#29. You can’t invest in generations
January 12, 2026
YoYou cannot invest in generational timeframes. There is no reliable generational thesis or company. Capital endures. Companies do not.
There’s something very sexy about saying you invest on a generational timescale. It’s become popular among “long-term allocators” and aspiring Warren Buffets building their holding companies. *Generations*
Oh did my investment miss earnings? I still see something nobody else sees over *generations*
That sounds amazing.
So how do you invest in generations?
You don’t. You can’t. Not like that.
When serious investors say “generational,” they don’t mean century long holding periods (in this essay though, I do mean exactly that). But they do mean capital that doesn’t need to resolve itself quickly. Even this softer definition still assumes a stability at the company level that does not exist. Long time horizons amplify, rather than smooth, organizational fragility.
When those investors talk about “generational” today (think AI company), what they usually really mean is five to ten years. That’s the outer boundary of institutional patience. Everyone claims patience. Structure reveals how much they actually have. And, for the majority of our financial world, the answer is very little.
Where long-term ends:
Public market capital (long/short, long-only, mutual funds): 1 day to 3 years. Beyond that, capital rotates, mandates change, and institutional-level theses are called.
Public company CEOs (non-tech): 7 years max (the average life span of a public CEO). Incentives are quarterly. Reputational risk is immediate. Incentives do not align with generational bets. No CEO is risking incentive comp to build a product your granddaughter will appreciate when she’s your age.
Venture funds: ~10 years. Venture funds may last 12–15 years, but their company-level exposure window is closer to a decade. Liquidity pressure arrives well before generational durability can be proven.
Big tech CEOs: the outlier edge in modern institutional patience. Non-founder tenures are impressive, from 6 years (Sundar) to 14 years (Tim Apple). Founder tenures go further. Founder-controlled firms with voting insulation represent the outer edge of modern corporate patience. Zuck has absorbed years of criticism with the Meta rewrite. Even still, horizons are measured in decades, not generations.
Long-term capital exists. It can persist across centuries. It exists in families, endowments, institutions. But long-term homes do not. Not structurally. Not historically. “Long-term investing” quietly conflates the two. Over a generational horizon, systems survive and compound while firms decay.
But (I hear you say) an investor with an infinite time horizon could do it. Generational companies exist (I know). In theory, yes. There are assets that have stood for generations. But generational endurance is only identifiable ex-post, not ex-ante. If you were a forever investor generations ago, which one would you (could you) pick?
Well, maybe you’d pick Bank of New York Mellon. Founded in 1784, started publicly trading in about 1792. 230 years is generations, no? Yes! Winner! I would also add Coca-Cola, P&G, Cargill, and Exxon. All trace their roots to the late 1800s. They are amazing examples of companies that have endured by adapting.
But the inherent difficulty (impossibility!) of translating this into an investment strategy is that these are only apparent ex post. These are a few companies out of thousands that crossed the threshold. It looks inevitable after the company survives. For example, the Coca-Cola of 1920, 1970, and today are not the same beasts. What has (amazingly) persisted was not a single compounding advantage, but an organization capable of serial reinvention. That distinction matters. Generational endurance such as this is visible only in hindsight. It cannot be reliably selected because it depends on a series of CEOs, management teams, and boards capable of building off of predecessors in a way that is impossible to predict (see Portfolios of People). Pointing to rare survivors as evidence of predictability mistakes outcome for process.
And not only is it nigh impossible to pick generational winners, but it’s become increasingly less likely that any will exist. Take this from BofA:
“Roughly a third of the S&P 500 has been replaced since 2015. In 1958, the average seven-year rolling lifespan of a company on the S&P 500 was 61 years. By the 1980s, it had dropped to 30 years, by 2016 it was 24 years, and by 2021 it was 16 years. If we continue on this road, by 2027, companies could last just 12 years.”
Multiple studies converge on the same conclusion: the typical public company now lasts a fraction as long as a human life. Could you confidently bet on any single name to exist in a generation? Buying today’s S&P for the next generation is not a bet on today’s companies. If you were hoping to invest at the company-level for generations, the traditional route, public markets, increasingly deny you that opportunity.
Index investing serves as a type of system. Indices buy a system one can hold, in theory, forever. But even those investors wouldn’t claim an index is a forever investment. Indexing solves for survivorship only, but an index still takes a direction (S&P 500, US, Equities) that preserves the stated exposure. It does not evolve to what the forever timeframe might require. Forever portfolios require governance.
So:
Human life expectancy (US) = 79 years
Average S&P tenure = 12-16 years
You will outlive most companies you invest in. Your grandchildren will outlive almost all of them. The crux of the issue with our generational investors is nothing financial, but human fallibility interacting with long timelines (see Primary Data and Human Fallibility). I get excited about generational technologies. I get excited about the companies that build them. But don’t conflate that with buy and hold forever. The durable object when investing with an enduring mindset is not the firm. Investing with a “generational” mindset is to have the capital allocation logic that continuously adapts.
“Forever” belongs to systems, not tickers. If you believe companies endure, buy stocks. If you believe capital endures, design systems. Most modern institutions understand this distinction. But they are not built to act on it.
