The U.S. stock market, as represented by key indexes, has produced solid returns for investors over the long run.  This fact has led many investors to conclude that 1) most stocks go up in the long run, and 2) buy-and-hold is a sure-fire strategy. 

I would like to argue that, while the initial observation of strong index performance is correct, those secondary conclusions are misleading and have serious implications for successful investing.

---

First, let us review some statistical characteristics of the U.S. stock market:

Stock concentration (in dollar terms):  Professor Hendrik Bessembinder of Arizona State University has published at least three research papers — one in 2018 (link), one in 2019 (link), and one in 2023 (link) — studying the patterns of stock returns.  His findings, in my words, can be summarized into the following two points: 

• Over the long term, most stocks generated little returns:  In rough terms and rounding up the numbers, over the past century, among all publicly listed U.S. stocks, nearly 60% generated lifetime returns that were lower than that of one-month Treasury bills.   The top 4% of stocks made up the entirety of the U.S. stock market gains — which means, in the U.S., the top 4% of stocks generated all the gains and the remaining 96% of stocks were a wash.  To make the point even more obvious, here is another finding by Prof. Bessembinder:  If stocks are grouped together by their lifetime returns (rounded up to the nearest 5%), the single most frequent outcome observed during the century, by far, is a complete capital loss of 100% — i.e., if an investor randomly picks a stock and buys and holds that stock, the most likely outcome is zero.  Zero!
• Over the long term, a few stocks contributed all the gains:  During the past century, the top three stocks — Apple, Microsoft, and Exxon Mobil — accounted for over 10% of the total wealth creation in the U.S. stock market; the top 11 stocks, 20%; the top 72 stocks, 50%; and, the entire wealth creation of the U.S. stock market during that century can be explained by fewer than 1,000 stocks, which is less than 4% of all U.S. stocks during that time period.  Outside of the U.S., a similar dynamic exists but is more extreme:  From 1990 to 2018, for non-U.S. stocks, the entire net wealth creation can be explained by less than 1% of stocks.

Stock concentration (in percentage terms):  Stock returns are concentrated not only in dollar terms but also in percentage terms.  In a blog piece I wrote in 2020 (link), I showed that across the U.S. and Chinese stock markets, over the past 40 years, there are only a few hundred stocks that generated more than 10x returns for investors in any given decade.  In terms of stocks that returned 100x, there are only a few dozen in the U.S. and a limited few in China.  So, not only most stocks cannot generate meaningful returns in dollar terms, but also most stocks cannot generate meaningful returns in percentage terms either.

Time concentration:  Stock markets demonstrate some peculiar patterns in that most returns seem to be concentrated in just a few days.  Consider the following:

• For the past 35 years from 1988 to 2022, the S&P 500 has delivered an annualized return of 10.4% (turning $1 into $32).  If the best 20 trading days are removed from the data, the annualized return would be reduced from 10.4% to 6.3% (turning $1 into $8.6).  If you miss those 20 days, your “would-have-earned” $32 shrinks to become $8.6.  That is huge!  Similar dynamics exist in other stock markets and other stock indexes as well.  In essence, long-term investment results are concentrated in a few days.
• A 2011 paper (link) demonstrates that from 1958 to 2009, over 60% of the cumulative annual excess return (i.e., return over the risk-free rate) in the U.S. stock market was earned on announcement days — days with pre-scheduled macroeconomic announcements — which is just 13% of trading days!
• A 2013 paper (link) shows that in the U.S., from 1980 to 2011, about 50% of the “realized excess stock market returns” were earned during the one day leading up to scheduled FOMC announcements. 

Volatility at single-stock level:  Despite the fact that most stocks are going nowhere in the long term, most stocks are going everywhere in the short term.  Take these two examples: 

• “80%”:  In 2004, Bryan Lawrence had a conversation with Warren Buffett and Buffett said that “the average stock goes up and down by 80% in a year.”  Lawrence couldn’t believe that number and went back and did the math for about 4,000 U.S. companies going back 20 years, only to realize that Buffett was right:  An average stock goes up and down by 80% in a year.  Lawrence has since been doing this calculation about once a year, and the resulting volatility number has remained largely unchanged.
• Increasingly volatile:  A 2000 paper (link) illustrates that while overall market volatility displayed no significant trend from 1926 to 1997, stock-level volatility more than doubled in recent decades.  In short, stocks are becoming more volatile while the overall stock market is not.

---

With those observations in mind and thinking about the behavior of the stock market, it is sensible to make this following statement: 

It is a few stocks and a few days that contribute most stock market returns over the long run; as for the rest of the stocks and the rest of the time, stocks just move up and down violently and in aggregate, produce little returns.   

---

Profound but counterintuitive.  It means a lot of different things, I think. 

It helps explain why almost no active investment managers can beat U.S. stock indexes.  While trying to jump from one set of stocks to another, switching between stocks and cash, managers think they are looking smart without realizing it is exactly what they are doing that is causing them to miss “super stocks” and miss “key days” — and the unfortunate yet inevitable fate of underperformance haunts these managers.  Remember, indexes do not skip “super stocks” nor skip “key days” — that is where the difference lies.

It also sheds light on why conventional buy-and-hold strategies have failed miserably in delivering satisfactory investment results.  Conventional wisdom guides us to build a “well-diversified” portfolio by buying and holding 15 to 30 uncorrelated stocks.  Given that most stocks produce no returns, how can investors come to believe this kind of conventional buy-and-hold portfolios to easily work out in their favor?  Owning 15 to 30 stocks, expecting most of these stocks to generate strong results, and therefore the portfolio to do well as a whole, is statistically unrealistic.  A more realistic view is this:  The more stocks an investor owns, the less time he or she will have in studying each stock, the lower the accuracy in identifying the right stocks.  That is exactly what is happening to almost all conventional buy-and-hold investors.  For most investors (barring a few unusually talented ones), the more you own, the less you know, the worse your investment results will be.  In this regard, it is sensible to argue that the conventional approach of buy-and-hold a “well-diversified” portfolio just doesn’t work for most investors.  Instead, I should argue and I think it is apparent to readers at this point, buy-and-hold is an abnormal strategy that works for abnormal stocks.

The volatility question is also interesting. Rising volatility at the single-stock level is a double-edged sword.  On the one hand, it makes concentrated investing less palatable for its increasingly volatile performance as underlying stocks are becoming more volatile.  On the other hand, it makes concentrated investing more attractive because stocks are swinging so hard that dedicated managers will have a better chance to purchase “super stocks” at advantageous prices. Therefore, in an environment where stock prices are becoming more and more volatile, concentrated investing, if done with skill and care, enjoys a better and better chance to succeed.

---

If the above is not ironic enough for you, the ultimate irony is this:  In institutional investing, most allocators want to “check the box” so they push managers to build “well-diversified” portfolios.  For managers who are commercially minded, they will answer to those requests.  Conventional wisdom does not work in investing.  By building “well-diversified” portfolios, underperformance is almost guaranteed.  So-called institutional best practices only reinforce this unfortunate path to mediocrity. 

Perhaps, it is unacceptable for managers to acknowledge to their clients that the managers themselves do not truly understand most of the 30 stocks that they own.  If they did, they would be fired.  Yet, if they do not own that many stocks, they will not have a “well-diversified” portfolio to show their clients, and those managers would be fired too.  In either case, they are likely to be fired.  You pay a price for being intellectually honest! 

Perhaps, the reason behind all this is that most investors and allocators are getting confused by themselves.  They are getting paid for thinking they have great ideas and for having ideas that convince other people, not for actually having great ideas.

See the difference!

---

Therefore, I increasingly think that a sensible investment approach is a “barbell” one:  Either go all in on index investing or go all in on concentrated investing.  Any approach in the middle has been statistically proven to not work for almost anyone.  To keep doing things that do not work is madness. 

For managers who genuinely care about investment performance, they must embrace an investment philosophy that reflects investment reality, not sheer fantasy or commercial interests.  They must do things that add real value, not that make stories that they think can convince other people at the next investor meeting.

(END)