Beyond the Three-Factor Model: RAFI, Profitability and the Evolution of Smart Beta
The Fama-French Three-Factor Model (1993) was a monumental leap forwards in empirical asset pricing. By adding size and value factors to the Capital Asset Pricing Model (CAPM), Eugene Fama and Kenneth French were able to explain much more of the variation in stock returns between two different diversified portfolios that CAPM failed to account for. Yet, despite its elegance, the three-factor model left a number of persistent anomalies unexplained; especially in accounting for why ‘HML’ (High Minus Low) value strategies can underperform as well as the well-known momentum anomaly.
This article primarily focuses on how two complementary frameworks—RAFI's Fundamental Indexing and Novy-Marx’s profitability research—have added more precision to value investing and advanced our understanding of asset pricing. We also explore momentum in more detail and uncover why Fama and French chose to leave it out of their revised Five-Factor Model (2015).
The Limitations of Traditional Value Investing
Traditional value strategies—especially those using price-to-book (P/B) ratios like Fama and French’s HML—often suffer from two flaws:
Accounting distortions: Book value omits intangible assets like brand value, software development, R&D, customer relationships, or human capital. This penalises innovative firms and overstates the capital base of outdated businesses. Modern economies are increasingly intangible-intensive, especially in tech, healthcare, and services. Ignoring intangibles means systematically understating the true economic capital of many growth companies and overstating it for asset-heavy ones.
No quality filter: Value stocks are often cheap for a reason. Without screening for profitability or reinvestment discipline, these strategies risk overloading on distressed firms with little upside potential.
This is where both RAFI and Novy-Marx make meaningful contributions — by offering more nuanced methods of distinguishing genuine value from junk.
Fundamental Indexing: The RAFI Approach
RAFI (Research Affiliates Fundamental Index), pioneered by Rob Arnott, rejects price-based weighting. Instead, companies are weighted by economic size, using measures like:
5-year average sales
Operating cash flow
Dividends
Book value (optionally adjusted for intangibles)
This leads to a systematic rebalancing mechanism — trimming outperformers and adding to underperformers, but always guided by fundamentals, not prices. The result is a contrarian, value-tilted strategy that has historically outperformed traditional market cap-weighted indices.
RAFI’s value strategies have delivered around 2% per annum excess returns over Fama-French value benchmarks in developed markets (Arnott et al. 2013), in part because they implicitly favour profitable companies and avoid value traps.
Whilst the 2% outperformance is striking, it's important to view the claim in context:
Much of the outperformance can be attributed to deeper value exposure: RAFI value portfolios tend to have a lower price/fundamentals profile than traditional Fama-French value—i.e., they are more ‘value-like’.
Fama-French Value is a simple academic construct, often implemented in a long-only top 30% (value) vs bottom 30% (growth) fashion using one variable (typically P/B). Real-world smart beta and fundamental value strategies (like RAFI) refine their exposure using better data and multiple metrics.
Click to Expand
|
Click to Expand |
Novy-Marx and the Profitability Revolution
In his landmark 2013 paper, ‘The Other Side of Value’, Robert Novy-Marx shows that gross profitability (revenues minus COGS, scaled by assets) is a powerful, independent predictor of returns — with similar strength to value.
‘Profitability is the other side of value.’
High-profit firms tend to outperform, even when controlling for size, value, and momentum. More importantly, profitability is negatively correlated with value, meaning that many cheap stocks are also low-quality — hence underperformance in naïve value portfolios.
Different Methods, Same Destination
Though they differ in approach, both RAFI and Novy-Marx seek to answer a deeper question:
How can we identify firms that are not only cheap, but financially strong?
How Novy-Marx Shaped the Fama-French Five-Factor Model
Fama and French took notice. In 2015, they expanded their model by adding:
RMW (Robust Minus Weak): profitability
CMA (Conservative Minus Aggressive): investment behaviour
These additions improved the model significantly — but made the original value factor statistically redundant in some datasets. This implied that traditional value was proxying for profitability and investment efficiency, but doing so imprecisely.
This wasn’t a refutation of value — it was a refinement.
Why Has Momentum Been Neglected?
When Fama and French expanded their original 3-factor model to a 5-factor model in their 2015 paper, ‘A Five-Factor Asset Pricing Model’, they included:
Market risk (MKT)
Size (SMB)
Value (HML)
Profitability (RMW)
Investment (CMA)
Momentum (as in Jegadeesh and Titman 1993) — the tendency of past winners to outperform past losers — was conspicuously left out. The reason being that there is no rational risk-based story:
Fama and French’s models are rooted in rational asset pricing theory. They aim to explain returns as compensation for systematic risk — not behavioural anomalies or market inefficiencies.
Fama and French (2015) argue that momentum does not have a clear risk-based justification. It looks more like a behavioural anomaly — a deviation from market efficiency — than a premium for bearing risk.
In their 2015 paper, they wrote:
‘We do not add a momentum factor to the model because it is a short-term return anomaly that has no clear link to the long-term economic risks that should matter in asset pricing.’
Whilst Fama and French exclude momentum, other major asset pricing frameworks—such as AQR’s multi-factor models (Asness et al.)—do include momentum alongside value, size, quality, and low volatility. Why? Because empirically, momentum works. It’s one of the most consistently observed return patterns across asset classes, regions, and time periods. Even if it lacks a rational explanation, its performance is statistically robust (Asness, Moskowitz, and Pedersen 2013).
Factor Momentum: Explaining More with Less
A growing body of research (Ehsani and Linnainmaa 2019; Moskowitz, Ooi, and Pedersen (2012) shows that factor momentum — the tendency for recent outperforming factors (value, quality, momentum, etc.) to persist — subsumes much of the return from individual stock momentum.
Key findings:
Momentum works not just at the stock level, but across style factors
Factor momentum portfolios are lower turnover, better diversified, and cheaper to run than individual stock momentum portfolios
Lower turnover
Factor portfolios adjust exposures less frequently than stock momentum strategies, as they’re based on smoother, aggregate trends across groups of stocks rather than volatile individual price movements.
Lower implementation cost
Trading a few broad factor exposures (like value or quality) is significantly cheaper than the high turnover required to maintain a fast-changing stock momentum portfolio, which suffers from transaction costs and slippage.
Broader diversification
Factor strategies operate across large groups of stocks, reducing idiosyncratic risk and ensuring more stable performance compared to concentrated individual stock positions.
Risk-based interpretation
Factor momentum can be viewed through the lens of changing macroeconomic risk premia or shifting investor sentiment, offering a more intuitive link to economic fundamentals than traditional stock-level anomalies.
Real-World Applications - The Smart Beta Approach
Other asset managers have integrated these findings into practically investable funds:
Dimensional Fund Advisors (DFA): moved from pure Fama-French Three-Factor model concentration to include profitability, investment, and momentum tilts
Avantis Investors: incorporates multiple quality signals alongside value in portfolio construction
These implementations tend to deliver better risk-adjusted returns than naïve value approaches, whilst retaining a systematic, rules-based discipline.
Conclusion: Better Tools for Modern Markets
Markets evolve. So must our models.
The Fama-French Three-factor model was revolutionary — but left room for improvement. Novy-Marx showed that profitability completes the picture. Arnott operationalised value more intelligently via fundamentals. Factor momentum added yet another layer of explanatory power.
Investors should no longer ask: ‘Is this stock cheap?’ They should ask: ‘Is it strong financially and cheap?’
Smart value today is disciplined, quality-aware, and factor-diversified. The ideas from RAFI and Novy-Marx help make that possible.
References
Arnott, R. D., Hsu, J., Kalesnik, V., and Tindall, P. 2013. ‘The Surprising Alpha from Malkiel’s Monkey and Upside-Down Strategies’. Journal of Portfolio Management 39 (4): 91–105.
Arnott, R. D., Harvey, C. R., Kalesnik, V., and Linnainmaa, J. T. 2021. ‘Reports of Value’s Death May Be Greatly Exaggerated’. Financial Analysts Journal 77 (1): 44–67.
Asness, C. S., Frazzini, A., Israel, R., and Moskowitz, T. J. 2015. ‘Fact, Fiction, and Value Investing’. Journal of Portfolio Management 42 (1): 34–52.
Asness, C. S., Moskowitz, T. J., and Pedersen, L. H. 2013. ‘Value and Momentum Everywhere’. Journal of Finance 68 (3): 929–985. https://doi.org/10.1111/jofi.12021
Ehsani, S., and Linnainmaa, J. T. 2019. ‘Factor Momentum and the Momentum Factor’. NBER Working Paper No. 25551.
Fama, E. F., and French, K. R. 2015. ‘A Five-Factor Asset Pricing Model’. Journal of Financial Economics 116 (1): 1–22.
Jegadeesh, N., and Titman, S. 1993. ‘Returns to Buying Winners and Selling Losers: Implications for Stock Market Efficiency’. Journal of Finance 48 (1): 65–91. https://doi.org/10.1111/j.1540-6261.1993.tb04702.x
Moskowitz, T. J., Ooi, Y. H., and Pedersen, L. H. 2012. ‘Time Series Momentum’. Journal of Financial Economics 104 (2): 228–250.
Novy-Marx, R. 2013. ‘The Other Side of Value: The Gross Profitability Premium’. Journal of Financial Economics 108 (1): 1–28.