Single-strategy and quant funds

The third of the roadmap's three models. A pod shop assembles many mediocre edges into a good portfolio. A single-strategy fund does the opposite: one edge, understood deeply, applied at whatever scale it tolerates.

That last clause is the whole lesson.

Capacity is the organising constraint

M1.5 and M2.7 already gave you the arithmetic, and here it becomes an organisational fact:

   gross alpha   ∝  Q          your edge scales with size
   trading cost  ∝  Q^1.5      impact grows with participation

   net profit peaks, then FALLS

For a single-strategy fund, that peak is a hard ceiling on the entire business. There is a size at which the strategy stops working, and no amount of capital, talent or technology moves it much — because the constraint is the market's ability to absorb the trading, not the fund's ability to think.

This produces behaviour that looks bizarre from outside and is entirely rational:

   ✓ closing funds to new investors while performing brilliantly
   ✓ returning capital to investors
   ✓ refusing to disclose anything, since the edge decays if copied
   ✓ paying employees rather than gathering assets

An asset manager's instinct is to grow. A capacity-constrained fund's instinct is to stay small and keep the returns, because the marginal dollar makes every existing dollar worse.

Medallion, the extreme case

Renaissance Technologies' Medallion fund is the sharpest illustration, and the most famous track record in finance — reported gross returns averaging in the region of 60–70% a year over decades, before very high fees.

Everything else about it follows from capacity:

• Closed to outside money since the 1990s. It runs employee capital.
• Profits distributed annually rather than compounded, holding the fund near its capacity limit by construction.
• Extreme secrecy — no publication, aggressive non-competes, staff drawn from mathematics, physics and code-breaking rather than finance.

The instructive part is what Renaissance did next. It launched separate institutional funds — RIEF and others — open to outside investors and running different, more capacious strategies. Those funds have performed respectably rather than miraculously. Same firm, same people, same technology; the difference is capacity. It is the clearest natural experiment in the industry that the edge, not the organisation, is the scarce thing.

The scalable end

At the other extreme sit firms whose strategies genuinely absorb capital, and their character is completely different:

                  CAPACITY-CONSTRAINED        SCALABLE
   ──────────────────────────────────────────────────────────────
   returns        very high                   market-like to good
   size           small, often closed         very large
   secrecy        total                       often publishes research
   fees           very high                   competitive, falling
   horizon        short — days or less        long — months to years
   example        Medallion                   AQR, Bridgewater

AQR is the clearest case of the scalable model: factor investing — value, momentum, carry, quality — implemented cheaply at enormous scale, with Cliff Asness and colleagues publishing their research openly. That openness is not altruism; it is coherent with the business. Factor premia are (arguably) compensation for risk rather than a secret, so explaining them does not destroy them, and being the credible academic voice attracts institutional capital.

Compare that with a firm whose edge is a decaying signal: publishing would be suicide. How open a firm is tells you what kind of edge it thinks it has. That single heuristic reads the industry surprisingly well.

Bridgewater is the large macro example — Ray Dalio, the Pure Alpha fund, a heavily codified investment process and a famously distinctive internal culture. Macro strategies trade the most liquid instruments on earth (rates, FX, index futures), so capacity is enormous, which is why it could grow to be among the largest hedge funds.

Who's who

   firm                     known for
   ──────────────────────────────────────────────────────────────
   Renaissance              Medallion — the best track record in
   Technologies             finance, closed to outsiders. Jim Simons,
                            a mathematician who hired scientists
                            rather than traders.
   D. E. Shaw               among the earliest quant funds; a notable
                            alumni tree, including Jeff Bezos.
   Two Sigma                technology-first quant, large data and
                            engineering organisation.
   AQR                      factor investing at scale, and unusual
                            openness — publishes serious research.
   Bridgewater              the largest macro fund; Pure Alpha; a
                            codified and much-discussed culture.
   TGS / Quadrature /       extremely secretive, reportedly very high
   XTX                      returns. XTX is notable as a large
                            non-bank FX market maker.
   ──────────────────────────────────────────────────────────────

Why most quant funds are not Medallion

A closing corrective, and it is M2.7's argument applied to an industry.

Medallion is quoted constantly, which creates the impression that quantitative investing routinely produces such returns. It does not. Medallion is the extreme tail of a distribution whose middle contains a great many funds delivering ordinary returns, and whose left tail contains funds that quietly closed.

Everything M2.7 said applies at the level of firms: survivorship bias (you hear about the survivors), multiple testing (thousands of funds have been launched; the best track record among them would look extraordinary even if none had an edge), and the standard error of a Sharpe estimate, which makes a few good years nearly uninformative.

Medallion's record is long and consistent enough to be genuinely beyond that critique — decades, not years. Almost nothing else is. Keeping both facts in mind at once is the mark of someone who has read past the folklore.

Source: Gregory Zuckerman, The Man Who Solved the Market, on Renaissance — the roadmap lists it under M6 and it belongs here too. AQR’s research library is free and unusually good; reading a paper by a firm that manages the strategy is a different experience from reading one by an academic. Flirting with Models interviews systematic managers about how their businesses actually work.