The Case Against Factor Investing: Criticisms and Debates

Intellectual Honesty Required

Factor investing criticism deserves serious attention. No investment approach survives scrutiny without honest examination of its weaknesses, and factor investing has attracted serious criticism from academics, practitioners, and skeptics. Some of these critiques are overblown; others are genuinely concerning.

This article presents the strongest arguments against factor investing, including concerns about factor decay and data mining. If you're going to tilt your portfolio toward factors, you should understand what might go wrong.

Criticism #1: The Data Mining Problem

The "Factor Zoo"

Academic finance has a publication problem. Researchers are rewarded for finding new factors, creating strong incentives to mine historical data for patterns—any patterns.

The result: a "factor zoo" of hundreds of claimed factors. By one count, over 400 factors have been proposed in published papers.

How many are real? Far fewer than 400.

Harvey, Liu, and Zhu (2016) published a devastating analysis arguing that most reported factors are likely false discoveries. Their logic: when you test enough combinations, statistical flukes will appear significant. Standard thresholds for statistical significance (p < 0.05) are far too loose when hundreds of hypotheses are tested.

Their estimate: a factor needs a t-statistic above 3.0—not the traditional 2.0—to be credible. By this standard, most claimed factors fail.

How Bad Is It?

Consider the methodology behind factor discovery:

1. Download decades of stock returns
2. Test various characteristics (momentum, value, quality, accruals, etc.)
3. Run regressions, find significant relationships
4. Publish the ones that work

The problem: step 3 involves thousands of implicit tests. With enough searching, you'll find patterns even in random data.

The Defense

Proponents argue that the major factors—market, size, value, profitability, momentum—survive this critique:

• They work across countries, not just US data
• They work across time periods, including out-of-sample
• They have economic rationale (behavioral or risk-based)
• They've been tested by independent researchers

The "factor zoo" problem affects exotic, narrow factors more than the canonical ones. But the critique is a reminder: not everything labeled a "factor" deserves the name.

Criticism #2: Factor Decay and Crowding

The Efficient Markets Counterattack

If factors represent true mispricings, shouldn't they disappear once discovered? Investors would pile in, arbitrage away the premium, and future returns would normalize.

This is exactly what some critics argue has happened.

The crowding argument:

1. Academics discover a factor
2. Asset managers create products around it
3. Trillions flow into factor ETFs
4. Prices adjust; the premium shrinks or vanishes
5. Factor investors are left with systematic exposure to something that no longer works

Evidence of Decay?

There's some evidence for factor decay:

Size premium: Largely absent in recent decades. Small-cap underperformance from 2000-2020 suggests the original premium may have been arbitraged away or overstated.

Value premium: Significantly weaker since the 1990s. Value's "lost decade" (2010-2020) was historically unprecedented in severity.

Momentum: Still appears in data, but crash risk may be higher as more capital chases the strategy.

AUM in factor ETFs has exploded from near-zero to hundreds of billions. It would be surprising if this had no effect.

The Defense

Counter-arguments:

Risk premiums can't be arbitraged away. If value stocks are genuinely riskier, their premium reflects compensation for risk, not mispricing. You can't eliminate the risk, so you can't eliminate the premium.

Behavioral explanations persist. If factors exist because of human psychology (overconfidence, herding, etc.), these biases don't disappear because academics wrote papers about them.

Factor timing is hard. Even if arbitrageurs know a factor exists, timing it is difficult. The premium persists because capturing it requires patience most investors lack.

The truth is probably somewhere in the middle: some premiums have shrunk, others persist, and we can't know for sure which is which.

Criticism #3: Recent Underperformance

Value's Lost Decade

No discussion of factor criticism is complete without addressing value's spectacular failure from 2010-2020.

During this period:

• Value stocks underperformed growth by massive margins
• Many value investors gave up in frustration
• Growth stocks (especially big tech) dominated returns
• Traditional value metrics looked increasingly broken

For investors who tilted toward value, this wasn't abstract. It meant watching your portfolio lag for a decade while "overpriced" tech stocks kept rising.

Is Value Dead?

Prominent voices declared value investing obsolete. Arguments included:

Intangible assets: Traditional value metrics (price-to-book) don't capture software, brands, network effects. "Expensive" tech stocks might be fairly valued by correct metrics.

Low interest rates: Near-zero rates favor growth stocks with cash flows far in the future. Value's underperformance tracked falling rates.

Winner-take-all dynamics: Digital economies create monopolies. Traditional value stocks (banks, energy, manufacturing) were structural losers.

Structural change: Maybe the world has genuinely changed, and historical premiums are irrelevant.

The Partial Recovery

Value partially recovered in 2021-2022 as:

• Interest rates rose
• Inflation returned
• Tech stocks corrected
• Energy and financials outperformed

Does this vindicate value? It's too early to conclude. One good year (or two) doesn't prove a factor works. Value believers argue mean reversion is happening; skeptics say it's just a cyclical bounce.

Criticism #4: Implementation Is Harder Than It Looks

Theory vs. Practice

Academic factor research often assumes:

• Zero transaction costs
• Unlimited shorting capacity
• No market impact
• Perfect rebalancing

Real-world implementation is messier:

Costs erode premiums. Factor ETFs charge 0.15-0.35%. Trading costs add up. If premiums are 2-3% and costs are 0.5%, you've lost a significant chunk.

Factor definitions matter. How you define "value" or "momentum" affects returns significantly. Some definitions work better than others. Academic papers might use definitions that are hard to replicate.

Turnover creates tax drag. Momentum strategies, in particular, require frequent trading. In taxable accounts, this generates short-term capital gains.

Tracking error is painful. Multi-year underperformance versus benchmarks causes many investors to abandon factor strategies at exactly the wrong time.

Behavioral Reality

Perhaps the most underrated criticism: factor investing is psychologically difficult.

Imagine you tilted heavily toward value in 2015. By 2020, you'd watched the market (dominated by growth stocks) beat you for five straight years. Friends who bought FAANG stocks were celebrating. Your "disciplined" value approach felt like an expensive mistake.

How many investors could stick with the strategy? Very few. And abandoning at the bottom is exactly the wrong move.

Factor investing assumes a patience that most investors don't have.

Criticism #5: Factors Aren't "Smart Beta"

The Marketing Problem

Asset managers branded factor investing as "smart beta"—implying it's a smarter version of index investing.

Critics argue this is misleading:

It's not smart; it's different. Factor investing makes active bets. Calling it "smart" implies superiority that isn't guaranteed.

Benchmark-hugging doesn't work. Many factor products have weak factor exposure to minimize tracking error. They're neither good index funds nor good factor portfolios.

Costs are higher than true indexing. If premiums are uncertain but costs are certain, maybe cheap index funds are "smarter."

Complexity for its own sake. Some multi-factor products are so complex that no one—including the managers—fully understands what they're doing.

The Defense

Factor investing isn't a magic formula. It's a framework for understanding returns. Used properly, it can:

• Improve risk-adjusted returns over long periods
• Help investors understand what bets they're making
• Provide diversification beyond market cap weighting

Used improperly—chasing hot factors, timing, over-complicating—it can underperform simple index funds.

A Balanced Conclusion

Having considered the critics, where does factor investing stand?

What the Critics Get Right

• Data mining is real. Many "factors" are statistical noise. Be skeptical of exotic factors.
• Premiums may have shrunk. Factor decay is plausible. Expect less than historical averages.
• Implementation is hard. Costs, turnover, and behavioral challenges erode real-world returns.
• Extended underperformance happens. Factors can lag for a decade. Can you handle that?

What the Critics Get Wrong (Or Overstate)

• The major factors aren't debunked. Market, value, momentum, quality have robust evidence across markets and time periods.
• Risk premiums persist. If factor premiums compensate for genuine risk, they can't be arbitraged away.
• Alternatives are worse. Stock picking has even less evidence of sustainable outperformance. Factor investing, for all its flaws, is more defensible than most active strategies.

A Pragmatic View

Factor investing isn't a magic formula, but it's a reasonable framework. The honest case:

Factor investing probably works, modestly, over long periods.

"Probably" because certainty isn't possible. "Modestly" because premiums are smaller than historical data suggests and costs erode them further. "Over long periods" because factors can underperform for years or decades.

This isn't a rousing endorsement, but it's realistic. Factor investing offers a slight edge with evidence behind it. That's more than most investment strategies can claim.

Key Takeaways

• Data mining is a valid concern: The "factor zoo" contains many false discoveries. Stick to well-documented factors with economic rationale.
• Factor decay may be real: Premiums might be smaller going forward. Expect less than historical averages.
• Value's lost decade happened: Extended underperformance is possible for any factor. Patience is required.
• Implementation erodes returns: Costs, taxes, and behavioral mistakes reduce real-world premiums.
• Factors aren't magic: They're a slight edge, not a guaranteed win.
• The alternative (stock picking) is worse: Factor investing has more evidence than most active strategies.
• Modest expectations are appropriate: 1-2% premium over the market, with significant tracking error, over 10+ year horizons.

Factor investing deserves skepticism. But informed skepticism leads to realistic expectations and appropriate position sizing—not abandonment of an evidence-based framework for stock market returns.

---

This concludes our Factor Investing series. Return to What Is Factor Investing? to revisit the fundamentals, or explore FactorIQ's tools to analyze your own portfolio's factor exposures.