.
That’s conformation bias in action. You see it often. In a politically charged year, examples of the bias were everywhere in 2020.
Today’s media makes it challenging to stay focused on the long term and what really matters. By its nature, financial planning can help people reset and refocus on what is controllable and important to them. We can’t prevent every error, but we sure can help people try to act more like the casino and less like gamblers.
Dan Moisand, CFP, has been featured as one of America’s top independent financial planners by numerous magazines. He practices in Melbourne, Fla. You can reach him at [email protected].