Cliff Asness's 10 Investing Pet Peeves, Revisited: What the Evidence Says

In January 2014 the Financial Analysts Journal published a short essay by Cliff Asness, co-founder of AQR Capital Management, called “My Top 10 Peeves.”¹ Each peeve is a belief that Asness thinks is common, often repeated, wrong or misleading, and costly to investors. Twelve years later the list still circulates, mostly as a set of quotable lines.

This guide grades each peeve against the economic theory behind it and the research published before and since. Asness asked for this kind of treatment. He called the pieces “mostly teasers,” said he would “blatantly ignore counterpoints,” and invited readers to disagree with both his arguments and whether the peeves even exist.¹ So the job here is to supply the counterpoints and see what survives.

My verdict: all ten hold up in substance, and four need a caveat. One caution before grading him on any of them. Several objections people raise against this paper are already in its footnotes. Where that is the case, I say so and quote him, because criticizing a version of the argument he did not make would be the kind of sloppy reasoning the paper is about.

The scorecard

The common thread is a habit of mind: ignore the label and work out the economics. A “permanent loss” still has a distribution of outcomes. Cash that “enters the market” changes hands. A smart beta fund is a market fund plus a bet. Holding a bond to maturity does not erase its price. The five peeves with the most direct use for a do-it-yourself investor are 1, 3, 6, 9, and 10, and they get the most space below.

1. Volatility versus “permanent loss of capital”

The claim. Many value investors say volatility is a quant’s toy and that real risk is the chance of losing money you never get back. Asness grants that this is “one fair way to think of risk” and says the critics are “not directly wrong.” His objection is that they attack a straw quant. Even the simplest quantitative framework carries two numbers: an expected return and the uncertainty around it. A stock that both camps agree is overvalued gets a negative expected return from the quant, not a label of “low risk.” Volatility, in his words, is “how much it’s likely to move versus the forecast of expected return.”¹

His best example is a cheap stock that gets cheaper with no change in fundamentals. The permanent-loss investor reports that nothing happened, since the stock is now even cheaper and will come back if held. Asness disagrees: “even if you are right, someone else now has the opportunity to buy it at an even lower price than you did. In a very real sense, you lost money; you just expect to make it back.” When AQR lost money shorting technology stocks about a year too early in 1999, “we didn’t get to report to our clients, ‘We have not lost any of your money. It’s in a bank we call “short NASDAQ.”’”¹

Theory and evidence. The split between expected return and variance is the starting point of portfolio theory.² The same year, Roy proposed “safety first,” which ranks portfolios by the probability of falling below a disaster level, a downside-focused measure close to what the permanent-loss camp means.³ Markowitz himself gave a full chapter of his 1959 book to semivariance, which counts only below-target outcomes.⁴ Quantitative finance never settled on volatility as the only measure, and Asness says so in his first footnote: symmetric volatility is “not the only weapon,” and it “would be particularly inappropriate for option-like securities with by-design asymmetric payoffs.”

The research that best supports his 1999 story is Shleifer and Vishny’s limits of arbitrage.⁵ Their model shows how a manager who is right about value can still be forced out: investors see the loss, “infer from this loss that the arbitrageur is not as competent as they previously thought,” and withdraw capital “even though the expected return from the trade has increased.” A temporary mark-to-market loss becomes permanent when the investor cannot hold through it.

Where it holds and where it needs care. Asness’s point holds. His own phrase for the dispute, “smart people talking in different languages,” fits better than calling either side wrong. Volatility is a measure of risk, and it is not a complete one. A retiree drawing income, a leveraged investor, and a fund facing redemptions can turn a drawdown into a permanent loss, while a 30-year-old saver mostly cannot. William Bernstein’s split between shallow risk (price swings) and deep risk (permanent loss of purchasing power) is a compatible way to say the same thing; we cover it in Stocks Are Always Risky.

For a DIY investor: keep expected return and risk as separate questions, and measure risk in more than one way. Volatility, maximum drawdown, time underwater, and the chance of being forced to sell all describe different ways to lose. Our portfolio risk metrics guide walks through each, and An Unrealized Loss Is Still a Loss makes Asness’s short-NASDAQ point for individual holdings.

2. Everything expensive gets called a bubble

The claim. Asness reserves “bubble” for “a price that no reasonable future outcome can justify.” Technology stocks in early 2000 qualified, he argues, short of assumptions like the companies “owning the GDP of the Earth.” Most other uses describe something “expensive,” with a “lower than normal expected return,” which is a different claim. His example was the 10-year Treasury in late 2013: priced for a low real return, but with plausible paths, including a Japan-style one, where buyers did fine.¹

Evidence. Greenwood, Shleifer, and You’s “Bubbles for Fama” is the best test of whether bubbles can be spotted from prices.⁶ They study every industry whose value-weighted return exceeded 100% over two years, in US data from 1926 to 2014 and international sectors from 1985 to 2014. Such run-ups did not, on average, predict unusually low returns. They did predict “a substantially heightened probability of a crash.” In the US data, the chance of a crash, defined as a drawdown of 40% or more within two years, was 20% after a 50% run-up, 53% after a 100% run-up, and 80% after a 150% run-up. The 40 US industries that doubled in two years since 1928 averaged a 7% gain over the next year and roughly zero over two years. Volatility, issuance, turnover, and the shape of the price path added information about which run-ups would crash.

That result supports Asness’s vocabulary. “This asset is expensive, so its expected return is lower” is one claim. “This asset has risen a lot” is a second. “This asset is likely to crash soon” is a third. Run-ups carry real information about crash risk, and they still do not translate into a reliable forecast of low average returns, which is roughly the middle ground Asness stakes out between “the word is verboten” and “everything is a bubble.” Theory explains why bubbles can last even when sophisticated traders see them: no single arbitrageur can attack alone, and they cannot coordinate when to sell.⁷

Postscript on the Treasury example. The 10-year Treasury yield was 3.04% at the end of 2013.⁸ A buyer who held that bond to maturity earned about 3% a year, while consumer prices rose 31.6% over the decade, about 2.8% a year.⁹ That left a real return near 0.25% a year, below the 1% to 2% Asness sketched, and the shortfall came from inflation. The bond never collapsed in price along the way. The test of his definition came later: in August 2020 the 10-year yield touched 0.52%, and in 2022 the aggregate bond market lost about 13%. Whether 2020 met his bar for a bubble is arguable. At 0.52%, it offered a negative real return unless inflation stayed below half a percent a year.

For a DIY investor: when someone says “bubble,” ask which of the three claims they mean. Valuation should shape your return expectations and your margin of safety. It is a poor timing signal. We cover the bubble evidence in more depth, including Bernstein’s behavioral signs, in Are We in an AI Bubble?

3. Judging strategies on three- to five-year track records

The claim. Investors hire and fire managers, strategies, and asset classes on trailing three- to five-year returns. Asness says that over those horizons “pretty much everything has shown some systematic, if certainly not dramatic, tendency to mean revert,” so buying what did well and dumping what did badly uses the data “backwards.” With a disciplined approach, he writes, “value and momentum are both good long-term strategies, but you don’t want to be a momentum investor at a value time horizon.”¹

Evidence for mean reversion. The academic record he leans on is real. De Bondt and Thaler found that extreme past losers beat extreme past winners over the following three years. Thirty-six months after formation, their loser portfolios beat the market by 19.6% and the winners lagged by about 5.0%, a 24.6-point spread, with much of it earned in Januaries.¹⁰ Fama and French estimated that predictable variation was about 40% of three- to five-year return variance for small-firm portfolios and about 25% for large-firm portfolios over 1926-1985.¹¹ Momentum runs in the other direction at shorter horizons: Jegadeesh and Titman found winners keep winning over three to twelve months, about 1% a month for six-month formation periods in 1965-1989, with part of the gain dissipating over the next two years.¹² Poterba and Summers found the same pattern of short-horizon continuation and long-horizon reversal, and warned that the tests “have little power.”¹³

Evidence on chasing. The strongest support comes from studies of real hiring and firing. Goyal and Wahal examined decisions by about 3,400 US plan sponsors from 1994 to 2003.¹⁴ Sponsors hired managers after strong excess returns, but that “does not deliver positive excess returns thereafter.” Fired managers often did fine after being fired, and “if plan sponsors had stayed with fired investment managers, their excess returns would be no different from those delivered by newly hired managers,” before counting the cost of switching. Berk and Green offer a rational version: if skilled managers face decreasing returns to scale, money flows in until their edge is gone, so past winners stop outperforming even when their skill was real.¹⁵

Our test. To check the sign of the effect directly, I used Kenneth French’s US factor returns from 1927 through August 2026: the market’s excess return, size (SMB), value (HML), and momentum.¹⁶ For each factor I compared every trailing three- and five-year return with the return over the following three or five years. Then I ran a simple chaser: each January, compare the factor with the best trailing return against the one with the worst, and check which did better next.

Seven of the eight correlations are negative, which leans Asness’s way, and all are small. Momentum is the exception at five years. The chasing test leans the other way. Across 90 to 94 January decisions, the trailing winner beat the trailing loser over the next window 52% to 53% of the time, with a median edge of about 4 percentage points over three years and 5 to 8 points over five. Most of that edge comes from ranking factors whose long-run averages differ. The market and momentum averaged 0.69% and 0.61% a month over the sample, against 0.35% for value and 0.17% for size, and they were the trailing winner most often. Ranking factors by trailing return partly ranks them by their long-run premium.

Where it needs a caveat. Within a single factor, the reversal Asness describes shows up weakly for the market, size, and value, and momentum keeps going. Across factors, the trailing record mostly restated which factors earn more over the long run. Neither result supports using the data “backwards” as a rule. A three- to five-year record carries little information beyond what the long-run evidence already says, and acting on it means trading on noise. Asness’s own phrase, “certainly not dramatic,” is the right size for the effect. In his 2024 paper he puts it as being “a (mild) contrarian” at three to five years.¹⁷

For a DIY investor: before adopting or dropping a strategy, write down the reason it should work, what it costs, and how long a bad stretch could last, using evidence that existed before the recent run. Value lagged growth for most of 2010 to 2020, and an investor who judged it on five-year windows would have sold near the bottom. We cover that stretch in Is Factor Investing Dead?

4. Looking for one culprit behind 2008

The claim. Asness says nearly everyone shares some blame for the 2007-2008 crisis: government, banks, rating agencies, the mortgage industry, and borrowers. His larger point is that the debate merges two events: a real estate and credit bubble, and a financial crisis that its collapse set off. “The collapse of the real estate/credit bubble did not have to lead directly to the financial crisis.” People who blame government are mostly explaining the bubble, people who blame banks are mostly explaining the crisis, and so they rarely engage each other.¹

Evidence. Ben Bernanke made the same distinction in 2012, separating triggers (“the proximate causes, if you will”) from vulnerabilities, “the structural, and more fundamental, weaknesses in the financial system and in regulation and supervision that served to propagate and amplify the initial shocks.”¹⁸ He used Asness’s counterexample too: the dot-com crash “destroyed as much or more paper wealth--more than $8 trillion” and produced only a short, mild recession. Research on the trigger side documents mortgage credit expanding fastest in ZIP codes with many subprime borrowers even as their relative income growth fell¹⁹ and laxer screening of loans that were easier to securitize.²⁰ Research on the amplification side covers opaque mortgage securities and runs on short-term funding.²¹ The official inquiry reflects the disagreement: the Financial Crisis Inquiry Commission’s report came with two dissents signed by four of its ten members.²²

For a DIY investor: the lesson extends past 2008. When a story names one cause for a market disaster, ask what set it off and what turned it into a crisis. Leverage and short-term funding usually decide the second part, and they are also what you control in your own finances.

5. Four phrases to retire

“It’s a stock picker’s market”

Asness doesn’t see why managers would get better at picking stocks just because the index is flat. He allows one rigorous version: when stocks move in lockstep there is little idiosyncratic volatility for anyone to exploit, so high dispersion is “a necessary (but not sufficient!) ingredient.” Later work supports that version. Von Reibnitz found that the most active funds’ outperformance over the least active is concentrated in months of high cross-sectional dispersion.²³ Sharpe’s arithmetic still applies in every market: before costs, the average actively managed dollar earns the market return, and after costs it earns less.²⁴

“Forecasting is exceptionally difficult right now”

Short-term market forecasting is always exceptionally difficult, Asness says, so the phrase implies that it is usually easy. The evidence agrees with him. Welch and Goyal found that the standard predictors of the equity premium performed poorly in and out of sample and “would not have helped an investor” time the market.²⁵ Their 2024 update with Zafirov added 29 newer predictors and data through 2021; more than a third lost significance even in-sample, and half of the rest did poorly out of sample.²⁶

“Arbitrage”

In academic finance, arbitrage means a riskless profit. In practice it often means “a trade we kind of like.” Asness accepts a middle ground: a long-short bet on closely related securities converging. Merger and convertible arbitrage can lose a lot when spreads widen before they converge, which is the Shleifer-Vishny mechanism again.⁵ “Relative-value trade” says the same thing without implying the trade is safe.

“There’s a lot of cash on the sidelines”

This is the strongest of the four, and it rests on an accounting identity. When a buyer spends cash on a stock, the seller now holds the cash. “There are no sidelines,” Asness writes: “Add us all up and there are no sidelines.” Money market funds held about $7.9 trillion in late September 2026,²⁷ and buying stocks with it would move that cash to whoever sold the stocks, leaving the total unchanged. He grants two real effects. Eager buyers can push prices up while the cash stays put, and companies can create or retire shares over time through issuance and buybacks.

For a DIY investor: treat these phrases as signs that the speaker is filling airtime. None of them should change an allocation.

6. Smart beta is active management

The claim. “If you deviate markedly from capitalization weights, you are, by definition, an active manager making bets.” Asness writes it as an equation: a smart beta portfolio equals the cap-weighted index plus the difference between the two, SB = CW + (SB - CW). The second term is a long-short portfolio, the same structure as the Fama-French value factor, HML.²⁸ Adding it to the market gives you “market exposure plus a separate bet.” Calling smart beta “a better way to get market exposure” gets the framing wrong. The bet “may be a good bet” (AQR sells several, and he discloses it), but it is still a bet.¹

Where it holds. It holds because of how aggregation works. The cap-weighted market is the one portfolio everyone can hold at once; any portfolio that differs from it needs someone on the other side.²⁴ A value tilt is a bet that cheap stocks will beat expensive ones, and someone else holds the expensive ones. Asness and John Liew made the longer argument the same year: smart beta repackages long-studied factor tilts, and they still think the good ones work.²⁹

The definitional caveat, which he already made. ETF marketing often uses “passive” to mean “tracks a published index mechanically.” Under that definition a value index fund is passive. Asness takes this up in footnote 8. He calls the index-following definition “quite odd,” since nearly any quantitative strategy can be written as an index. For a second definition, passive as low turnover, he writes “I partially withdraw my peeve,” while still calling low turnover “a very weak definition of passive.” The clean way to resolve it is to separate two questions. Is the portfolio run by rules or by discretion? Does it hold the market or tilt away from it? A factor ETF is rules-based on the first and active on the second.

For a DIY investor: judge a tilt as a tilt. Know the bet, its expected premium, and how long it can lag. Our guide Every Portfolio Is a Long-Short Portfolio shows how to measure the size of the bet with active share and tracking error.

7. Hedge funds judged against the wrong benchmark

The claim. Hedge funds are neither fully hedged nor fully long. Asness puts their long-run equity exposure at about 40% to 50%, citing a median rolling seven-year beta of 0.47. In strong years the press says they lost to the S&P 500; in down years it praises them for losing less. Both judgments ignore the partial market exposure.¹

Evidence. His 2001 paper with Robert Krail and John Liew showed that the exposure is larger than it looks.³⁰ Hedge funds hold illiquid assets whose marks lag the market, so a simple regression understated their beta. Using the CSFB/Tremont index from 1994 to 2000, the simple beta was 0.37. Adding three months of lagged market returns raised the summed beta to 0.84, and the estimated alpha went from about +2.6% a year to about -4.5% a year. Getmansky, Lo, and Makarov traced the same smoothing to illiquidity, which understates volatility and inflates Sharpe ratios.³¹ Fung and Hsieh’s seven-factor model, built from liquid market factors, “can explain up to 80 percent of monthly return variations” in diversified hedge fund portfolios.³²

Asness came back to this in 2025, arguing that some alternatives should carry a beta of 1.0, with equity futures added on top, so investors add the alternative without giving up stock exposure.³³ AQR sells products built that way, and the argument stands on the arithmetic regardless.

For a DIY investor: before paying an alternative-strategy fee, split the return into market beta, other factor exposures, and what remains. Only the remainder is worth a premium fee. Two of AQR’s own funds get that treatment in AQR Funds for Bogleheads.

8. High-frequency trading

The claim. Asness calls the backlash against HFT “massively overwrought.” HFT is mostly market making done by machines. It broke up an expensive dealer cartel, narrowed bid-ask spreads, and made trading cheaper, so small investors whose orders fit within the quoted size “are obvious winners.” He adds that HFT firms as a group make “far less money in aggregate than most would guess.” He discloses that AQR is algorithmic but nowhere near high-frequency.¹

Evidence for. Hendershott, Jones, and Menkveld used the NYSE’s staggered rollout of automated quotes in 2003 as a natural experiment. Algorithmic trading narrowed spreads and reduced adverse selection, especially for large stocks.³⁴ Brogaard, Hendershott, and Riordan found that HFT pushes prices toward efficient values, including on the most volatile days.³⁵ The Flash Crash study by Kirilenko and coauthors found that the most active high-frequency traders did not change their trading pattern as prices fell on May 6, 2010.³⁶

Evidence for a caveat. Budish, Cramton, and Shim showed that continuous trading creates brief, mechanical arbitrage opportunities “available to whomever is fastest,” which sets off an arms race in speed that investors pay for through wider spreads.³⁷ Aquilina, Budish, and O\u2019Neill then measured it using London Stock Exchange message data. Latency-arbitrage races occurred about once a minute per FTSE 100 stock, accounted for about 20% of trading volume, and acted like a 0.5 basis point tax on trading, roughly $5 billion a year across global equity markets.³⁸ Six firms won more than 80% of the races.

Those figures fit both sides. A half basis point is far smaller than the spreads HFT replaced, which supports Asness. A few billion dollars a year spent on speed that comes from how exchanges match orders is a real cost, and a market-design fix such as frequent batch auctions could remove it. Larry Harris, whom Asness cites, takes the same split view: HFT lowers costs, and parts of it still deserve regulatory attention.³⁹

For a DIY investor: very little. Asness’s advice holds: if trading against fast traders worries you, buy an index fund and trade rarely. Use limit orders, and avoid trading at the moment news breaks.

9. Buybacks that “prevent dilution”

The claim. Companies issue stock options to executives, then buy back shares so shareholders own the same fraction of the company as before. Asness calls this “idiocy.” Shareholders do own the same fraction, “So what? They now own as much of a company that happens to have a bunch less cash.” The buyback hides the cost of the options without removing it. He made the longer case in 2004, before option expensing was required.⁴⁰ The accounting rule he refers to, FAS 123(R), was issued in December 2004 and took effect for larger companies in fiscal years starting after June 15, 2005.⁴¹¹

Evidence. Repurchases do track option activity. Kahle found that firms announce buybacks when executives hold many options and when employees hold many exercisable options, and that the market reacts less positively to buybacks at firms with many employee options outstanding.⁴² Bens, Nagar, Skinner, and Wong found firms repurchase more when options dilute diluted EPS and when earnings fall short of EPS growth targets.⁴³ An earlier Federal Reserve study estimated that S&P 500 buybacks cut shares outstanding by about 2% a year in the late 1990s, but that only about half of the repurchased shares were retired on net because of option exercises.⁴⁴

Where it needs a caveat. Asness writes that the only rational reason to buy back stock is that management thinks it is undervalued. Footnote 20 adds a second, raising leverage. That still leaves out the most common reason. A mature company with more cash than good projects can return it to shareholders through a buyback at a fair price, and repurchases are more flexible than dividends and defer taxes for shareholders who don’t sell. Executives surveyed by Brav, Graham, Harvey, and Michaely described repurchases as a flexible use of residual cash after investment, alongside timing the market and raising EPS.⁴⁵ So the narrow claim stands: a buyback done to offset option dilution does not make the options free. The broad claim, that undervaluation is the only good reason to repurchase, goes too far.

Since 2023, US companies have paid a 1% excise tax on net repurchases. The tax base nets out the value of stock issued during the year, including stock issued to employees, so the tax code measures buybacks net of issuance.⁴⁶

For a DIY investor: headline buyback figures are gross. S&P 500 companies repurchased a record $942.5 billion of stock in 2024,⁴⁷ and the number says nothing about how much stock they issued to employees. Look at the change in diluted share count, or net payout yield, instead. Our shareholder yield guide has a net payout calculator for this.

10. Bonds have prices too

The claim. “You should own bonds directly, not bond funds, because bond funds can fall in value but you can always hold a bond to maturity and get your money back.” Asness, who started his career in fixed income, calls this perhaps the least harmful peeve and perhaps the most annoying. A bond fund is a portfolio of bonds marked to market every day. When rates rise, the individual bond falls in value just like the fund; its price simply isn’t on the statement. You can sell the fund after a drop, buy the individual bonds it held, and hold those to maturity, so the two choices are “just the same thing.”¹

He adds that individual-bond investors usually reinvest maturing bonds in new long bonds through a ladder. The ladder as a whole “never itself matures,” which is the same complaint they make about funds.

Evidence. The arithmetic is not in dispute. TreasuryDirect explains that a security’s price moves opposite to market yields even though it pays par at maturity.⁴⁸ Vanguard calls the advantage of individual bonds “the ‘principal at maturity’ myth” and says holding to maturity “primarily provides an emotional rather than economic benefit.”⁴⁹ 2022 gave the plainest test in decades. The US aggregate bond market lost about 13%, with the Bloomberg US Aggregate index down 13.01% and AGG’s NAV down 13.06%,⁵⁰ and individual bonds with similar maturities lost similar amounts in market value, whether or not anyone saw a price.

Footnote 21 lists the good reasons to choose one or the other: taxes, commissions and spreads, fees, diversification (which favors funds for bonds that can default), active management, and the risk of trading costs caused by other fund investors.

Where it needs a caveat. The footnote leaves out liability matching. If you need a known amount on a known date, a Treasury or TIPS maturing on that date removes interest-rate risk relative to that need, because its duration shrinks to zero as the date approaches. A conventional bond fund keeps a roughly constant duration. Campbell and Viceira show that for a conservative long-horizon investor, long inflation-indexed bonds are close to the riskless asset.⁵¹ That reason for an individual bond is about matching cash flows to spending. The mark-to-market loss is still real; what changes is that the bond’s value moves with the value of the liability. Defined-maturity bond ETFs do the same job in fund form. We cover building one in How to Build a TIPS Ladder.

For a DIY investor: choose between bonds and bond funds on cost, taxes, diversification, and whether you are matching a dated liability. Do not choose individual bonds to avoid seeing losses. Our Austria century bond guide shows what holding to maturity does and does not protect, with a calculator.

What twelve years added

• Bonds: 2022’s rate shock tested peeve 10 on millions of investors at once. Fund holders saw the loss; individual-bond holders had the same loss without the statement line.
• Factor patience: value’s long lag through 2020 and its rebound in 2021-2022 played out peeve 3 at exactly the horizon Asness warned about.
• HFT: the latency-arbitrage studies put a size on the cost of the speed race. It is small per trade and real in total, which leaves Asness mostly right and gives the critics one concrete point.
• Buybacks: Congress taxed net repurchases in 2022, and its net-of-issuance base matches Asness’s point about employee stock.
• Market efficiency: Asness’s 2024 paper “The Less-Efficient Market Hypothesis” argues that relative stock prices have become less efficient since about 1990, with the value spread returning to dot-com levels in 2019-2020 and staying wide longer.¹⁷ He concludes value “should be more lucrative for those who can stick with it over the long-term, but also harder to stick with,” and cites Peeve 3 directly. My inference, which he does not state, is that judging on three- to five-year windows has become more costly.

Key takeaways

• Asness’s ten peeves hold up in substance after twelve years. Four need caveats: trailing-return chasing, HFT, buyback motives, and liability matching with individual bonds.
• Volatility and permanent loss are two ways to describe risk. Forced selling, leverage, and withdrawals are what turn the first into the second.
• In US factor data from 1927 to 2026, the factor with the best trailing three- or five-year return beat the worst one over the next window 52% to 53% of the time, mostly because factors with higher long-run averages tend to lead. The trailing record added little beyond that.
• A factor or smart beta fund is the market plus a long-short bet. Judge the bet on its own.
• Buybacks that offset option grants do not make the grants free. Track the net change in share count, since gross buyback totals leave out stock issued to employees.
• An individual bond loses value when rates rise, the same as a fund. Individual bonds earn their place by matching dated spending; holding to maturity does not hide a loss.

Frequently asked questions

What are Cliff Asness’s 10 pet peeves?

From his 2014 Financial Analysts Journal essay: (1) dismissing volatility in favor of “permanent loss of capital,” (2) calling every expensive asset a bubble, (3) judging strategies on three- to five-year returns, (4) blaming one culprit for the 2008 crisis, (5) the phrases “stock picker’s market,” “forecasting is exceptionally difficult right now,” loose “arbitrage,” and “cash on the sidelines,” (6) calling smart beta passive, (7) misjudging hedge funds against the wrong benchmark, (8) demonizing high-frequency trading, (9) buybacks to “prevent dilution” from options, and (10) believing individual bonds avoid the price risk of bond funds.¹

Is volatility a good measure of risk?

It is a useful one and an incomplete one. Volatility measures how far results can stray from what you expect. It misses asymmetric payoffs, drawdown length, and forced selling, which is why downside measures date back to Roy and Markowitz in the 1950s.^{3, 4} Use it alongside drawdown and liquidity needs.

Is there really no cash on the sidelines?

For investors as a group, there isn’t. Every share bought is sold by someone, so buying stocks moves cash from one holder to another. Prices can rise when buyers are more eager than sellers, and companies can change the number of shares through issuance and buybacks, but the stock of cash is not “waiting” to enter the market.

Is a smart beta ETF active or passive?

It is usually passive in implementation and active in exposure. It follows rules mechanically, but its holdings differ from market weights, which is an active bet relative to the market portfolio.²⁹

Are individual bonds safer than bond funds?

No, not for the same credit quality and duration. Both fall in value when rates rise. Individual bonds can be the better choice for matching a known future expense, or for tax or cost reasons, but holding to maturity does not undo a price drop.⁴⁹

Should I switch funds after five years of bad returns?

Not on the returns alone. Plan sponsors who fired managers after poor returns and hired recent winners did no better afterward,¹⁴ and in our factor test the trailing winner beat the trailing loser 52% to 53% of the time, mostly because of differences in long-run factor premiums. Switch if the reasoning, the cost, or your own needs changed.

Related guides

• Every Portfolio Is a Long-Short Portfolio: measuring the active bet in any fund.
• Shareholder Yield: dividends plus buybacks minus issuance.
• Are We in an AI Bubble?: what bubble research can and cannot tell you.
• Is Factor Investing Dead?: why factors can lag for a decade.
• Austria’s 100-Year Bond: what holding to maturity protects.
• Stocks Are Always Risky: shallow risk versus deep risk.
• AQR Funds for Bogleheads: low-beta alternatives in a simple portfolio.

Sources

1. Asness, C. S. (2014). My Top 10 Peeves. Financial Analysts Journal 70(1): 22-30. The source of every quotation from Asness in this guide unless noted.
2. Markowitz, H. (1952). Portfolio Selection. Journal of Finance 7(1): 77-91. Expected return and variance as separate inputs.
3. Roy, A. D. (1952). Safety First and the Holding of Assets. Econometrica 20(3): 431-449. Minimizing the probability of a disaster outcome.
4. Markowitz, H. (1959). Portfolio Selection: Efficient Diversification of Investments. Yale University Press (Cowles Foundation Monograph 16). Chapter IX covers semi-variance.
5. Shleifer, A., & Vishny, R. W. (1997). The Limits of Arbitrage. Journal of Finance 52(1): 35-55. Performance-based withdrawals can force arbitrageurs out of trades whose expected return has risen.
6. Greenwood, R., Shleifer, A., & You, Y. (2019). Bubbles for Fama. Journal of Financial Economics 131(1): 20-43. Crash probabilities by run-up size (Table 3) and returns after the 40 US episodes.
7. Abreu, D., & Brunnermeier, M. K. (2003). Bubbles and Crashes. Econometrica 71(1): 173-204. Bubbles persist because arbitrageurs cannot coordinate their selling.
8. U.S. Department of the Treasury. Daily Treasury Par Yield Curve Rates. 10-year yield of 3.04% on December 31, 2013 and 0.52% on August 4, 2020.
9. U.S. Bureau of Labor Statistics. CPI for All Urban Consumers (CPI-U), series CUUR0000SA0. 233.049 in December 2013 and 306.746 in December 2023.
10. De Bondt, W. F. M., & Thaler, R. (1985). Does the Stock Market Overreact?. Journal of Finance 40(3): 793-805. Loser and winner portfolios 36 months after formation (p. 799).
11. Fama, E. F., & French, K. R. (1988). Permanent and Temporary Components of Stock Prices. Journal of Political Economy 96(2): 246-273. Predictable share of 3- to 5-year return variance, 1926-1985.
12. Jegadeesh, N., & Titman, S. (1993). Returns to Buying Winners and Selling Losers: Implications for Stock Market Efficiency. Journal of Finance 48(1): 65-91. Momentum over 3- to 12-month horizons, 1965-1989.
13. Poterba, J. M., & Summers, L. H. (1988). Mean Reversion in Stock Prices: Evidence and Implications. Journal of Financial Economics 22(1): 27-59. Short-horizon continuation, long-horizon reversal, and the low power of the tests.
14. Goyal, A., & Wahal, S. (2008). The Selection and Termination of Investment Management Firms by Plan Sponsors. Journal of Finance 63(4): 1805-1847. About 3,400 plan sponsors, 1994-2003.
15. Berk, J. B., & Green, R. C. (2004). Mutual Fund Flows and Performance in Rational Markets. Journal of Political Economy 112(6): 1269-1295. Flows compete away the returns to skill.
16. French, K. R. Data Library. Fama-French research factors and momentum factor, monthly, built from the 202608 CRSP database. Summitward’s calculation; the script and output are in our public research repository.
17. Asness, C. S. (2024). The Less-Efficient Market Hypothesis. Journal of Portfolio Management 51(1). Value spreads since 1990 and the case that relative prices have become less efficient.
18. Bernanke, B. S. (2012, April 13). Some Reflections on the Crisis and the Policy Response. Speech at the Russell Sage Foundation and The Century Foundation conference, New York. Triggers versus vulnerabilities; the dot-com comparison.
19. Mian, A., & Sufi, A. (2009). The Consequences of Mortgage Credit Expansion: Evidence from the U.S. Mortgage Default Crisis. Quarterly Journal of Economics 124(4): 1449-1496.
20. Keys, B. J., Mukherjee, T., Seru, A., & Vig, V. (2010). Did Securitization Lead to Lax Screening? Evidence from Subprime Loans. Quarterly Journal of Economics 125(1): 307-362.
21. Gorton, G. (2009). The Subprime Panic. European Financial Management 15(1): 10-46. Opacity of mortgage securities and the run on short-term funding.
22. Financial Crisis Inquiry Commission (2011). The Financial Crisis Inquiry Report. Majority report with a joint dissent (Hennessey, Holtz-Eakin, Thomas) and a separate dissent (Wallison).
23. von Reibnitz, A. (2017). When Opportunity Knocks: Cross-Sectional Return Dispersion and Active Fund Performance. Critical Finance Review 6(2): 303-356.
24. Sharpe, W. F. (1991). The Arithmetic of Active Management. Financial Analysts Journal 47(1): 7-9.
25. Welch, I., & Goyal, A. (2008). A Comprehensive Look at the Empirical Performance of Equity Premium Prediction. Review of Financial Studies 21(4): 1455-1508.
26. Goyal, A., Welch, I., & Zafirov, A. (2024). A Comprehensive 2022 Look at the Empirical Performance of Equity Premium Prediction. Review of Financial Studies 37(11): 3490-3557. 29 newer predictors, data through 2021.
27. Investment Company Institute. Money Market Fund Assets. Weekly data; total money market fund assets of $7.89 trillion for the week ended September 30, 2026.
28. Fama, E. F., & French, K. R. (1993). Common Risk Factors in the Returns on Stocks and Bonds. Journal of Financial Economics 33(1): 3-56. The HML long-short value factor.
29. Asness, C., & Liew, J. (2014, September). Smart Beta: Not New, Not Beta, Still Awesome. Institutional Investor.
30. Asness, C., Krail, R., & Liew, J. (2001). Do Hedge Funds Hedge?. Journal of Portfolio Management 28(1): 6-19. Contemporaneous versus summed betas, CSFB/Tremont index, 1994-2000.
31. Getmansky, M., Lo, A. W., & Makarov, I. (2004). An Econometric Model of Serial Correlation and Illiquidity in Hedge Fund Returns. Journal of Financial Economics 74(3): 529-609.
32. Fung, W., & Hsieh, D. A. (2004). Hedge Fund Benchmarks: A Risk-Based Approach. Financial Analysts Journal 60(5): 65-80.
33. Asness, C. (2025, March 28). Should Hedge Funds Hedge? Why Some Alts Should Have a Beta of 1.0. AQR Perspectives.
34. Hendershott, T., Jones, C. M., & Menkveld, A. J. (2011). Does Algorithmic Trading Improve Liquidity?. Journal of Finance 66(1): 1-33. The 2003 NYSE autoquote rollout as a natural experiment.
35. Brogaard, J., Hendershott, T., & Riordan, R. (2014). High-Frequency Trading and Price Discovery. Review of Financial Studies 27(8): 2267-2306.
36. Kirilenko, A., Kyle, A. S., Samadi, M., & Tuzun, T. (2017). The Flash Crash: High-Frequency Trading in an Electronic Market. Journal of Finance 72(3): 967-998.
37. Budish, E., Cramton, P., & Shim, J. (2015). The High-Frequency Trading Arms Race: Frequent Batch Auctions as a Market Design Response. Quarterly Journal of Economics 130(4): 1547-1621.
38. Aquilina, M., Budish, E., & O’Neill, P. (2022). Quantifying the High-Frequency Trading “Arms Race”. Quarterly Journal of Economics 137(1): 493-564. London Stock Exchange message data.
39. Harris, L. (2013). What to Do about High-Frequency Trading. Financial Analysts Journal 69(2): 6-9.
40. Asness, C. S. (2004). Stock Options and the Lying Liars Who Don’t Want to Expense Them. Financial Analysts Journal 60(4): 9-14.
41. Financial Accounting Standards Board (2004, December 16). FASB Issues Final Statement on Accounting for Share-Based Payment. Statement 123(R).
42. Kahle, K. M. (2002). When a Buyback Isn’t a Buyback: Open Market Repurchases and Employee Options. Journal of Financial Economics 63(2): 235-261.
43. Bens, D. A., Nagar, V., Skinner, D. J., & Wong, M. H. F. (2003). Employee Stock Options, EPS Dilution, and Stock Repurchases. Journal of Accounting and Economics 36(1-3): 51-90.
44. Liang, J. N., & Sharpe, S. A. (1999). Share Repurchases and Employee Stock Options and Their Implications for S&P 500 Share Retirements and Expected Returns. Federal Reserve Board, Finance and Economics Discussion Series 1999-59.
45. Brav, A., Graham, J. R., Harvey, C. R., & Michaely, R. (2005). Payout Policy in the 21st Century. Journal of Financial Economics 77(3): 483-527. Survey of 384 executives plus 23 interviews.
46. Congressional Research Service. The 1% Excise Tax on Stock Repurchases (Buybacks). IRC section 4501; repurchases after December 31, 2022, netted against stock issued during the year, including to employees.
47. S&P Dow Jones Indices (2025, March 19). S&P 500 Q4 2024 Buybacks. 2024 buybacks of $942.5 billion, gross of issuance.
48. TreasuryDirect. Understanding Pricing and Interest Rates. U.S. Department of the Treasury.
49. Vanguard. Bonds versus bond funds. Section titled “The ‘principal at maturity’ myth.”
50. iShares. iShares Core U.S. Aggregate Bond ETF (AGG) fact sheet. Calendar-year 2022 returns for the fund NAV and the Bloomberg US Aggregate Bond Index.
51. Campbell, J. Y., & Viceira, L. M. (2001). Who Should Buy Long-Term Bonds?. American Economic Review 91(1): 99-127.