Introducing …and the Cross-Section of Expected Returns for after-work quants

Takeaway. With a crowded factor zoo, a t-stat near 2 is often too low—Harvey–Liu–Zhu argue for a much higher multiple-testing hurdle (popularly remembered as t > 3) before celebrating a “new” factor.

…and the Cross-Section of Expected Returns (Campbell Harvey, Yan Liu, Heqing Zhu, 2016) is the scepticism classic every after-work factor notebook should keep open. (Related Hub note: harvey-liu-zhu-factor-zoo.)

Fit for learners: Sets the t>3 multiple-testing bar so you stop overfitting “new” factors in notebooks.

This Hub note is a plain-language intro. Not investment advice. Read beside Taming the Factor Zoo and Thousands of Alpha Tests.

The problem

Researchers and blogs keep mining characteristics until something clears |t|≈2. If many hypotheses were tried—including unpublished ones—false discoveries soar. Folklore single-test cutoffs mislead.

The idea / core intuition

• Document how large the cross-section factor literature became (the zoo).
• Argue for higher t-hurdles under multiple testing than the single-test 1.96 rule of thumb.
• Push better design: out-of-sample, economics, replication—not only in-sample stars.
• Does not say every factor is fake; it says your bar is probably too low.

Why after-work Python learners care

Python makes mining 50 sorts before dinner easy. This paper is the cultural seatbelt: code less greedily, validate more honestly.

Tiny evening exercise: simulate 100 independent noise “factors”; count how many clear |t|>2 by luck vs a higher hurdle like 3; write the lesson for your next real idea—learning only.

One-evening reading checklist

1. Define “factor zoo” in plain language.
2. Explain why multiple testing raises the t bar.
3. Note the popular “t > 3” memory aid—and that you should still read the paper’s nuance.
4. List two non-t-stat checks for every factor notebook.

Download PDF

Download PDF

Landing: NBER w20592

Important disclaimer

Not investment advice. This Hub note is educational only. Harvey–Liu–Zhu is a methodology warning—not a buy/sell list of factors or ETFs.

What this is not

• Not CEF / subsidy framing
• Not Quant or Macro course CTA
• Not a “factors are dead” clickbait thread

Next step

Single course CTA: SoloQuant Academy’s Introduction to Python Programming—statistics makes more sense once you can simulate noise in a few dozen lines of honest code.