This Paper Could Change How You Invest

High-level takeaway

Fama and French (1993) showed that multiple common risk factors — not just market beta — explain most of the cross‑sectional variation in average stock returns. This finding transformed empirical asset pricing and practical portfolio construction (factor investing). Practical implication: long‑run expected returns appear driven by exposure to identifiable factors; investors can tilt portfolios to these factors (via funds/ETFs) to raise expected returns. Debate remains whether these are compensated risks or persistent mispricings, and implementation (transaction costs, capacity) matters.

Key facts, timeline, and headline numbers

• Paper: Fama & French (1993, Journal of Financial Economics); widely cited (~15,000 citations reported).
• Sample period used in tests: July 1963 – December 1991.
• Core empirical result (3‑factor model):
  • Time‑series R² across 25 size × book‑to‑market portfolios ranged ~0.83–0.97 (average ~0.93).
  • 21 of 25 portfolios had R² > 0.90.
• Comparison to CAPM:
  • CAPM (market beta alone) explained ~60% of cross‑section return variation.
  • Fama‑French 3‑factor explained ~90%.
• 2015 update: Fama‑French 5‑factor model (adds profitability and investment factors) improved explanatory power toward ~95%.
• Factor proliferation: a 2016 census identified ~316 distinct published factors (the “factor zoo” phenomenon).

Assets, instruments, and vehicles referenced

• Stocks: large vs small cap; value vs growth.
• Bonds: mentioned in passing (paper looked at stock and bond factors), though video focuses on stocks.
• Long‑short factor portfolios (standard factor construction).
• ETFs and funds: Dimensional Fund Advisors (DFA); Avantis Investors (Canadian ETFs via CIBC); general reference to low‑cost funds and ETFs implementing factor exposures.
• Tools: Portfolio Visualizer (for regressions/time‑series analysis).

Factors and nomenclature

• Market factor: overall market return (same concept as CAPM market beta).
• SMB — Small Minus Big: size factor (long small caps, short big caps).
• HML — High Minus Low: value factor (long high book‑to‑market/value stocks, short low book‑to‑market/growth stocks).
• RMW — Robust Minus Weak: profitability factor (long high profitability, short weak).
• CMA — Conservative Minus Aggressive: investment factor (long low investment/conservative firms, short high investment/aggressive firms).
• Typical construction: factor portfolios are built as long‑short portfolios to capture premiums.

Methodology (step‑by‑step summary)

1. Sort the universe of stocks into 5 size groups × 5 book‑to‑market groups → 25 portfolios (captures combinations such as small‑value, large‑growth, etc.).
2. Construct factor returns:
   • Market factor.
   • SMB (long small, short big).
   • HML (long high B/M, short low B/M).
   • (For the 5‑factor model, add RMW and CMA.)
3. Run time‑series regressions of each test portfolio’s excess returns on the factor returns:
   • Regression outputs: factor loadings (exposures/betas), alpha (intercept = unexplained return), R² (explanatory power).
4. Evaluate alphas and R² across portfolios to assess model fit.
5. Robustness checks: test the model on portfolios sorted by other characteristics (e.g., dividend‑to‑price, earnings‑to‑price).

Key empirical observations and exceptions

• Once market, size and value exposures are accounted for, most of the variation in returns across diversified portfolios is explained; most portfolios show near‑zero alphas.
• Exception: small‑cap growth stocks produced much lower returns than the model predicted (a negative alpha), leaving some puzzles for further research.
• Many active funds that beat benchmarks did so via factor tilts rather than unique manager skill; those exposures can often be replicated more cheaply.

Interpretation, debate, and caveats

• Two main interpretations of factor premiums:
  • Compensated, undiversifiable risks (risk premia).
  • Persistent mispricings exploited by investors.
• The joint‑hypothesis problem makes it hard to definitively resolve whether factors are risk premia or mispricings.
• Factor proliferation (“factor zoo”) highlights the need for careful selection, robust testing, and skepticism about data mining and publication bias.
• Implementation caveats: transaction costs, capacity constraints, turnover, and exact portfolio construction materially affect realized returns. Factor tilts raise expected returns but do not guarantee outcomes.

Practical applications and industry implementation

• Factor investing products:
  • Dimensional Fund Advisors historically implemented Fama‑French research (many DFA strategies are available as ETFs in some markets).
  • Avantis Investors offers factor‑based products/ETFs (Canadian distribution via CIBC noted).
• Portfolio construction:
  • Use factor exposures (tilts) to target higher expected returns.
  • Prefer diversified, low‑cost implementation and account for trading/transaction costs.
• Performance evaluation:
  • Use multi‑factor regressions (3‑ or 5‑factor) to decompose an active manager’s returns into factor exposures and alpha.

Explicit recommendations and cautions

• Evidence supports tilting toward certain factor exposures for higher long‑term expected returns, but:
  • Mind transaction costs, capacity, and implementation details.
  • Recognize ongoing debate over whether factors reflect true risk premia or mispricings.
  • Treat factor returns as long‑term expected outcomes — not guarantees. Continue to diversify and control costs.

Disclosures and presenter notes (video)

• Presenter: Ben Felix, Chief Investment Officer, PWL Capital.
• Disclosure: Ben Felix / PWL Capital use Dimensional Fund Advisors funds extensively; he was not paid by Dimensional or Avantis to mention them.
• People and references mentioned: Eugene Fama and Kenneth French (authors), William (Bill) Sharpe (CAPM/Nobel reference), John Cochrane (comment on “factor zoo”), a 2016 factor census.

Primary sources and presenters mentioned

• Ben Felix (presenter; CIO at PWL Capital)
• Eugene Fama (co‑author)
• Kenneth French (co‑author)
• Dimensional Fund Advisors (factor strategy implementer)
• Avantis Investors (factor product provider)
• John Cochrane (commentator)
• William (Bill) Sharpe (CAPM reference)

Short checklist to apply this in practice

• Identify desired factor tilts (SMB, HML, RMW, CMA, market).
• Choose implementation vehicle (low‑cost funds/ETFs, institutional funds, or direct factor portfolios).
• Run multi‑factor regressions on your portfolio to measure exposures, alpha, and R² (tools: Excel, Portfolio Visualizer, etc.).
• Monitor transaction costs, capacity, and changing factor behavior; avoid overfitting to historical signals.