Gardner’s Alpha

Introduction

In 2013, a group from AQR wrote a paper entitled “Buffet’s Alpha”, which confirmed that Warren Buffet’s superior stock returns can be explained, in retrospect, solely by his exposure to the value, quality, and low-volatility factors (i.e., buying “a wonderful business at a fair price”) and the use of the 1.6-times leverage that results from investing his insurance float. This was not a surprise. Buffett, himself, had explained his approach repeatedly in his letters to shareholders. However, it was useful to get confirmation from the data that his outperformance indeed came from this and not something else.

While the outperformance of many great investors, including the greatest of all time, Warren Buffett, can now be understood via factors, there are other great investors, whose outperformance remains “unexplained”. Top on my list of unexplained outperformance is that of David Gardner, one of the co-founders of the Motley Fool. While his outperformance may not be a match for the all-time greats, the success of his strategy is perhaps the most mysterious, leaning on unusual variables and sometimes betting directly against the familiar ones, most notably by buying only “overvalued” stocks.

Gardner’s Performance

The Motley Fool’s services focus on investors who are early in the wealth accumulation phase, adding to their portfolio on a regular basis. While they report the performance of the individual recommendations going back decades, it would take some work to translate that into a fixed portfolio that would accurately reflect how well they had achieved their investment goals.

Nonetheless, there is little question in my mind that Gardner has significantly outperformed the market since the mid-90s. Indeed, it would be very difficult not to do so when you have invested in at least three 100-baggers and at least five 50-baggers. (I say “at least” because I’m pretty sure I missed some when I was trying to collect the lists of these.)

Indeed, Peter Lynch’s famous “10-baggers” are so common amongst his stock picks that they don’t even track or celebrate them. Instead, they celebrate the first “spiffy pop”, which is when the stock you own rises more in one day that you paid for the shares initially. When you have to invent new terms for the absurd returns on your investments, you’re doing pretty, pretty good.

Gardner started recommending growth stocks in the mid-90s, just as the Internet went mainstream, and he retired from stock picking last year, just before the recent destruction in growth stocks. What a legend.

If you are wondering why someone with this track record, who has a “never sell” philosophy, is not amongst the richest people in America, the reasons, I think, are that he started from a smaller base (investing only his own money) and that he is much more diversified than a typical portfolio manager.

Whereas Buffett would hold as few as 3–4 stocks and a typical portfolio might be considered diversified at 10–20 stocks, Gardner’s strategy uses many more positions. He advocates for having your number of positions be between one and two times your age. Hence, someone of David’s age would have 50–100 positions. A 100-bagger return on 1/50-th of your portfolio nicely triples your wealth, but it doesn’t put you on the cover of Time magazine.

The latter difference is reflected in my Horseshoe Theory of investing styles. Buffett is the quintessential “quality” investor, in my parlance, which means he focuses on highly predictable companies. That predictability allows for extreme concentration. David Gardner, on the other hand, is in the “hyper-growth” category, again in my parlance. Owning companies whose future fundamentals are extremely unpredictable requires far more diversification.

I should note that David also has a brother, Tom (the other Motley Fool co-founder). For me, Tom’s approach fits nicely within the quality camp, which means his outperformance is something that I think we understand. It is David’s hyper-growth strategy that remains unexplained.

Real Option Theory

With that mystery in mind, I was struck by this discussion with Michael Mauboussin on the Long View podcast (starting at 37:15) where he mentioned his 1999 paper on valuing companies using “real options”.

The idea is to use option theory to try to value the possibility that a business could find new opportunities, e.g., Amazon’s creation of AWS. While it may seem as though the discount rate of a DCF already includes the possibility of upside as well as downside, the article points out that there are critical differences. Most importantly, increased uncertainty nearly always decreases the DCF value of a business: the increased downside resulting from more uncertainty almost always outweighs any added upside. In contrast, increased uncertainty always increases the value of an option.

The result is that an investor who is trying to invest in real options would want to own companies with more uncertainty, whereas a quality investor like Buffett wants to own companies with less uncertainty. This contrast is one of the inspirations for my aforementioned Horseshoe Theory of investing styles (with the other being Buffett’s observations that the distinction between value and growth is unhelpful for quality investors).

While I was familiar with the paper, and indeed inspired by it previously, it had not hit me until listening to the podcast how directly this theory relates to David Gardner’s investing style. The table below lays out, on the left, the four criteria that Mauboussin gives for identifying companies with the most real option value and, on the right, the key properties that David Gardner looks for in the stocks that he picks:

Mauboussin noted, in his 1999 paper, that market leaders usually have the most option value as they typically get the “first call” on new opportunities. For example, they attract the best employees, who come up with more new ideas, and they often receive the first pitch about potential acquisitions.

Gardner’s stock picks, which he dubs “Rule Breakers”, focus heavily on market leaders, with three of his six criteria pointing exactly at this property. He wants companies that are the clear “top dogs” in their industries and have already shown strong market returns — nearly a direct translation of “market leaders”. He also wants them to have a sustainable competitive advantage to give us confidence that they will remain the top dog.

Gardner’s first and most important rule, though, is not to buy “top dogs” but to buy “top dogs in an important, emerging industry”. I would argue that important, emerging industries are those that most strongly correlate with “uncertainty in the sector”. Industries that are just hitting their mainstream adoption have high growth rates that make predicting future fundamentals difficult, and those that are also clearly important are likely to draw competition, which further clouds the picture. All of this is negative for a quality investor, but as Mauboussin notes, ideal for seeking option value.

The table above shows four of the six traits of a Rule Breaker, including the most important trait, but there are two other traits worth, both worth noting.

The fifth trait of a Rule Breaker is “strong consumer appeal”. While Gardner states that this one is not a strict requirement, it is nice to have as it indicates a potentially large runway for growth.

Mauboussin gets at the same idea when he explains why you want to focus on “market leaders”. In addition to getting first call on new opportunities, market leaders often take advantage of economies of scale and scope. I would argue that strong consumer appeal is a way to get that scale and scope. The idea of finding “riches in niches” is a great strategy for quality investing, but those seeking real option value want products with broad appeal.

Valuation

The final trait of a Rule Breaker is that it be “widely perceived as overvalued”. Here, we do see a split between Mauboussin and Gardner, with Mauboussin (in the podcast) wanting to buy real option value at the lowest possible price and Gardner leaning into a perception of the stock being too expensive.

Part of this, I think, comes from Gardner’s intention to not focus on valuation. Buffett fans will remember the story of how he almost did not buy See’s Candy because their asking price appeared too high to him at the time. Luckily, the owners relented and accepted Buffett’s price. In a Sliding Doors-type moment, Gardner had a formative experience with the opposite outcome: he refused to buy Yahoo stock because it did not get all the way down to his valuation, only to watch the company massively outperform afterward. Gardner responded intelligently by dropping that valuation filter from his process.

In other interviews, Gardner has said that he generally thinks the market correctly incorporates the next 3–5 years of growth into the price. From my own experiments, I think the best estimate is probably 5–7 years; however, I agree with the basic idea. Commentators will sometimes say that the market is expecting a company to have double-digit growth “forever”, but that can’t possibly be true, as the proper price of such a company would be infinity. If the price is finite, then the market is assuming a finite window of above-average growth, and I think an assumption of around 5 years is likely typical.

Let’s put these pieces together. If we have a market leader with a sustainable competitive advantage, then the core of the existing business is of high quality, making it fair, in my view, to predict five years forward. In that case, if the company is trading with a P/E minus growth under 20, say, then I think can argue that the existing business is trading in the neighborhood of fair price. That means that its real option value is free, and as Mauboussin explained, market leaders are the most likely companies to have option value, so the stock is likely underpriced.

That provides an explanation for the outperformance of his strategy when the existing business is of high quality, but Gardner often invests in even less predictable companies, many of them not even yet profitable (e.g., Tesla). The simple analysis above cannot handle those cases.

However, I think we can still understand his outperformance via real options. Unprofitable companies have no value in their existing business, so the market must be pricing some amount of option value. However, such companies are even more uncertain, which means their option value is even higher, increasing the chances that is not fully appreciated by the market.

Conclusion

While this simple analysis does not give us mathematical proof that the theory of real options fully explains Gardner’s outperformance, it is by far the closest I have gotten to feeling that I really understand why the strategy has worked. I think this formalism does capture the essence of his strategy, and I am grateful to Michael Mauboussin for teaching it to us.