One of the great privileges of our work is that it demands a beginner’s mindset and a commitment to continuous learning, no matter how experienced we may be. To that end, I find myself increasingly focused on better understanding the ongoing collision between the worlds of investing and gambling. The line between the two has always been hazy. Both involve probabilities, a range of outcomes, and, at times, significant emotional and financial consequences.
Yet those similarities quickly lose their relevance once we consider a critical distinction: even a mediocre investment strategy should generate profits over time, while gambling carries a negative expected return. Casual gamblers who believe they will make money are ignoring the massive and consistent profits generated by casinos, sports-betting platforms, and related businesses. Where, exactly, do they think that money comes from? Of course, we all know the outlier stories: someone who won the lottery, hit a big sports bet, or made a fortune owning some obscure digital token.
But if we engage in this kind of activity more than a handful of times, probabilities inevitably assert themselves. That is why the growing convergence of investing and gambling is so troubling. Owning NVIDIA or Tesla stock is not, in and of itself, gambling. But a betting mindset around these names has clearly taken hold. Listen to the bullish case, and you often hear a striking level of certainty about the future (AI chip demand, robotaxis, and other transformative narratives). This is not materially different from the gambler at a blackjack table who believes today is their lucky day because of the day of the week, the dealer’s hair color, or some other comforting superstition masquerading as insight.
The reality is straightforward. A competent blackjack player has roughly a 42% chance of winning (49% for the dealer and about 9% for ties). The odds are similar in games like roulette. Play long enough, and the probability of losing all the money on the table approaches 100%. This can be demonstrated mathematically through what statisticians call the “gambler’s ruin problem.[i]”
None of this is to suggest that everyone who gambles is foolish or suffering from addiction. For many people, gambling is simply entertainment. When wagers are made with the clear understanding that losses are highly likely, the harm can be limited. Investing, however, is different; here we are talking about life savings, retirement security, or the core capital a foundation relies on to fund its mission and support its community.
Returning to AI and robotaxis, the most troubling aspect of today’s environment is that very few shareholders seem focused on probabilities. Yes, NVIDIA’s and Tesla’s businesses could absolutely thrive over the next decade, and they could even exceed what is implied by current stock prices; but again, what are the odds? And no, throwing up our hands and saying that no one can know for sure does not grant permission to ignore the question entirely.
The red flags become more visible as position sizes grow. When NVIDIA represents 0.5% of a portfolio, it’s reasonable to be less rigorous about probabilities. That is not much different from walking into a casino and placing a small portion of one’s savings on a single number in roulette, with a 35-to-1 payout. The loss is controlled and manageable, while the upside is significant. But when someone allocates 7.5% or more of their life savings to a single idea, they must have an informed view of the realistic probabilities and expected payoffs across different scenarios. In the highly plausible future where heavily concentrated S&P 500 investors lose 30%, 40%, or even 50%, the most damning aspect of their mistake will be that they entered a losing game without ever acknowledging that possibility.
The larger issue is that as capital markets begin to resemble casinos, aggregate investor returns decline regardless of how carefully any individual participates. When zero-revenue NASDAQ companies command market capitalizations in the billions, index investors are exposed whether they like it or not; capital is being risked in these names simply by virtue of owning the market.
Where does this lead? Comparisons to the late 1920s or the late 1990s are increasingly common. At the same time, it is fair to say the world is different today, and simplistic historical comparisons can be just as unhelpful as reflexively condemning investors in NVIDIA or Tesla. Still, it is worth asking what drives such widespread confidence, or overconfidence. Quite simply, there appears to be little sense that negative consequences are possible. This, in turn, reflects poor or nonexistent memories. Consider the Morningstar chart below.
There are two key takeaways. First, look at the small section on the far right where the line rises steadily. It may appear brief, but it actually represents roughly a 16-year period. Now look at the long horizontal stretches that Morningstar highlights, which represent periods where markets generated no returns. There are many of them, and they often lasted a very long time. Today’s investors behave as if such periods are highly improbable and not worth serious consideration.
The uncomfortable truth is that extended bear markets cannot be prevented. We cannot, and should not, try to predict when the next will occur, but we can say with certainty that the odds of another extended one are 100%. Our responsibility is to allocate client capital with this reality firmly in mind. That means an unwavering commitment to diversification, respecting probabilities at every level of the investment process, and considering a wide range of outcomes. (And it explains our preference for free cash flow, share buybacks, and dividends, all of which help reduce downside risk.) Above all, it means never, ever ignoring history.
We can hope that markets become less casino-like, but it is difficult to see how that happens without some form of calamity first. All we can do is focus on protecting long-term client outcomes and remember a simple truth: the house always wins. One question worth asking is whether today’s deep value investors are, in fact, the house.
[i] In probability theory, the gambler’s ruin problem proves that repeated play with negative odds leads to certain loss for a player with finite capital.