Buffett’s bad advice

Warren Buffett is a national treasure. In today’s world of Wall Street charlatans and Silicon Valley megalomaniacs, he is a fixed point of rationality and decency. We should amend the Constitution to require mandatory lectures for U.S. Senators where Buffett explains repurchases and crypto.

However, he’s not perfect. Like all of us, Buffett was born into this world naked and ignorant of the benefits of portfolio diversification. Since his birth in 1930, he’s acquired clothes, but sadly he’s failed to acquire an appreciation for diversification. While he has a keen eye for bargains in the stock market, he remains tragically blind to the greatest free lunch in finance.

Buffett has long condemned portfolio diversification and advocated concentrated portfolios. Let’s consider his statement at the Berkshire Hathaway 1996 annual meeting:[1]

You know, we think diversification is — as practiced generally — makes very little sense for anyone that knows what they’re doing.

… Diversification is a protection against ignorance.

…. I mean, if you look at how the fortunes were built in this country, they weren’t built out of a portfolio of 50 companies. They were built by someone who identified a wonderful business. Coca-Cola’s a great example. A lot of fortunes have been built on that.

… there is less risk in owning three easy-to-identify, wonderful businesses than there is in owning 50 well-known, big businesses.

 … If you find three wonderful businesses in your life, you’ll get very rich. And if you understand them — bad things aren’t going to happen to those three.

It is more in sadness than in anger that I must inform you that this advice is just wrong. Really wrong. Spectacularly wrong.

Portfolio concentration is a dumb idea. Diversification is a good idea. Buffett’s not the first smart person to get it wrong, as shown in these words from Keynes in 1934 (quoted by Chambers, Dimson, and Foo (2014)):

As time goes on, I get more and more convinced that the right method in investment is to put on fairly large sums into enterprises which one thinks one knows something about and in the management of which one thoroughly believes … It is a mistake to think that one limits one’s risk by spreading too much between enterprises about which one knows little and has no reason for special confidence.

Keynes, Keynes, Keynes. Dude, you know I love you, but you’re the one who’s making a mistake. I forgive you since you were writing decades before the work of Markowitz, Samuelson, and Sharpe.

There’s one thing that Buffett gets right: Diversification is a protection against ignorance. That’s true, and that’s why we need it. We humans are all unavoidably ignorant about the future. Many bad things will someday happen, and they are inherently unpredictable.

Here’s Buffett denying this inherent feature of the human condition: And if you understand them — bad things aren’t going to happen to those three. Nonsense. No human has the ability to identify firms that bad things “aren’t going to happen to.” Buffett seems to say that when you “understand” a stock, you attain mystical knowledge of things to come.

Buffett’s claim reminds me of the joke by American humorist Will Rogers:

buy some good stock and hold it till it goes up, then sell it. If it don't go up, don't buy it.

This joke is funny because NO ONE KNOWS FOR CERTAIN WHICH STOCK WILL GO UP. I’m sorry to use all caps, but desperate times call for desperate measures.

Buffett has done many wonderful things that have enriched Berkshire shareholders, but his failure to diversify was an unforced error. Let’s take his decision to concentrate Berkshire’s portfolio into Coca-Cola as of 1996, a position which he has maintained to this day.[2] Coca-Cola has massively underperformed the market since 1996. So why did Buffett buy it? Because Buffett is a human being, and Coke’s underperformance was inherently uncertain. It is folly to believe that Buffett or anyone else has magical powers to identify three firms that will outperform with certainty going forward.

I’m not saying that Buffett should abandon everything and put his money into index funds. But he could change his ways and make his stellar performance even better. He could take the insights embodied in his actual concentrated portfolio (e.g., Coca-Cola) and invested in a diversified portfolio of similar firms (e.g., firms similar to Coca-Cola). This diversified approach would produce higher returns with lower volatility, according to Frazzini, Kabiller, and Pedersen (2018), results which I’ll explain later.

The debate about diversification vs. concentration raises many questions, including whether portfolio managers should choose concentrated portfolios if their clients already have other assets in diversified portfolios. But here, I want to focus on a simpler question: If you were putting your entire personal wealth into a portfolio, is it better to hold three companies or 50 companies? My view is that holding three companies is a great approach if you want to retire in poverty; holding 50 companies is what I prefer.

Why do people hold concentrated portfolios?

Historically, individual investors have been woefully underdiversified. For a discussion, see Barber and Odean (2013) and Beshears, Choi, Laibson, and Madrian (2018). Contributing factors include:

• Overconfidence
• The illusion of control
• Familiarity bias
• Survivorship bias

Overconfidence

I’ve previously mentioned overconfidence leading to gambling. Adam Smith called it the “overweening conceit which the greater part of men have of their own abilities.” Buffett’s statement that “bad things aren’t going to happen to those three” is an example of overconfidence.

Goetzmann and Kumar (2008) study portfolio diversification by U.S. retail investors and find pervasive underdiversification which they attribute partly to overconfidence. They find that low diversification is associated with low returns, and say:

The systematic under-performance of less diversified investors is somewhat puzzling. Why do those investors systematically accept lower returns?

… most investors could have improved the performance of their portfolios by simply investing in one of the many available passive index funds.

The illusion of control

Buffett’s claim, that bad things won’t happen if you understand the stock, reflects the illusion of control. Because you’ve carefully selected these three stocks, you think that you control the outcome. It’s an illusion. You can only control what stocks you buy; a cruel and unforgiving world controls what happens next.

Let’s consider the old saying about diversification: don’t put all your eggs in one basket. Mark Twain gave his own twist:

 Put all your eggs in the one basket and - WATCH THAT BASKET.

Twain’s advice might be reasonable in a situation where you are guarding the basket from minor threats, such as racoons or small children. But for larger threats, that strategy won’t work. It doesn’t matter how intently you watch the basket; if a tornado hits your house, the eggs won’t survive.

The best way to protect your eggs from tornadoes is not to watch one basket intently; that’s an illusion. You should put your eggs in different baskets in different houses.

Stock market threats are tornadoes, not raccoons. They cannot be perfectly predicted or controlled, only mitigated with diversification. Don’t believe the illusion that you control the tornado; you don’t.

Now, in many instances, Buffett actually does have control in a way that others would not. It is no illusion that Buffett can provide financing to companies or otherwise improve their prospects. But that’s a different kind of control, one that reflects not knowledge of future events, but the ability to influence them.

Familiarity bias

Familiarity bias is the tendency of investors to buy companies they know. Here’s Huberman (2001):

People simply prefer to invest in the familiar. People root for the home team, and feel comfortable investing their money in a business that is visible to them.

One version of familiarity bias is home bias, the tendency of investors to overweight their home country and underweight foreign countries. Home bias is pernicious because it generates not just under-diversification, but perverse anti-hedging. That is, in theory Japanese investors should underweight Japanese stocks and overweight U.S. stocks, because Japanese stocks are more correlated with other sources of risks that they face (real estate, employment, etc.) Instead, Japanese investors anti-hedge by overweighting Japan. Mistake.

The most extreme form of anti-hedging is owning your employer’s stock. Here’s Barber and Odean (2013):

Investors who overinvest in the stock of their employer (company stock) are left exposed to the fortunes of their employer (idiosyncratic risk). Famously, Enron employees had 62% of their retirement plan assets invested in company stock at the end of 2000. By December 2001, the company had declared bankruptcy and its employees had lost both their jobs and a large fraction of their retirement income.

Thankfully, this particular self-destructive behavior has greatly diminished in recent years, due to the rise of target-date funds and other smarter options for 401K accounts.

Warren Buffett is a walking, talking embodiment of familiarity bias. His daily diet includes milkshakes, Coke (five cans a day), and Egg McMuffins. So Berkshire bought shares in Dairy Queen, Coca-Cola, and McDonalds.

Survivorship bias

Survivorship bias is when a particular sample excludes some relevant outcomes. For example, suppose mutual fund company XYZ advertises that 100% of its existing funds have outperformed the market. Should we conclude that XYZ funds invariably outperform? No, because XYZ has not reported the performance of its funds that no longer exist (they were closed due to bad performance).

If we only see successful concentrated portfolios, we’d wrongly conclude that concentration is good. Consider this proposal from poet Alexander Pope in 1720, from Chancellor (2000):

I daily hear such reports of advantages to be gaind by one project or other in the stocks, that my spirit is up with double zeal, in the desire of our trying to enrich ourselves. . . . Let but Fortune favor us, & the World will be sure to admire our Prudence. If we fail, lets een keep the mishap to ourselves.

When you keep mishaps to yourself, you create what Hirshleifer (2020) calls “social transmission bias,” where only successful investors brag about their concentrated portfolios, leading to widespread under-diversification and the myth that concentrated portfolios usually have good outcomes.

Let’s consider Buffett’s statement:

I mean, if you look at how the fortunes were built in this country, they weren’t built out of a portfolio of 50 companies. They were built by someone who identified a wonderful business. Coca-Cola’s a great example.

This is Buffett engaging in social transmission bias. Only good stories get spread. Suppose I rewrite as follows:

I mean, if you look at how the fortunes were built in this country, they weren’t built out of a portfolio of 50 companies. They were built by someone who bought a Powerball ticket. Edwin Castro of Altadena, California, who won $2.04B in February 2022, is a great example.

Castro became a billionaire with a concentrated portfolio consisting of a single Powerball ticket costing $2. Should we conclude that buying Powerball tickets is a good investment? Nonsense.

Here’s Odean (1999) describing how investors may learn the wrong lesson from their own success:

Survivorship bias may also favor overconfidence. Traders who have been successful in the past may overestimate the degree to which they were responsible for their own successes—as people do in general (Ellen J. Langer and Jane Roth, 1975; Dale T. Miller and Michael Ross, 1975)—and grow increasingly overconfident.

I’m not saying that Buffett’s success is 100% luck. But neither is it 0% luck.

Buffett’s failure to diversify

Systematic evidence on Buffett’s underdiversification comes from Frazzini, Kabiller, and Pedersen (2018). The good news is that historically, Buffett has picked great stocks, with a fantastically impressive annual alpha of 5.4% over many decades.

The authors say:

Buffett’s returns appear to be neither luck nor magic, but, rather, reward for leveraging cheap, safe, quality stocks … explaining Buffett’s performance with the benefit of hindsight does not diminish his outstanding accomplishment. He decided to invest based on these principles half a century ago. He found a way to apply leverage. Finally, he managed to stick to his principles and continue operating at high risk even after experiencing some ups and downs that have caused many other investors to rethink and retreat from their original strategies.

Let me comment briefly on leverage. Buffett bought low volatility stocks and levered them up (about 1.7-to-1, according to the authors) using Berkshire’s unique corporate structure. Smart move. Leverage is an efficient way to benefit from undervalued safe stocks; portfolio concentration is not.

While Buffett did amazingly well, he could have done better if he’d diversified.

The authors study Buffett’s portfolio of public stock holdings from 1980 to 2017. Using the actual stocks that Buffett owned, they construct portfolios that are similar along the dimensions of profitability, valuation, and risk. Their portfolios involve short selling and leverage and are constructed with the benefit of hindsight; they are not meant to be realistic alternatives. These portfolios are a sort of robot Buffett who tries to imitate the real Buffett. Robo-Buffett, unlike the real Buffett, holds diversified portfolios.

Does robo-Buffett beat real Buffett? Yes. Robo-Buffett outperforms real Buffett by 7% per year, if you allow robo-Buffett to use shorting and leverage. If robo-Buffett is forced to be 100% long, it still beats the market, but it does not beat the real Buffett.

Why does the unconstrained robot beat the human? The authors find:

… This result may have arisen because our systematic portfolios have factor tilts similar to Buffett’s but hold a much larger number of securities, thus benefiting from diversification.

Catastrophic concentration

Buffett, a resident of Omaha, Nebraska, became a billionaire using concentration. Other residents of Omaha weren’t so lucky.

For various historical reasons including familiarity/home bias, as of 1999 many Omaha residents were invested in just three stocks: Level 3, InaCom, and Enron. They were following Buffett’s advice:[3]

Interest by Omaha residents in local stocks is understandable, given that one widely held local stock helped build the city's wealth: holding company Berkshire Hathaway Inc., run by Warren Buffett. "There was a time when you could probably attribute 75% to 85% of Berkshire ownership to three or four ZIP Codes in Omaha," says George Morgan, a stockbroker with Kirkpatrick Pettis, an Omaha-based brokerage firm.

Many of Berkshire's early investors became centimillionaires, says Mr. Morgan, so "it was difficult to tell people 'You shouldn't own [Level 3] ... you should diversify.' "

When you hold only three stocks, you face the possibility that all three simultaneously collapse in value. That’s exactly what happened:

Between 1999 and early 2002, the shares of a local computer dealer, a broadband-service provider and finally, Enron Corp. -- headquartered here until 1986 and still home to 750 retirees and several hundred employees -- all plummeted in value, wiping away a big chunk of many residents' stock-market gains. … If an Omahan had bought $100 worth of each of the three companies' stocks at the beginning of 1999 and held on to them, the portfolio would be worth less than $9 today.

… the signs of distress caused by the decline of the three stocks are easy to spot. Driving through a cornfield that has been turned into million-dollar homes, Patrick McNeil, president of McNeil & Co., a custom-home builder, points out several "Level 3 squeezes." These are homes either sold at a loss or abandoned midconstruction after Level 3 stock began dropping from its $132 peak in March 2000. It now trades for below $4.

Lesson: holding just three stocks is a terrible idea. Holding the same three stocks that your neighbors also hold is even worse.

In some cases, the very act of building a concentrated portfolio can contribute to poor performance. When you build a concentrated portfolio, when you buy XYZ, you will raise XYZ price, hurting yourself on the way up (because you are buying at a higher price). And when you try selling later, you may drive down XYZ price, hurting yourself on the way down (because you are selling at a lower price). In contrast, in diversified portfolios, you can spread your trading across multiple stocks to minimize market impact.

I’ve previously defined the Iron Law of Return-Chasing Flows: money chases trailing returns. Combine that with concentrated portfolios, and you have a toxic combination. Here’s van der Beck, Bouchaud, and Villamaina (2024):

Many active funds hold concentrated portfolios. Flow-driven trading causes price pressure, which pushes up the funds’ existing positions resulting in realized returns. … The combination of price impact and return chasing causes an endogenous feedback loop and a reallocation of wealth to early fund investors, which unravels once the price pressure reverts.

This process may describe the experience of Cathie Wood’s ARK Innovation Fund, according to Jason Zweig of The Wall Street Journal:[4]

As of October 2019, ARK Innovation had nearly one-quarter of its $1.6 billion in assets in nine stocks. It owned more than 5% of the total shares at each of them.

… As ARK Innovation’s returns flared up, investors pumped money in. In 2020, its assets ballooned more than ninefold from $1.9 billion to $17.7 billion.

… Sooner or later, though, performance falters, the hot money flees and managers have to dump stocks. That pushes down the prices of those stocks, worsening the fund’s performance and sending even more hot money stampeding out.

… ARK Innovation’s assets peaked at about $25.5 billion in mid-2021 and have shriveled to $6.3 billion. The fund has lost an average of 27.9% annually over the past three years.

I’ve previously discussed Shiller’s concept of “naturally occurring Ponzi schemes,” of which self-inflated returns are one manifestation. You don’t need concentration to get these feedback loops, but it sure helps. Here we see one of the many ways that Buffett has an advantage over Wood. Berkshire is not an ETF and thus it never needs to sell due to outflows.

An extreme example of self-inflated returns is Bill Hwang’s Archegos. Hwang followed Buffett’s recipe: deep fundamental research and concentrated portfolios. He also included some new ingredients: massive leverage and lying to his brokers about his positions. Hwang was convicted of multiple fraud counts in July 2024 and is currently awaiting sentencing.

Just say no to portfolio concentration, kids.