What this covers
Tyler Cowen interviews Cliff Asness, a quantitative investment strategist, on why systematic factor strategies—momentum and value investing chief among them—persistently outperform markets. Asness constructs his argument around a threefold framework: every market regularity stems either from data accident, behavioral error, or compensation for genuine risk-bearing. He contends that momentum and value succeed not because markets are broken but because they exploit durable mispricings born from how investors systematically over and underreact to information. The conversation moves between theory and practice, touching on why most active managers underperform, how fees can spoil even sound strategies, and why Discipline as Edge matters more than most investors admit.
The discussion ranges across the psychology of investing, the nature of true risk, and disagreements with academic consensus. Asness disputes the notion that Gene Fama—even the architect of efficient markets theory—believes markets price everything perfectly; Fama himself teaches that perfect efficiency is an extreme and unlikely hypothesis. Asness develops a distinction between volatility and genuine risk, arguing that [Risk as Hurting When It Hurts to Be Hurt](concept:Risk as Hurting When It Hurts to Be Hurt) is what matters—an asset safe in bad times is different from one that fails precisely when your life is falling apart. He explores Leverage Aversion as a market driver, explains why momentum should be added as a sixth factor alongside Fama and French's model, and makes a stark claim: the costliest error in finance is convincing the world that something near-certain is actually certain, as happened with housing before 2008. The conversation also touches on high-frequency trading's effects on market fairness, the debate over what widespread indexing means for active managers, and why low-risk stocks paradoxically outperform—an anomaly discovered by Fischer Black decades ago but largely forgotten.
Asness argues that systematic factor strategies like momentum and value reliably beat the market not because markets are broken but because they exploit durable behavioral errors (over/underreaction) and risk premia that most investors cannot stick with through painful stretches — and that the chief danger in finance is mistaking near-certainty for certainty.
- Momentum and value 'work' statistically over long horizons but have excruciating streaks that deter most investors, which is precisely why they persist.
- Anomalies require someone on the other side of the trade, explained by either risk-bearing or behavioral over-extrapolation.
- Telling investors something is riskless when it only mostly protects them (e.g. money market funds) is the most dangerous condition in markets.
Value and momentum are empirically powerful long-term systematic strategies that survive because most investors abandon them during inevitable bad periods.
- A momentum investing strategy—buying a portfolio of what has risen over the past 6 to 12 months and selling what has fallen—appears to reliably beat the market over the long term, despite seeming an insane proposition.
“A momentum investing strategy is the rather insane proposition that you can buy a portfolio of what’s been going up for the last 6 to 12 months, sell a portfolio of what’s been going down for the last 6 to 12 months, and you beat the market. Unfortunately for sanity, that seems to be true.”
- Even strategies that work long-term don't get arbitraged away because most investors cannot live through the bad periods—they decide the strategy will never work again at exactly the wrong time, which is why discipline is the scarce resource that preserves the edge.
“I think it is harder than you might guess, even if something works long term, to have it go away because a lot of investors can’t live through the bad periods. They decide why it’s never going to work again at the wrong time.”
- Systematic strategies like value, momentum, quality, and low-risk investing survive in a sweet spot: good enough to matter greatly if you can follow discipline, but not so easy that the world piles in and arbitrages them away; if everyone did them yesterday they would disappear.
“If everyone did them yesterday, they would go away. They work... in kind of a sweet spot. Good enough to be really important if you can follow discipline, not so good enough that the world looks at it and goes, “this is easy.””
- Value investing—buying what is cheap on price relative to fundamentals (earnings, book, sales) and underweighting the expensive, including buying five-year losers—is a very good long-term strategy, opposite in spirit to momentum, and works because if any pricing errors exist the cheaper-looking stock with identical fundamentals will outperform.
“buying what’s cheap in pretty much any way — price divided by something reasonable, price divided by earnings, books, sales. Just buying five-year losers... is also a very good strategy, very much the opposite in spirit to momentum.”
- The biggest mistake ordinary investors make is overtrading by being 'momentum investors at a value time horizon'—chasing what has worked over five years (a value horizon where reversal applies) and bailing out of strategies after a few tough years out of nausea, rather than either committing to disciplined 6-12 month momentum or holding value for 3-10 years.
“They are momentum investors at a value time horizon... If you wait five years and buy what’s worked for five years, you can call that a negative value investor or a momentum investor working with the wrong numbers.”
- Almost any market regularity—such as momentum—is larger in small-cap stocks than in large-caps (e.g., ~100-125 bps in large-caps vs ~250-300 bps in small), but this is not as good a deal as it sounds because small-cap risk and fluctuations are also larger.
“you bought the one‑third of stocks with the superior 6 to 12 months returns, you’d probably make 100, 125 basis points extra long‑term, on average. You do that in small stocks. It’s more like 250 to 300.”
- Momentum should be added as a sixth factor to Fama and French's factor model because it adds tremendous return versus their model and is negatively correlated with value, which makes the combination of two independently profitable strategies statistically even stronger.
“I don’t see how momentum is not their sixth factor. It adds a tremendous amount of return versus their model.”
Factor strategies succeed by exploiting behavioral investor errors like overconfidence, overreaction, and over-extrapolation rather than pure risk compensation.
- Because the average dollar can't beat the average (a mathematical identity, not Lake Wobegon), any active strategy like value or momentum requires someone to take the opposite side; that counterparty is either willingly bearing risk for a lower expected return, or making a behavioral error by over-extrapolating, and you must always be able to answer 'who is on the other side?'
“This is not Lake Wobegon. We can’t all beat the index. It’s actually a precise mathematical identity.”
- Every empirical regularity that predictably works (or fails) does so for one of three reasons: complete accident (data mining), a behavioral story (someone making a systematic error you exploit), or risk (in a rational world you get paid for bearing risk); these are not mutually exclusive and can coexist.
“Every empirical regularity, meaning something that works or is terrible with some predictability, it does so for one of three reasons. The worst one is complete accident. You’ve data mined.”
- Looking at the data, the two most common investor errors are overconfidence and over-extrapolation; people are overconfident that they correctly understand new news yet underreact in incorporating it, and that overconfidence eventually leads to overreaction.
“the most common error is two kinds. They’re not exactly the same thing but overconfidence and over extrapolation.”
- The two leading behavioral explanations for momentum are underreaction (anchoring and adjustment—prices move on news but not all the way, so they continue) and overreaction (positive feedback as people chase prices); both can be simultaneously true and the mix can vary over time.
“The first is called underreaction. Simple idea that comes from behavioral psychology, the phenomenon there called anchoring and adjustment. News comes out. Price moves but not all the way.”
- Nobody, including the godfather of efficient markets Gene Fama, believes markets price things perfectly; Fama deliberately tells his Chicago class that markets are almost certainly not perfectly efficient, because perfect efficiency (all information fully reflected) is an extreme hypothesis.
“he looks at the class and goes, “markets are almost certainly not perfectly efficient.” And this is University of Chicago with the godfather of efficient markets”
Most active managers and hedge funds fail to beat passive indexing because they charge too much, don't hedge adequately, and fall victim to the same behavioral errors they claim to exploit.
- The counterparty to factor strategies includes both stupid individual stock pickers and constrained or biased professionals: hedge funds appear to fight the value effect by buying expensive stocks, and leverage-constrained mutual fund managers get pushed into higher-beta stocks and likely overpay for them.
“they seem to buy more expensive, not cheap stocks... they are fighting the value effect. I think at least some of it is coming even from the very “smart” investors.”
- If you bought one of every hedge fund that would take your money, you'd likely be better off estimating their average stock-market exposure and buying an index fund, because as a group hedge funds don't hedge enough—running ~0.8 correlation to the S&P 500 over seven years—and charge too much.
“You’d probably be better off figuring out what their average exposure to the stock market is and go buying an index fund.”
- Active management is overrated; while certain systematic strategies can win, on average people try too hard to beat the market and pay too much for it, so most investors would do fine simply following low-cost indexing à la Jack Bogle.
“Active management I think is overrated... on average I think people try too hard to beat the market and pay too much for it.”
- As alternative strategies move into mutual fund vehicles ('liquid alts') and distribution to non-accredited investors grows, these products have largely replicated the same hedge fund problems—not fully hedged and probably still too expensive—even though reasonable core strategies exist.
“I think they’ve largely been replicating some of the same problems... they’re not fully hedged and are probably still too expensive.”
- There is no investment process so good that there isn't a fee high enough to make it bad; fees can convert even genuinely effective strategies into poor net outcomes for investors.
“There’s no investment process so good that there’s not a fee high enough that can’t make it bad.”
- A properly run systematic quantitative manager should not have strong opinions on individual stocks or sectors, because the strategies rely on broad statistical averages across thousands of long and short positions; if your systematic manager tells you to put it all on one sector, you should run.
“A quantitative systematic manager like me shouldn’t know a lot... If I have an idea, worse, if I have a very strong opinion, I’m just doing my stuff wrong.”
- Whether widespread indexing makes it easier for active managers to win is genuinely unresolved: intuition says fewer competitors makes winning easier, but the average-can't-beat-the-average constraint means if everyone is passive there's no one willing to take the other side of your bet, creating a paradox finance theorists still debate.
“Instinctively, you want to say it’s going to be easier to beat the market if fewer people are trying... On the other hand... if everyone else is passive, how do you induce them to take a bet away from passive?”
Valuation measures forecast returns over decade-long horizons but fail completely over months, so investors using backward-looking assumptions systematically overshoot future expectations.
- Valuation measures like the Shiller CAPE for stocks and real yields for bonds are statistically powerful for forecasting returns over the next ~10 years—high prices predict lower returns and cheap prices higher returns—but are terrible forecasters over the next month or year.
“That is a terrible forecaster for the next month or the next year... But over the next 10 years... it’s statistically powerful. When things are expensive, you’ve made less money. When things are cheap, you’ve made more money.”
- A 50/50 portfolio of US stocks and bonds is currently the worst-valued ever: stocks (by Shiller CAPE) are more expensive than ~90% of history and bonds (by real yield) are worse than ~90% of history, and because these two are rarely both at the 90th percentile simultaneously, the combined portfolio sits at its worst level ever—priced to make about 2.5% over inflation versus a historical 5%.
“That number is more expensive in roughly 90 percent of 100 some odd plus years of history.”
- If high prices on both stocks and bonds lead to lower-than-normal future returns, then individuals and institutions who base retirement and portfolio forecasts on historical returns are using too-high assumptions—even without any crash—which quietly worsens the retirement funding problem.
“If they’re using anything like history and if we’re right that high prices on both stocks and bonds lead to lower than normal returns, it doesn’t have to be a bubble... they’re using too high of an assumption.”
- The excess return investors demand on stocks over bonds is strongly a function of the relative volatility of stocks versus bonds experienced over the prior ~20 years (a generation); people are partly prisoners of their lived market experience, which explains the empirical strength of the Fed model.
“how much excess return they need on stocks versus bonds is a function very strongly of the last 20 years’ relative volatility of stocks and bonds.”
Market anomalies like low-risk outperformance and momentum's negative skewness reflect genuine pricing errors arising from leverage constraints and behavioral biases, not rational risk.
- Contrary to the theory that risk should be rewarded, stocks that are low risk—low beta, low volatility, low leverage—outperform in a fairly strong way; this anomaly, discovered by Fisher Black in 1970 and then forgotten for 35 years, is another effective systematic form of investing.
“Stocks that are low risk, backwards to theory, right? Risk should get paid. Stocks that have low betas, low volatilities, low fundamental risks like low leverage, outperform in a fairly strong way.”
- Momentum has negative skewness—a bad left tail with standard deviations you're not supposed to see—but those momentum crashes have historically tended to occur only in strong markets, not weak ones, which is why they are less worrying than value's drawdowns that hit during the Depression and global financial crisis.
“Momentum itself... it has negative skewness... It has standard deviations you’re not supposed to see. Big events. They have tended to be more, maybe this is luck, but they have tended to only occur in strong markets, not in weak markets.”
- Because leverage is costly or constrained, conservative and constrained investors who want more return move money into fewer, higher-risk assets rather than levering a diversified portfolio; this orphans low-risk assets, making them too cheap, while concentration risk is mispriced as less dangerous than leverage risk to investors' detriment.
“He points out that low-risk assets are orphans now... those assets that are orphaned will be a little too cheap, because no one wants them. People who want to be aggressive will a little too much be willing to be undiversified.”
- Many advanced risk models invoking fifth moments, skewness, and volatility-of-volatility are utter nonsense—Ptolemaic epicycles that fit but don't explain—though coskewness is at least a sensible candidate risk factor because it captures whether very bad events for an asset coincide with very bad market events.
“a fair amount of those risk models I think are utter nonsense. I don’t think the fifth moment... is a good measure.”
Bubbles should be reserved for assets where no rational scenario supports ownership even under aggressive assumptions; conviction about certainty, not mere expensiveness, causes financial disasters.
- The most dangerous condition in finance is convincing the world that something is a certainty when it isn't; the financial crisis's single primary cause was the assumption that real estate prices can't go down, and disasters like Lehman and the Reserve Fund breaking the buck were amplified because people believed those things were safe—whereas the tech bubble's collapse was more benign precisely because no one thought tech stocks were riskless.
“any time you convince the world that something is a certainty but it’s not, it’s the most dangerous time humanly possible.”
- A bubble should be reserved for an asset where, using your best aggressive assumptions about future growth (or inflation for bonds) and the current price, you cannot construct any scenario in which a rational investor would want to own it—not merely something that looks expensive; the late-1999 tech market qualified because aggressive return assumptions still produced returns below bonds.
“A bubble to me is something still subjective... is something I’ve tried my best to come up with future assumptions of growth... and current price, and I can’t come up with assumptions that would lead any rational investor... to want to own this.”
- Despite ~2% nominal yields and bonds being more expensive than ~90% of recorded history, bonds are not a bubble because a plausible workout scenario exists—Japan—where low yields persisted and bond investors did fine; by the proof-by-contradiction standard, the existence of even one plausible workable scenario disqualifies the bubble label.
“to come up with... a reasonable scenario that could happen in the next 20 years... for a workout, I need one word. Japan. It wasn’t that hard.”
The defining risk of a portfolio is not volatility but whether losses occur when everything else in your life is also bad, making leverage less dangerous than concentration.
- A genuine measure of risk isn't whether an asset occasionally loses money, but whether it loses money precisely when everything else in your life is also bad; an asset with occasional giant losses that occur in good times is something you can live with, while one that fails when everything else fails is truly risky.
“If it’s a good measure of risk, it doesn’t just hurt occasionally. It hurts you when it hurts to be hurt.”
- In theory all investors should own the same single best portfolio of risky assets (highest return for risk) and then lever it up if aggressive or add cash if conservative, rather than shifting money from low-return to high-return assets, because shifting sacrifices diversification while leverage preserves it.
“we should all agree on the best portfolio of risky assets. We should all own the same one, because it’s the highest return for risk. Those of us who are aggressive should apply some leverage to it.”
- The most dangerous things are those that appear to protect you but only mostly do so—like money market funds that look stable because short maturities round to a dollar—because they encourage risk-blind behavior; the same logic appears in safety equipment, where helmets or seat belts may induce more aggressive behavior (risk compensation).
“I think the thing people don’t appreciate is how dangerous things are that you think protect you, but only mostly protect you.”
High-frequency trading and market competition have genuinely improved fairness and cost efficiency for retail investors by compressing spreads and eliminating economically irrational front-running.
- High-frequency trading has made markets more just and fair, especially for small investors, because competitive market-makers have driven bid-ask spreads and execution costs far lower than the old regime, and because no one bothers to front-run small-dollar investors since there is no money in it.
“High frequency trading has made the world more just and fair, particularly for small investors.”
- Wayne Gretzky succeeded despite being undersized and slower than peers because he anticipated where the puck would be while others skated to where it was—a real, almost mutant anticipatory ability rather than physical superiority.
“he skated where the puck was going to be, while everyone else was skating where the puck was. Having watched him, I just think it was true.”
Personal financial anxiety from childhood instilled an unease about risk that paradoxically fueled professional achievement but left a hard-to-turn-off vigilance.
- Growing up middle-class with a trial-attorney father whose income swung between good and zero years, and whose parents over-shared financial anxiety, instilled a sense of impending doom that paradoxically fueled achievement but left a hard-to-turn-off unease that doesn't make for a relaxed person.
“I had a sense of impending doom as a child that I think was oddly a positive for achievement... but doesn’t make you a happy, relaxed person and is impossible to turn off after it’s no longer useful.”