Whenever we find ourselves forced to make decisions with limited information, we are most likely to use a decision rule or heuristic based on prior experience. This approach is easy to understand, when we think about the alternatives. If we did assume that “this time is different” from all of the others, how would we make a rational decision about how it works? If I were to tell you that I am showing you a magic trick and it works because there is something about it that is different from your day to day experience but I won’t tell you what it is, how could you possibly know what will happen next? The simple answer is that you can’t. Therefore, you are going to start with a set of assumptions based on how similar looking things behaved in the past and you will naturally go from there.
This gets us into trouble when it comes to investing money for at least 3 reasons. First, the variance of returns that we are likely to see is much greater than what we are accustomed to dealing with. Second, past performance is almost always the best predictor of future performance. Third, most measurements that we run into in nature follow a normal distribution. The fact of the matter is that each of these common expectations breaks down when it comes to stock picking.
Normal Returns Should be Around the Mean Right?
First, a portfolio of global equities that includes large-cap, mid-cap, and small-cap stocks as well as international stocks and emerging market companies will have an expected return of almost 12%. However, the standard deviation of such a portfolio can easily exceed 18%. We know that if returns are normally distributed, then the mean plus or minus 1 standard deviation will cover about 2/3 of the distribution. This means that the odds are close to 1 in 6 that you will see a result above or below this range. It’s like rolling a standard six-sided die. Any time you see a 1 you get a result below this range and any time you roll a 6 you see a result above this range. Neither outcome is particularly rare.
This implies that an annual return below negative 6% is pretty common. This should be clear because an observation of -6% is just the mean minus 1 standard deviation. This is not rare. It does not signal any special circumstance. To be completely honest it’s shouldn’t even be news. This is perfectly normal – but that’s not how it feels at the time. When you saw a return of -6.24% in 2018 you probably looked around for someone to blame – the Fed, the President, punishment for past sins, or whatever. Oh no – the sky is falling. A host of explanations are called upon to explain this dire condition – this retirement crisis – this first step to doomsday.
The next year the return was just as likely to be roughly 30%. In fact in 2019 the return was 28.88%. Woo hoo! The special problem must have been solved. But if we simply add 2 numbers we see that this is nothing other than a mean plus 1 standard deviation because 12 + 18 = 30%. Again, not news. No special explanation is needed. This is no more unusual than the -6% return from the year before. No special action or response is justified because this is what normally distributed variables do all the time.
The S&P 500 has shown average returns of 10% for decades. However, because the variance is so high, it has only produced annual returns between 8 and 12% 7 out of the last 98 years. In fact, its more likely to be above 30% or below -6% then it is to be in this range. But most of our experience with normal distributions is with settings where the variance is much smaller than the mean. This sets us up to over-react to stock returns, even though nothing unusual is going on.
Obviously, I Should Pick a Mutual Fund Based on the Track Record Right?
This myth has been debunked hundreds of times, yet it still drives Billions of dollars every year into the 5-Star funds. Morningstar rates mutual funds based on recent performance. The top 10% of funds over the past 10 years are extremely likely to be listed as 5-star funds for this year. Every year almost 80% of money that goes into mutual funds goes into 4 and 5 star funds. However, when we look at 5-star funds we find that less than half of them will beat the market in the next year. (This percentage is actually much lower if we account for fees and taxes.) Not only does past performance NOT suggest that this fund will be better than average next year. It is actually more likely to be in bottom section of the distribution of returns in the following year. This suggests that what we see is actually a reversion to the mean. This means that a fund that beats the market in year t is more likely to be below the average in year t + 1.
In fact, a simple heuristic of only picking 1-star rated funds will consistently beat the strategy of picking only 5 star funds. Where else does this happen in life? Its like saying that the basketball player most likely to go scoreless in the next game is Kevin Durant, or Steph Curry. We know from experience that that is almost never going to happen. The player’s past performance is by far the best predictor of what he will do in the next game. If you have no weather report, your best bet is almost always to assume that tomorrow will look pretty much like today. This approach works in thousands of different settings. It just stinks when it comes to stock picking.
Past Winners are Truly Past
Past performance of stock pickers is not a useful indicator of future performance. But the larger issue is even more surprising than this. YouTube includes an interesting short video of Warren Buffett speaking and at a Berkshire Hathaway annual meeting (https://www.youtube.com/watch?v=z9a_w8GPnB4) In this clip Buffett shows 2 lists of companies. One is the largest companies in the world in 1989. Let’s call that List A. The second is the largest companies in the world in 2021. Let’s call that List B. I will not repeat the lists here because they are easy to see with the link above. If you look at List A three things will jump out at you. One, 13 of the companies listed were based in Japan, and 6 were based in the US. In fact, the top 4 were all Japanese banks. The largest company was Industrial Bank of Japan and was worth roughly $104 Billion and the 20th largest firm was worth $30 Billion. If we fast forward about 30 years to 2021 and look at list B, we see that 13 of the firms are based in the US and none are based in Japan. Several of the firms on List B didn’t even exist in 1989 including Facebook Alphabet, and Amazon. In addition, we see that Apple, as the largest firm on the list has a market value of over $2.05 Trillion, and the smallest firm, United Health is worth roughly $305 Billion. Furthermore, the value of the US-based firms on the list adds up to $11.3 Trillion. This would be over twice the size of the entire GDP of Japan in 2021 (4.94 Trillion). The old list was not a great predictor of future success either at the firm or country level.
An Average Company will Give the Average Return – Right?
If market returns look normally distributed, it would seen to follow that returns on individual stocks would look that way as well. But there is a key piece of information left out of this line of thinking. A stock can only fall by 100% over any period of time. However, it can go up by thousands of percent over the same period. As a result, the median return has to be lower than the market average. In fact a more rigorous analysis of stock returns includes a clear case of “fat tails.” There are outliers that dominate the total. The S&P 500 has yielded a roughly 10% return over the past 8 decades. However, if you remove the top 35 stocks from the sample, your return is close to 0. The outliers are really all that matters. Of course, everyone knows who the outliers are after the fact, but no one knows who they will be before it happens.
The investment firm Correlation Ventures looked at the returns on venture capital investments for over 21000 ventures from 2004 to 2014. They found that 65% lost money; 2.5% generated profits between 10 and 20 times the investment; 1% generated returns of more than 20 times the investment; and 0.5% had returns of at least 50 times the investment or more. This is where the vast majority of the industry’s profit comes from. The outliers are about all that matters.
If you are like me, you are probably thinking that this is not a big surprise. Most small businesses fail, and most small companies that do survive never generate large profits. Surely, I am much better off if I focus on the larger firms. Again, you would be wrong. The distribution of returns on these larger firms looks remarkably similar to the returns on the very small ones in the sense that a small number of outliers explains the vast majority of the total gain.
J.P. Morgan Asset Management once published the distribution of returns for the Russell 3000 Index. This is the index of the 3000 largest firm in the US. Looking at returns since 1980 they found that 40% of these stocks lost at least 70% of their value and never recovered. Virtually all of the index’s overall gain came from 7% of the firms on the list because they did better than the average by at least 2 standard deviations. Again, the outliers are all that matters. (See The Smartest Portfolio You’ll Ever Own, by Daniel Solin.)
In 2020 Warren Buffett’s net worth was roughly $85 Billion. Roughly 43% of that amount came from one company. Were it not for Apple, Mr. Buffett would have to survive on only $44 Billion. I think he would still be able to pay his rent, but that doesn’t change the fact that the outliers drive everything. How big are the outliers? Since 1980 the Russell 3000 has increased by a factor of 73. Roughly 40% of the firms that were in the index in 1980 are now bankrupt. The top 7% did so well that they overwhelmed these outcomes and produced a success level almost beyond comprehension. Again, the outliers drive everything.
This produces a skewed distribution of outcomes. At any time when we look at stock returns, the firms that went out of business drop out of the discussion. The returns of the surviving firms appear to be normally distributed, but this is due to the fact that we are looking at a biased sample. If the firms that drop out of the discussion are included, we get a very different picture. The end result is that the average firm actually under-performs the market average. This is yet another reason that buying the entire market in the form of an index fund is the way to go. When a firm drops out of the S&P 500, it is replaced with a new one, but as the index holder you do nothing. The problem takes care of itself.
Before, you start to think that this only applies to the smaller firms – check this out. Over the past 100 years or so America has seen about 3000 car companies. Of this number, by the summer of 2008, 3 of them survived, and technically 2 of those three were bankrupt. In other words 99.9% of US based car companies went out of business, and 2 of the 3 survivors are only here because the government bailed them out. The outliers drive everything and you will not know who they are until most of the money invested in the sector is lost.
Why You Will Try to Pick Stocks Anyway
When it comes to stock picking, all you are looking for is the outliers. Not only do you not know who the outliers are gong to be 30 years from now, it’s a pretty safe bet that these outliers are not among the most successful firms in existence now, and many of them don’t even exist. This is a strong argument for why you shouldn’t waste your time trying to pick them. You are much better off simply buying the entire list. The outliers will carry you to riches even though you have no idea who they will be, or even what country they will be based in.
So why are you going to try it anyway? Because you can be wrong dozens of times. As long as you get the big one right, the rest won’t matter. When we look at the level of individual stocks the variance of returns is so high, it becomes easy to find a litany of anecdotes that make stock-picking look like a good idea, especially since almost everyone who gets it wrong doesn’t want to talk about it. It reminds of a line from High Rollers by Ice T – “you’ll never get caught cause you got nerves of ice, and you’re much smarter than the crooks on Miami Vice — Right You want to be a high roller!!!”
The premise of this book is that doing well with money has a little to do with how smart you are and a lot to do with how you behave, and behavior is hard to teach, even to really smart people. The research that suggests that stock picking won’t work for you is overwhelming, but human nature will lead you to try it anyway. For those folks (and that’s most of you) who will insist on traveling this path have a nice day. We are done here. For the other folks, let’s talk about how to do the best we can in the real world.
