This interview features quant investor Wes Gray, who argues that real investment edge comes from pairing a long-term strategy with patient capital, not from building smarter models. He says momentum (buying winners) beats value (buying cheap stocks) in data, but value is easier for humans to stick with during downturns. Key holdings: QVAL ETF (his concentrated value fund with ~40 stocks), DFA (a giant firm whose size has eroded the 'price-to-book' value signal), and Vanguard (its massive scale risks political backlash).
At a Glance Wes Gray, founder of Alpha Architect, discussed purification strategies for quantitative factor investing on the Invest Like the Best podcast. The core argument is that adopting more concentrated and pure factor exposures (such as value and momentum) can enhance investment performance. K
Wes Gray (Founder of Alpha Architect, former Marine, PhD in Finance from the University of Chicago) articulates his quantitative factor investing philosophy in the interview: true excess returns do not come from smarter models, but from pairing long-term strategies with long-term capital and using tax structures to force investor patience. Core judgment: the momentum factor outperforms the value factor on a pure data basis, but behaviorally, the value factor is easier to stick with.
Wes Gray argues that quantitative investing should be strictly divided into two distinct domains, and investors must clearly understand which game they are participating in.
High-Frequency/Market-Making Side: Represented by Renaissance Technologies and Getco, the core lies in microsecond-level information advantages and competition among machine learning algorithms. Gray describes a client in Chicago: "Their 40 math/physics PhDs all focus on a single central machine learning algorithm, trading at the microsecond level, with annualized returns of 30-40%. But their ROA (return on assets) might be equivalent to buying the S&P 500—they are not investors, they are traders." (Meaning: high returns come from high leverage, not genuine investment alpha.)
Factor Investing Side: The domain where Gray and O'Shaughnessy operate—"using quantitative tools for long-term investing, minimizing the burden of human decision-making." The core is not about building better models than AQR or anyone else, because "we can all run Compustat data until our faces turn blue; no one has a true advantage in models."
Key Distinction: High-frequency trading is a "winner-takes-all" zero-sum game requiring continuous innovation; the core advantage of factor investing lies in "patience and the ability to endure pain." Technology will not change the behavioral pattern that "people always throw the baby out with the bathwater."
Gray advocates for a highly concentrated factor exposure strategy that does not neutralize industries, standing in stark contrast to the "risk-first" philosophy of mainstream quantitative institutions.
Core logic: If an investor hires you to capture the value factor premium, you should maximize exposure to that factor rather than diluting it through industry neutralization, risk parity, or other means. "If everyone has to play like a Vanguard fund, then mechanically, that almost destroys any ability to generate excess returns or endure pain."
Number of holdings: The QVAL ETF holds approximately 40 stocks. Gray believes that 30–50 is a reasonable range for "maximizing active risk without being foolish." "We can achieve the diversification we need while still delivering high-purity 'blue ice meth' — then investors can add their own 'baby powder' to dilute it as they see fit."
Boundaries of risk management: Gray acknowledges the risks of extreme concentration (e.g., retail stocks may account for up to 50% of the current portfolio) but argues that within the context of a global value + momentum combination, such risks are manageable — momentum often forms a "convex hedge" against value.
Gray admits: If purely based on data, momentum is the superior choice; but as a human, he chooses value.
Data perspective: Momentum performs better in robustness studies across all asset classes and all time periods. "Fama and French, in their 'Anatomy of Anomalies' paper, call it the premier anomaly—they can't even explain it with the efficient market hypothesis." (Meaning: Even the staunchest supporters of efficient markets acknowledge that momentum cannot be explained by risk models.)
Behavioral perspective: Gray admits, "I would continue to hold value even if it dropped 99%, but I would likely abandon momentum once it underperforms." He cites Kahneman's System 1/System 2 framework, arguing that momentum's "counterintuitive" nature makes it difficult to stick with under pressure. "If I were a robot, I would choose momentum. But I am human, raised by Ben Graham."
Comparative data:
| Dimension | Value Factor | Momentum Factor |
|---|---|---|
| Empirical Robustness | Strong (multiple markets, multiple periods) | Stronger (all asset classes, all periods) |
| Academic Consensus | Explainable by risk models | Unexplainable by the efficient market hypothesis |
| Behavioral Persistence | High (intuitively reasonable) | Low (counterintuitive, easily abandoned under pressure) |
| Gray's Personal Choice | Value (behaviorally sustainable to zero) | Momentum (data-superior but behaviorally unfeasible) |
Gray bluntly states that the price-to-book ratio is a "terrible" value factor, citing two reasons: supply-side arbitrage and mechanical flaws.
Supply-side arbitrage: DFA (Dimensional Fund Advisors) manages approximately $600 billion in assets, and its core strategy is to buy stocks with low price-to-book ratios. "Anytime you have a factor, and there's a company with $600 billion in permanent capital dedicated to it, that factor gets arbitraged away." (In other words, DFA's scale has already eroded the excess returns of the price-to-book factor.)
Mechanical flaws: Stock buybacks reduce book value, passively inflating the price-to-book ratio, which causes genuinely value-creating companies to be excluded from the "cheap" stock category. Gray recommends alternative metrics such as the enterprise multiple, arguing that these indicators can "add a spark of mispricing on top of the risk premium."
Key warning: Gray notes that if Alpha Architect's factors (such as the enterprise multiple) also accumulate to $500 billion in scale, they too would be arbitraged away—but "before that happens, we will have already created significant value for early investors."
Gray argues that the current passive investing boom is not permanent capital allocation, but a new form of performance chasing, carrying significant overlooked active risks.
Core Argument: The proper use of passive investing is "as a portfolio tool to efficiently capture market average returns at low cost," but the actual marketing claims it "outperforms 95% of active managers" — which is essentially performance chasing. "If massive amounts of capital chase a strategy that seems to always win, has infinite supply, and makes anyone who doesn't buy it look foolish, I don't recall any historical macro equilibrium ending well."
Overlooked Active Risk: Investors buying S&P 500 index funds are far from holding a "global market portfolio" — this is an extreme active bet on large-cap U.S. stocks. "By feeling passive, you are actually making a huge active bet — and you don't even know it."
Vanguard's Long-Term Risk: Gray believes Vanguard's biggest risk is not market-related but political. "When a senator realizes that a one-stop shop owns 20% of the entire U.S. corporate sector, things turn sour." He draws a parallel to Amazon: excessive scale inevitably invites regulation, and regulation destroys the advantages of the business model.
Gray builds Alpha Architect’s business model on the Shleifer-Vishny (1997) “limits to arbitrage” theory: pairing smart strategies with short-term capital inevitably leads to a lose-lose outcome.
Problem Diagnosis: Everyone in the market knows which strategies work (value, momentum), but these strategies are “long-term trades” while capital is “short-term capital.” The result: the manager wins first, then underperforms and gets fired, the client loses, and no one benefits.
Solution: Compete not on model sophistication, but on “investor education” and “capital durability.”
Ideal Client Profile: Gray uses the acronym “EDUCATED” — engineer types, process-driven, distrustful of Wall Street, wanting to understand the logic behind the system; or private business owners, who naturally have a 10–20 year mindset.
| Position | Guest Stance | Key Data |
|---|---|---|
| QVAL ETF (Alpha Architect Value ETF) | Bullish (own product) | Approximately 40 holdings, concentrated exposure to value factor |
| DFA (Dimensional Fund Advisors) | Neutral (respectful but notes its scale has arbitraged away the book-to-price factor) | Manages approximately $600 billion in assets |
| Vanguard | Risk warning (excessive scale may invite political risk) | No specific data provided |
| Best Buy | Cited as a value stock example | Held by Gray, representing "Amazon fear" |
| AQR | Neutral to slightly positive | No specific data provided |
| Renaissance Technologies | Cited as a benchmark for high-frequency trading | Annualized return of 30-40% (as quoted by Gray) |
| Getco | Cited as a market-making example | Original team left after acquisition by KCG |
1. "The real advantage isn't building a better mousetrap, but pairing educated capital with a strategy." (Gray) — Everyone can run the same data, model differences are negligible; the true moat is finding long-term capital that understands the strategy and can endure the pain.
2. "Momentum beats value in the data, but as a human I choose value — because I wouldn't abandon value even if it drops 99%, but I would abandon momentum once it underperforms." (Gray) — A core insight from behavioral finance: the optimal strategy is not the executable one; self-awareness matters more than data.
3. "Book-to-price has been arbitraged away by DFA's $600 billion in permanent capital — it's not a bad factor, but if you want to maximize the value premium, you should pick metrics that still have a 'spark' of mispricing." (Gray) — Factors themselves can be arbitraged away by scale; choosing alternative metrics (e.g., enterprise value multiples) not yet targeted by large-scale capital is a sensible strategy.
4. "Passive investing is not permanent capital — it's performance-chasing in disguise. When everyone buys the S&P 500, they are actually making an extreme active bet on large-cap U.S. stocks, without knowing it." (Gray) — S&P 500 index funds are far from a "global market portfolio"; investors take on massive active risk while feeling passive.
5. "Vanguard's biggest risk is not the market, but politics — when a senator realizes one entity owns 20% of U.S. corporate equity, regulation will step in." (Gray) — Excessive scale inevitably invites non-market intervention, analogous to Amazon's regulatory risk.
6. "Good real estate investors are rich not because they are smart, but because they are locked into low-cost assets by tax liabilities, forced to hold for 20 years." (Gray) — Alpha Architect uses tax structures to create "forced patience," turning behavioral disadvantages into advantages.
7. "Captain Redinger-style kindness — making you better by torturing you. We do the same with investors: explain why this strategy is so terrible, and if you still want to stay, you have at least a small chance of success." (Gray) — Educating investors about the "pain" of a strategy is more important than selling the "returns"; this is key to solving the "limits to arbitrage" problem.
8. "If you could tell me that passive capital is permanent and will be held for 50 years, then all arbitrage theories would hold. But can you tell me that?" (Gray) — The essence of the passive investing boom is another form of short-term performance-chasing, not permanent capital allocation, so active strategies still have room to survive.