Quality Factor Investing: The One That Survived

Most factors that get discovered in an academic paper quietly die the moment real money tries to trade them. Quality is the rare exception, and understanding exactly why it has held up tells you more about building a durable portfolio than almost any other single lesson in factor investing.

Why the Factor Zoo Made “Quality” a Question Worth Asking

By the mid-2010s, academic finance had a credibility problem it mostly created for itself. Researchers had published somewhere north of 300 distinct return-predicting signals — accruals, idiosyncratic volatility, share issuance, distress risk, dozens of variations on momentum, and plenty of variables nobody outside a working paper had ever heard of. Financial economists started calling this pile of signals the “factor zoo,” a phrase that stuck because it captured the absurdity: you cannot build an investment philosophy around 300 things that are each supposedly worth a few percentage points of extra return a year, especially when many of them are correlated, some are just data mining with extra steps, and a handful only worked because a researcher tried forty specifications and reported the one that worked.

In 2016, Tobias Moskowitz and colleagues, and separately Kewei Hou, Chen Xue, and Lu Zhang in a widely cited replication effort, ran a huge share of the zoo back through cleaner data and tighter statistical standards. A large fraction of the published anomalies failed to replicate at conventional significance levels once you accounted for multiple testing. Earlier, Jonathan Lewellen, along with R. David McLean and Jeffrey Pontiff in their 2016 Journal of Finance study “Does Academic Research Destroy Stock Return Predictability?”, had already shown something almost as damaging to factor believers: even the anomalies that did replicate tended to lose roughly half their return once they were published, presumably because traders read the paper and arbitraged the edge away, or because the original result was partly a fluke of the sample period.

Quality-related signals — profitability, low leverage, earnings stability, and the composite “Quality Minus Junk” factor formalized by Clifford Asness, Andrea Frazzini, and Lasse Pedersen in their 2019 paper of the same name — are among the few that came through this stress test looking intact. That doesn’t mean quality investing is a free lunch. It means the premium has proven durable enough, across different data sets, time periods, and research teams, that treating it as noise requires ignoring a genuinely large and repeated body of evidence. Heading into 2026, with a market still digesting an unusually long stretch of AI-driven capital spending, elevated valuations in mega-cap technology, and a Federal Reserve that spent 2022 through 2024 relearning what happens to over-levered balance sheets when rates rise fast, the practical case for quality as a standing allocation — not a tactical bet — has arguably gotten stronger, not weaker.

What “Quality” Actually Measures Inside a Factor Portfolio

The word “quality” gets used loosely in ordinary investing conversation — a “quality company” might just mean one an investor likes. In factor construction, it has to mean something narrower and more measurable, because an index or ETF can’t rely on judgment calls. Nearly every institutional quality methodology, whether it’s MSCI’s, S&P’s, or AQR’s QMJ framework, converges on three pillars.

Profitability

This is usually the largest weight in a quality score. Common inputs are return on equity (ROE), return on assets (ROA), gross profitability (gross profit divided by total assets, a measure championed by Robert Novy-Marx’s 2013 research showing it predicted returns about as well as book-to-value), and cash-flow-to-assets. The intuition is straightforward: a company that reliably converts revenue into profit without heavy financial engineering is doing something structurally right, whether that’s pricing power, a durable cost advantage, or simple operational discipline.

Earnings Stability

A firm with a 22% ROE built on smooth, predictable earnings is a very different animal from one with the same average ROE built on wild year-to-year swings. Quality methodologies penalize earnings variability — measured as the standard deviation of year-over-year earnings growth, or sometimes the variability of return on equity itself over a five- to ten-year lookback. Stability matters because it lowers the odds of a nasty surprise that forces a stock out of an index or triggers a dividend cut.

Balance Sheet Strength (Low Leverage)

The third pillar penalizes debt. Metrics include debt-to-equity, debt-to-assets, or sometimes interest coverage. This is the piece that most directly explains why quality tends to hold up in credit-stress environments: a company financed conservatively doesn’t need to refinance into a hostile bond market, doesn’t face covenant breaches, and isn’t forced into dilutive equity raises when liquidity tightens.

Some frameworks add a fourth screen for accounting quality or “earnings persistence,” penalizing firms whose reported earnings diverge sharply from operating cash flow — a red flag popularized by accounting researcher Richard Sloan’s work on the accrual anomaly. AQR’s QMJ model folds growth and payout behavior in as well, on the theory that quality firms grow profits steadily and don’t need to dilute shareholders to fund that growth.

How an Index Provider Turns These Ideas Into a Rules-Based Portfolio

Knowing the concepts is one thing; building a tradable index from them is another. MSCI’s USA Sponsored Quality Index — the benchmark behind iShares’ QUAL — is a useful, fully documented example of the mechanics.

MSCI starts with the broad MSCI USA parent index, then scores every constituent on three variables: return on equity (trailing), debt-to-equity, and earnings variability (the standard deviation of year-over-year EPS growth over five years). Each raw metric is converted into a z-score — how many standard deviations above or below the market average a company sits — winsorized to control for outliers, then the three z-scores are combined into a single composite quality score, typically with profitability given roughly equal or slightly higher weight than the other two pillars depending on the exact vintage of the methodology. Stocks are ranked by composite score, and the index is built to capture roughly the top third of the parent universe by float-adjusted market cap, subject to sector-neutrality constraints so the index doesn’t just become an unintentional bet on, say, software companies over utilities. Constituents are weighted by the product of their quality score and their market capitalization, then the index is rebalanced semiannually, in May and November.

S&P’s methodology for the S&P 500 Quality Index (which underlies Invesco’s SPHQ) uses a similar three-legged stool — ROE, accruals ratio (a proxy for earnings quality relative to cash flow), and financial leverage — but selects only the top 100 stocks from the S&P 500 by composite score rather than a broader third of the market, which makes SPHQ a more concentrated, higher-conviction quality bet than QUAL’s broader sleeve. Active or semi-active approaches, like Vanguard’s VFQY or JPMorgan’s JQUA, use analogous fundamental screens but layer in a manager or model that can deviate from strict mechanical rebalancing dates, generally at a lower turnover cost but with slightly less transparency about exactly when and why a name is added or dropped.

The practical upshot for an investor comparing products: two funds both labeled “quality” can hold meaningfully different stocks, because “top third of a broad universe, sector-neutral” and “top 100 of the S&P 500, concentrated” are structurally different baskets even when both start from the same three fundamental ideas.

Quality’s Relationship to Value, Momentum, and Low Volatility

Quality doesn’t live in isolation. It has a specific, fairly consistent correlation structure with the other major equity factors, and that structure is exactly why many multi-factor portfolios pair quality with something else rather than running it alone.

Quality and value are, on average, mildly negatively correlated. Expensive, high-growth compounders often score well on quality but poorly on classic value screens (low price-to-book, low price-to-earnings), while a statistically “cheap” stock is frequently cheap precisely because its earnings are unstable or its balance sheet is stretched — the exact traits quality penalizes. This negative correlation is the reason a combined “quality-value” sleeve, sometimes marketed as looking for “quality at a reasonable price,” tends to smooth the ride better than either factor run in isolation.

Quality and momentum tend to show a small positive correlation, because markets frequently reward improving fundamentals with price strength before that improvement becomes obvious in a trailing accounting screen. Quality and low-volatility strategies overlap the most of any factor pair — both tend to favor large, stable, defensively positioned businesses — which is useful to know because it means combining a quality fund and a low-volatility fund in the same portfolio buys you less true diversification than the fund names might suggest.

The practical lesson: a portfolio holding both a quality tilt and a low-volatility tilt is making a bigger, more concentrated bet on “stable, defensive, bond-like equities” than either label alone implies, and that concentration shows up specifically in periods when interest rates move sharply — which brings us to quality’s real weakness.

When Quality Underperforms: The Junk-Rally Problem and Rate Sensitivity

However, “survived” doesn’t mean “always wins,” and treating quality as a factor that never has a bad year would be its own mistake. Quality has two well-documented weak spots.

The first is what practitioners call a junk rally: sharp, early-cycle recoveries where the most beaten-down, highly levered, lowest-quality companies bounce hardest because they were priced for bankruptcy and didn’t go bankrupt. The initial recovery off the March 2020 low is the textbook case — heavily shorted, debt-laden airlines, cruise lines, and energy names posted some of the largest percentage gains of any group in the following several months, while high-quality, already-expensive compounders participated in the rally but lagged the junk cohort on a relative basis for a stretch. A quality tilt, by design, systematically underweights exactly the kind of stock that leads a junk rally.

The second weak spot is duration-like sensitivity to interest rates. Because quality (and low-volatility) portfolios skew toward large, stable, often slower-growing businesses whose cash flows are relatively predictable and far out in time, they behave a bit like long-duration bonds when rates move. In 2022, as the Fed raised its policy rate from near zero to over 4% inside twelve months, quality strategies that were also expensive on traditional valuation multiples took a real hit alongside long-duration growth stocks, even though their balance sheets were exactly what you’d want in a tightening cycle. Low leverage protects a company’s solvency; it does not automatically protect its stock price from a valuation multiple compressing because the discount rate went up.

Neither weak spot invalidates the factor. They just mean quality is best understood as a structural tilt with a specific personality — defensive, balance-sheet-focused, prone to lag in junk rallies and in sharp-rate-shock years — rather than a strategy that’s supposed to top every one-year leaderboard.

Notice the pattern across the two credit-driven shocks (2008 and 2020): quality’s balance-sheet screen provided real downside cushioning. In the 2022 rate shock, the gap narrows, because that drawdown was driven by valuation compression rather than solvency fear — exactly the scenario where quality’s leverage advantage matters less and its duration-like profile matters more.

A Worked Example: Scoring Two Companies and Sizing a Quality Tilt

Composite scoring is easier to internalize with numbers than with definitions. Imagine two hypothetical mid-cap industrial companies, “SteadyCo” and “LeverCo,” both trading at a similar market capitalization of roughly $8 billion.

On the surface, LeverCo’s headline ROE of 19.8% looks nearly as good as SteadyCo’s 21.4%. A screen that stopped at ROE alone would treat these two companies as close cousins. Once you convert each metric into a market-relative z-score — say the universe average debt-to-equity is 0.90x with a standard deviation of 0.65x, and the universe average earnings volatility is 8.5 points with a standard deviation of 5.1 points — the picture changes sharply.

SteadyCo’s leverage z-score comes out around +0.54 (better than average, since lower leverage is favorable and (0.90 − 0.55) / 0.65 ≈ 0.54), while LeverCo’s comes out around −1.85 ((0.90 − 2.10) / 0.65 ≈ −1.85). On earnings stability, SteadyCo scores roughly +0.84 ((8.5 − 4.2) / 5.1 ≈ 0.84) versus LeverCo’s roughly −1.78 ((8.5 − 17.6) / 5.1 ≈ −1.78). Even with roughly comparable profitability z-scores of, say, +0.60 for SteadyCo and +0.45 for LeverCo, the composite (a simple average of the three) lands at about +0.66 for SteadyCo versus about −1.06 for LeverCo. On a rules-based index, that gap is large enough to place SteadyCo comfortably inside the top-quality tercile and to exclude LeverCo entirely, despite the two firms looking almost identical on the metric most casual investors check first.

Now scale that logic to a portfolio. Suppose an investor holds a $250,000 equity allocation and decides to route 20% of it, or $50,000, into a quality-factor ETF rather than a plain S&P 500 fund, keeping the remaining 80% broadly diversified. If the quality sleeve had compounded at an illustrative 12.5% annualized versus 11.8% for the broad market over a hypothetical decade — a gap broadly consistent with QUAL’s live track record against the S&P 500 since its 2013 inception, though the actual annual spread has varied considerably year to year and was occasionally negative — the $50,000 quality sleeve would grow to roughly $163,700 versus about $155,900 in a plain index fund, a difference of roughly $7,800 on that slice alone, before accounting for the modestly higher expense ratio (0.15% versus roughly 0.03% for a plain-vanilla S&P 500 fund) that a quality ETF typically carries.

Quality Factor ETFs Compared

Expense ratios and inception dates reflect published fund documentation and are subject to change; verify current figures on the issuer’s fact sheet before investing.

Common Mistakes Investors Make With Quality Factor Investing

Treating “quality” and “safe” as synonyms. Quality reduces balance-sheet and earnings-stability risk. It does nothing to reduce valuation risk. A high-quality company purchased at 40 times earnings can still fall 40% if growth disappoints and the multiple compresses — 2022 proved that in painful detail.

Doubling up on quality and low-volatility without realizing the overlap. Because the two factors share a large amount of common exposure to big, stable, defensive names, stacking a quality fund and a minimum-volatility fund often just concentrates a portfolio further into the same handful of mega-cap staples and health-care names rather than adding a second, distinct source of diversification.

Judging the factor on a one- or two-year window. Quality’s edge, like every factor’s, shows up over full market cycles that include both credit-stress periods (where it should shine) and junk rallies or momentum-driven bull runs (where it should lag). Abandoning a quality allocation after a single disappointing year — 2020’s early recovery months, for instance — is a classic way to sell low right before the factor’s next strong stretch.

Ignoring sector concentration inside “quality.” Because technology and health-care firms often screen well on profitability and balance-sheet metrics, quality indexes can end up with meaningful sector tilts relative to the broad market. An investor layering a quality ETF on top of an existing large-cap growth-heavy portfolio may be adding far less true diversification than the fund’s factor label implies.

Paying active-fund fees for what is functionally a rules-based screen. Several quality products charge north of 0.25% for a strategy whose economic logic is close to fully codified in academic literature and reproducible in a low-cost index fund at half that price. The fee gap compounds meaningfully over a multi-decade holding period.

A Practical Checklist Before Adding a Quality Tilt

• Confirm which three or four metrics the fund actually screens on — ROE, leverage, earnings stability, accruals — and read the index methodology document, not just the marketing page.
• Check overlap with any existing low-volatility, dividend-growth, or large-cap growth holdings before assuming a quality fund adds real diversification.
• Review sector weights against the broad market to understand what implicit bets you’re taking (quality funds frequently run overweight technology and health care, underweight utilities and energy).
• Compare expense ratios across at least three quality products — the spread between 0.12% and 0.29% adds up meaningfully over 20-plus years of compounding.
• Decide on a holding horizon of at least one full market cycle (generally five-plus years) before evaluating whether the tilt “worked,” since single-year comparisons are dominated by noise.
• Size the position as a tilt, not a replacement — most practitioner research suggests a 10% to 30% factor sleeve layered on a core diversified holding, rather than an all-or-nothing bet.
• Revisit the position after any large rate move; quality’s duration-like sensitivity means a sharp rate shock changes its near-term risk profile even though the underlying balance sheets haven’t changed.

Key Takeaways

• Quality factor investing systematically favors companies with high, stable profitability and low leverage — it is one of a small number of published factors that has held up in post-publication and out-of-sample testing, unlike most of the 300-plus signals in finance’s so-called “factor zoo.”
• Index construction typically combines return on equity, debt-to-equity, and earnings-growth variability into a composite z-score, though exact metrics and concentration differ meaningfully between providers such as MSCI and S&P.
• Quality has historically cushioned drawdowns during credit-stress shocks (2008, 2020) more effectively than during valuation-driven, rate-shock declines (2022), because its edge comes from balance-sheet strength, not immunity to multiple compression.
• Quality tends to lag during junk rallies, when the most levered, previously distressed companies bounce hardest, and it overlaps heavily with low-volatility strategies, which limits its diversification benefit when the two are combined.
• Low-cost, transparent quality ETFs exist at expense ratios between roughly 0.12% and 0.15%; investors should be cautious about paying meaningfully more for a strategy whose logic is largely public and reproducible.

Frequently Asked Questions

Is quality factor investing the same as buying “blue chip” stocks?

Not exactly. Blue-chip investing is an informal, reputation-based judgment about well-known, established companies. Quality factor investing is a rules-based, quantitative screen on specific metrics — return on equity, leverage, and earnings stability — that can include lesser-known mid-cap names that pass the screen and exclude some famous large-cap names that don’t, if their balance sheet or earnings volatility disqualifies them.

How is the quality factor different from the value factor?

Value looks at price relative to fundamentals (low price-to-book or price-to-earnings, for example), while quality looks at the fundamentals themselves, independent of price. A stock can be high quality and expensive, high quality and cheap, low quality and expensive, or low quality and cheap. The two factors are typically mildly negatively correlated, which is why many practitioners combine them.

Why did quality underperform during parts of 2020 and 2021?

The sharp initial recovery off the March 2020 low favored heavily discounted, highly leveraged companies that had been priced for potential bankruptcy and then didn’t fail — a classic junk rally. Quality strategies, by design, underweight exactly that kind of stock, so they participated in the broader recovery but lagged the most speculative names for a stretch.

What’s a reasonable expense ratio for a quality factor ETF?

Among widely available U.S. quality ETFs, expense ratios generally range from about 0.12% to 0.29%. Since the underlying methodology is largely public and mechanical, there’s a reasonable argument for favoring funds at the lower end of that range unless an actively managed alternative has a clearly differentiated, well-documented process.

Can quality factor investing replace a core index fund entirely?

Most factor researchers, including the authors of the original QMJ paper, frame quality as a tilt layered on top of a diversified core holding rather than a full replacement for broad market exposure. A concentrated, sector-tilted quality-only portfolio takes on idiosyncratic risks — including the sector concentration and rate sensitivity described above — that a broadly diversified core position is designed to avoid. Investors weighing this trade-off alongside other portfolio construction basics may find it useful to first review some of the more common index fund mistakes to avoid before layering on any factor tilt.

Does quality factor investing work outside U.S. large-cap stocks?

Yes. MSCI, S&P, and other providers publish quality indexes for international developed and emerging markets, and the underlying academic research — including Asness, Frazzini, and Pedersen’s international tests in the original QMJ paper — found a comparable, though not identical, premium outside the United States. Liquidity, disclosure quality, and accounting standards vary more by country, though, which can make the metrics themselves noisier in some emerging markets.

References

• Asness, Clifford S., Andrea Frazzini, and Lasse Heje Pedersen. “Quality Minus Junk.” Review of Accounting Studies, 2019.
• McLean, R. David, and Jeffrey Pontiff. “Does Academic Research Destroy Stock Return Predictability?” The Journal of Finance, 2016.
• Novy-Marx, Robert. “The Other Side of Value: The Gross Profitability Premium.” Journal of Financial Economics, 2013.
• Hou, Kewei, Chen Xue, and Lu Zhang. “Replicating Anomalies.” The Review of Financial Studies, 2020.
• MSCI. “MSCI USA Quality Index Methodology.” MSCI Index Methodology Documentation.
• S&P Dow Jones Indices. “S&P 500 Quality Index Methodology.” S&P DJI Methodology Documentation.