The short version
- Berkshire Hathaway and the Medallion Fund are both extreme-success stories, but they are not two paths a single retail investor can choose between — only one is open to outside capital.
- Medallion's reported ~39% net annualized return (1988–2018) depends on capacity constraints that disappear at scale; the strategy is not a "scaled-up" version of anything individual investors can run.
- For a long-horizon individual, the portable lesson from Simons is not "use AI to beat the market" — it is "let rules execute the trades your emotions would mismanage."
The interesting question is not "Buffett or Simons?" — it is whether the framing of that comparison helps an individual investor at all. One vehicle is buyable in any brokerage account. The other has not accepted outside capital in roughly three decades and is reportedly hard-capped at a level of assets that institutional finance considers small. Most retail readers asked to "pick a path" between them are being offered a choice that only exists on paper.
This piece does the comparison anyway, but with the friction of reality kept in view: what each track record actually shows, what a long-horizon individual investor can and cannot port from each, and where the analogy breaks.
The two records, as the public record actually states them
Per Berkshire Hathaway's 2024 annual letter, the company's per-share market value compounded at approximately 19.9% per year from 1965 through 2024, against approximately 10.4% for the S&P 500 with dividends reinvested over the same window (Berkshire Hathaway 2024 Shareholder Letter). That is roughly 60 years of audited, replicable-in-public-markets performance. Berkshire Class B shares are buyable in any taxable account or IRA.
The Medallion Fund's record is more remarkable in magnitude and far less accessible in practice. The most-cited account is Greg Zuckerman's 2019 book The Man Who Solved the Market, supplemented by Bradford Cornell's 2020 paper "Medallion Fund: The Ultimate Counterexample?" published in The Journal of Portfolio Management. Both report that Medallion produced gross annualized returns near 66% and net (after Renaissance's high fees) near 39% from 1988 through 2018, with negative years rare and shallow. The fund has been closed to outside investors since approximately 1993 and runs primarily Renaissance employees' own capital. Reported AUM has been kept near a low-tens-of-billions ceiling — small by hedge-fund standards, almost certainly because the strategy's edge degrades as positions grow.
| Dimension | Berkshire Hathaway | Medallion Fund |
|---|---|---|
| Method | Concentrated equity, long-horizon value | Short-horizon statistical arbitrage |
| Holding period | Years to decades | Days, sometimes shorter |
| Reported net annualized | 19.9% (1965–2024) | ~39% (1988–2018) |
| Benchmark over same window | S&P 500: 10.4% | S&P 500: ~10% (1988–2018) |
| Capacity | ~$1T+ market cap, scales with float | Reportedly capped near low tens of $B |
| Open to outside capital | Yes (BRK.A / BRK.B) | No, closed since ~1993 |
| Replicable by retail | Partial — buy the vehicle | Effectively zero |
Sources: Berkshire Hathaway 2024 Shareholder Letter; Zuckerman, The Man Who Solved the Market (Portfolio, 2019); Cornell, The Journal of Portfolio Management, 2020.
What the Medallion record really tells us
Cornell's framing in the 2020 paper is more useful than the headline number. He notes that Medallion's returns over the studied window were essentially uncorrelated with market beta — the fund did not depend on equities going up, which is what makes the record so anomalous against efficient-market priors. But the same paper points out the unavoidable corollary: a strategy whose edge is uncorrelated with the broad market and whose AUM is held below a small ceiling is, by construction, not a strategy a $50 trillion equity market can produce many copies of. The capacity constraint is not a footnote on Medallion's record; it is the load-bearing assumption.
This matters because retail investors routinely encounter products marketed as "AI-driven" or "quant-style" with the implicit suggestion that the Medallion lineage applies. It does not. A publicly listed ETF is, by definition, capacity-unconstrained relative to Medallion. Whatever edge it claims has to survive both daily creation/redemption flows and AUM that can scale into the billions or tens of billions. That is a different problem than the one Renaissance solves.
Medallion is not a scaled-down hedge fund — it is a deliberately scaled-down hedge fund. Stripping the capacity ceiling does not give you a bigger Medallion; it gives you a worse one.
What the Berkshire record really tells us
Berkshire's record is in some ways the harder one to argue with, precisely because it is replicable in a way Medallion is not. An investor who bought Berkshire B shares in 1996 (when they were issued) and held to 2024 captured most of what made the company a multi-decade compounder. There is no closed door, no fee structure carving off the edge, and no capacity argument standing in the way.
The honest caveats are different ones. The early Berkshire compounding (1965–1980s) happened at a fraction of today's company size; recent decades have produced returns much closer to the S&P 500 with lower volatility, which is what we should expect — Sharpe (1991), in "The Arithmetic of Active Management," makes the structural point that as an active investor's footprint becomes large enough to move markets, the alpha they extract converges toward the market's. Berkshire is roughly $1 trillion in market capitalization. It cannot meaningfully outperform the index from here without that movement showing up in the index itself.
Buffett has effectively said this in recent letters. The reasonable expectation for the next 20 years of Berkshire is "modest premium to the S&P 500, with materially lower drawdown" — not 19.9% per year. That is still an excellent outcome. It is not the 1965–2000 outcome.
Where the comparison stops being useful
The deeper point is methodological. Buffett's edge over six decades came from temperament and capital allocation: holding through drawdowns, sizing concentrated positions, owning insurance float as a low-cost funding source. Simons's edge came from infrastructure: data feeds, latency, models, and a research culture that treated alpha as a perishable good to be replenished. These are not two flavors of the same skill. They are different occupations.
An individual investor reading "AI-quant" pitches in 2026 is rarely being offered Simons's actual edge. They are being offered a factor product — value, quality, low-volatility, momentum — wrapped in modern marketing language. Those products are legitimate and well-studied (see our analysis of QQQM vs. QRFT for one example), but they are descendants of Fama-French-style factor research, not of Renaissance.
What is actually portable from each
From Berkshire: hold long enough that taxes and behavioral mistakes become rounding errors. Diversify within the equity allocation. Treat drawdowns as a feature of equity ownership, not a bug to time around. The discussion of whether buy-and-hold still works in 2026 resolves toward "yes, if you actually hold."
From Renaissance, the portable lesson is structural rather than technical: rules execute trades that emotions would mismanage. An individual investor cannot replicate Medallion's edge, but they can replicate the discipline that lets Medallion's edge compound — pre-committed rebalancing bands (Daryanani 2008; Vanguard 2024), written allocation rules, and a refusal to override the system mid-drawdown. The Mulden editor's open-source dashboard exists for exactly this — applying the same rebalancing literature to a single buy-and-hold portfolio. The framework is the deliverable, not the prediction.
At-a-glance scoreboard
| Category | Winner | Margin |
|---|---|---|
| Replicability for retail | Berkshire | Decisive — Medallion is closed |
| Reported alpha (within window) | Medallion | Material — but capacity-bounded |
| Multi-regime longevity | Berkshire | 60 years across cycles |
| Useful lessons for individual investors | Both, differently | Buffett: temperament. Simons: rules. |
Editor's read
If forced to anchor a long-horizon portfolio to one of these two philosophies, the editor leans heavily Buffett — not because the value approach is intrinsically superior, but because it is the only one of the two that an individual can actually inhabit. Medallion's record is fascinating evidence about market microstructure and the limits of efficient-markets theory. It is not an investing path. The honest takeaway is that the Renaissance lesson worth keeping is procedural — rules-based execution, drawdown discipline, refusing to override the system on emotion — and that lesson is fully compatible with a buy-and-hold ETF allocation.
FAQ
Can a retail investor buy into the Medallion Fund?
No. Medallion has been closed to outside capital since approximately 1993. Renaissance Technologies runs other funds open to outside investors, but those have not produced returns comparable to Medallion's, which is itself informative about where the edge actually lives.
If quant works for Renaissance, why doesn't it work for retail "AI ETFs"?
Capacity. Medallion's reported edge depends on staying small. A publicly traded ETF is, by design, scalable to billions in AUM and faces continuous creation/redemption flows. A strategy that survives both is by definition a different strategy than the one Medallion runs.
Has Berkshire underperformed the S&P 500 in recent years?
Over rolling 5- and 10-year windows since the mid-2010s, Berkshire's results have been close to the S&P 500's, sometimes ahead and sometimes behind, with materially lower drawdown. This is consistent with Sharpe's (1991) point that very large active managers tend to converge toward the index they sit inside.
Is Berkshire effectively a value-tilted index fund now?
That is a defensible characterization. The float-funded insurance business plus the operating subsidiaries plus the equity book combine into something that behaves more like a low-volatility, value-tilted equity fund than like a concentrated stock-picking vehicle. Buffett himself has guided expectations in this direction.
What should a long-horizon individual actually take from this comparison?
Two things, kept separate. (1) Long-horizon equity ownership through a low-cost vehicle is the workable version of "the Buffett path" for almost everyone. (2) The lesson worth porting from Renaissance is rules over instinct — written allocation, written rebalancing bands, no mid-drawdown overrides.
What this comparison can and can't tell you
The Berkshire record is 60 years of audited performance across multiple regimes — the 1970s inflation, the 2000–2002 dot-com bust, 2008, 2020. The Medallion record is 31 years (1988–2018 in the most-cited window) of a closed fund whose detailed methodology has never been independently verified. Both are useful evidence about what is possible, but both should be read with the appropriate humility about survivorship and selection. We are not seeing the funds that ran similar strategies and failed. The base rate for "long-running quant fund that produces 39% net" is, by definition, very small. Treat any single-fund record — including Medallion's — as upper-bound evidence rather than typical evidence.
Key takeaways
- Berkshire and Medallion are not "two paths" between which a retail investor can choose. Only Berkshire is a vehicle individuals can buy.
- Medallion's reported edge is structurally inseparable from its capacity ceiling. "Bigger Medallion" is not a thing the strategy admits.
- The portable Buffett lesson is temperament; the portable Simons lesson is rules-based execution. Both fit comfortably inside a long-horizon ETF allocation.
- Marketing language using "AI" or "quant" rarely describes a Renaissance-style edge. More often it describes a factor product, which is a different and well-understood thing.
- For most readers, the practical question is not Buffett vs. Simons. It is whether you have written allocation and rebalancing rules and follow them.
The editor holds broad-market and factor-tilted ETFs through low-cost vehicles; does not hold BRK.A or BRK.B at the time of writing; has no access to Medallion or any Renaissance product.
Methodology
Berkshire performance figures are from the Berkshire Hathaway 2024 Shareholder Letter (audited per-share market value, 1965–2024) and benchmark figures from S&P 500 total return over the same window. Medallion figures are as reported in Zuckerman, The Man Who Solved the Market (Portfolio, 2019), and analyzed in Cornell, "Medallion Fund: The Ultimate Counterexample?" The Journal of Portfolio Management, 2020. The arithmetic-of-active-management argument is from Sharpe (1991), Financial Analysts Journal. Macro context (10-year Treasury 4.39%, VIX 16.99) is from FRED, asof 2026-05-01. No claims in this article rest on proprietary or unverifiable performance figures.
This article is for educational purposes and does not constitute personalized financial advice. See Disclaimer.