Factor Investing Is Also About Diversification
Factor investing is generally presented as a way to increase expected returns. Indeed, many investors deliberately tilt their portfolios towards groups of companies that have historically delivered higher returns than the wider market. Value investors favour relatively cheap companies. Profitability strategies favour companies with stronger operating profitability. The objective is to capture additional risk premia over long periods. Whilst that is reason enough for many to pursue factor investing, it is not the only valid reason to do so.
Indeed, combining the market, value and profitability premia may also provide a form of diversification. These return drivers do not reliably perform well or badly at the same time. An investor who spreads exposure across them is therefore less dependent on any single premium being rewarded over a particular period.
The historical evidence provides a striking illustration. Across the complete calendar-year histories available from the Fama–French Data Library, the market, value and profitability premia have never all been negative in the same year in either the United States or developed markets excluding the United States.
That does not mean a factor-tilted equity portfolio cannot lose money. It means the different sources of return within that portfolio have historically experienced their difficult periods at different times.
Three Sources of Equity Returns
A conventional diversified equity portfolio is dominated by the market risk premium.
The market premium is the return from owning the broad stock market above the return available from relatively safe short-term assets. Investors expect to receive this premium because shares expose them to uncertain earnings, economic downturns and potentially severe falls in value.
Factor investing introduces exposure to additional sources of expected return.
The value premium is the difference between the returns of companies with high book-to-market ratios and companies with low book-to-market ratios. In simpler terms, it measures the relative performance of cheap ‘value’ stocks against expensive ‘growth’ stocks.
The profitability premium is the difference between the returns of companies with robust operating profitability and companies with weak operating profitability.
Fama and French incorporate these characteristics into their five-factor model, alongside market, size and investment factors. Their research finds that average stock returns are related not only to market exposure and valuation, but also to differences in profitability and corporate investment (Fama and French 2015).
Value and profitability strategies are normally justified by their higher expected returns. However, their imperfect relationship with the market, and with each other, means that they may also diversify the source of those returns.
How the Analysis Was Conducted
The analysis uses annual factor returns from the Kenneth R. French Data Library.
For the United States, the following three series were taken from the Fama–French five-factor dataset:
Rm − Rf, representing the return of the value-weighted US stock market above the one-month US Treasury bill rate;
HML, representing the return of high book-to-market stocks minus low book-to-market stocks; and
RMW, representing the return of robustly profitable stocks minus weakly profitable stocks.
The US five-factor dataset begins in July 1963. As 1963 does not contain a complete 12 months, the analysis begins with the first full calendar year in 1964 and runs to 2025. This provides 62 complete annual observations. The underlying factors are constructed from value-weighted portfolios formed using independent sorts on company size and the relevant company characteristic.
The same three series were then taken from the developed markets excluding the United States five-factor dataset.
The developed ex-US data begin in July 1990, so the first complete calendar year is 1991. The available annual sample therefore covers 1991 to 2025, providing 35 complete observations. Returns are expressed in US dollars, include dividends and capital gains and are not continuously compounded.
Emerging markets were excluded from the analysis.
For each calendar year, the three factors were classified simply as positive or negative. A year was considered simultaneously negative only where:
Rm − Rf < 0
HML < 0
RMW < 0
The test therefore asks whether, in any complete calendar year, equities underperformed Treasury bills, value underperformed growth and robustly profitable companies underperformed weakly profitable companies at the same time.
What Do We Find?
In the United States, there was no calendar year between 1964 and 2025 in which the market, value and profitability premia were all negative.
The same result appears outside the United States.
In developed markets excluding the United States, there was no calendar year between 1991 and 2025 in which all three premia were negative.
Across 62 complete years in the United States and 35 complete years in developed ex-US markets, at least one of the three return drivers was positive in every calendar year.
This is an unusually intuitive demonstration of factor diversification.
It does not establish that the factors are negatively correlated. Nor does it show that they perfectly offset one another. It shows that their periods of underperformance have not historically coincided.
Sometimes the market premium has been negative whilst value or profitability has remained positive. At other times, equities have produced strong returns whilst value stocks have underperformed growth stocks. Profitability has also experienced periods of weakness that have not necessarily coincided with poor returns from the other two premia.
The factors have taken turns.
Why Might the Premia Perform Differently?
The three factors represent different economic exposures.
The market premium is affected by broad changes in economic growth, corporate profits, interest rates and investor risk appetite. When investors become worried about recession or financial instability, share prices across much of the market may fall together.
The value premium is relative. It depends on whether cheap companies outperform expensive companies.
Value can struggle during periods in which investors place increasingly high valuations on companies expected to deliver rapid future growth. Even where the stock market rises strongly, value may produce a negative relative return because growth stocks rise by more.
The profitability premium compares companies with robust operating profitability against those with weak profitability.
Profitable companies may be better placed to withstand difficult economic conditions, but the premium can still be negative. For example, weaker or more speculative companies may outperform sharply during a market rebound when investors become more willing to accept risk.
These return drivers are connected, but they are not identical.
A company can be cheap without being highly profitable. A highly profitable company can also trade at an expensive valuation. The economic conditions that favour one characteristic may therefore be less favourable for another.
That is the source of the potential diversification benefit.
Diversification Does Not Require Opposite Returns
Diversification is sometimes understood too narrowly.
Two investments do not have to move in opposite directions for their combination to be useful. They merely need to avoid moving perfectly together.
Imagine two strategies that both have positive long-term expected returns. One may perform poorly during a period in which the other performs relatively well. Combining them can reduce the portfolio’s dependence on either strategy producing its expected premium immediately.
This is important because individual factors can underperform for prolonged periods.
An investor holding only the market must accept periods in which equities underperform cash. An investor concentrating heavily in value must accept that expensive growth companies can outperform for many years. A profitability strategy can also fall behind when weaker or more speculative businesses lead the market.
Combining the market, value and profitability premia does not remove any of those risks. It spreads exposure across several sources of uncertainty rather than relying entirely on one.
Factor investing can therefore be viewed as diversification through different expected-return drivers.
Positive Factor Returns Do Not Necessarily Prevent Losses
There are, however, important caveats to the findings to point out.
HML and RMW are long–short relative returns. They are not the absolute returns earned by a conventional long-only investment fund.
A positive HML return means value stocks outperformed growth stocks. It does not necessarily mean value stocks made money.
Suppose the broad equity market falls by 20%. Value stocks might fall by 15%, whilst growth stocks fall by 25%. Value has outperformed growth by 10 percentage points, meaning the value premium is positive, but an investor in value stocks has still suffered a substantial loss.
The same applies to profitability. Robustly profitable companies can outperform weakly profitable companies whilst both groups decline in absolute terms.
A factor-tilted equity portfolio therefore remains an equity portfolio. Its absolute returns will usually continue to be dominated by the market.
The historical result does not mean:
‘One of the factors always protects the portfolio from losses.’
It means:
‘At least one of the three measured return spreads has historically been positive in every complete calendar year examined.’
That is a diversification benefit, but not a guarantee of capital preservation.
Diversification Within Equities Is Not Asset-Class Diversification
Factor diversification should also be distinguished from diversification across asset classes.
A portfolio combining market, value and profitability exposure remains vulnerable to risks that affect equities generally. During a severe downturn, positive value or profitability premia may merely reduce the size of the loss relative to a market-only portfolio.
High-quality bonds, cash and other defensive assets serve a different purpose. Their returns are not constructed as relative spreads within the stock market and they may respond differently to changes in economic growth, inflation and interest rates.
Factor investing can diversify the equity allocation, but it does not replace broader portfolio diversification.
This distinction is especially important when discussing risk with investors. A multi-factor equity fund may be better diversified than a highly concentrated growth fund, but it should not be described as low risk simply because it holds several factors.
The Result Is Not a Law of Nature
The absence of a simultaneous negative year is a historical observation, not a permanent relationship.
There is no mathematical or economic mechanism preventing the market, value and profitability premia from all being negative in the same year.
The finding is also dependent on the methodology used.
First, the analysis uses calendar years. Testing individual months, quarters or rolling 12-month periods could produce simultaneous negative observations.
Calendar-year boundaries are ultimately arbitrary. A negative period beginning in July and ending the following June could be divided between two calendar years and disappear from an annual analysis.
Second, the result uses the specific Fama–French definitions of the factors. Other models may define value and profitability differently.
The Fama–French value factor is constructed using book-to-market ratios. A live value fund might also consider measures such as price-to-earnings, price-to-cash-flow or enterprise value relative to operating profits.
Similarly, profitability can be measured using operating profitability, return on equity, gross profitability or other quality-related measures.
Third, the international data are measured in US dollars. A UK investor evaluating returns in sterling could experience different results because of currency movements.
Finally, academic factor returns exclude some of the practical complications faced by investors, including fund fees, taxes, trading costs, capacity constraints and differences between theoretical portfolios and investable products.
The correct conclusion is therefore limited but still meaningful:
Across the complete annual histories currently available from the Fama–French Data Library, the market, value and profitability premia were not simultaneously negative in the United States or developed markets excluding the United States.
It would be incorrect to claim that they can never be simultaneously negative.
Why This Matters for Investors
Expected returns are uncertain.
Although market, value and profitability exposure may each be associated with higher expected returns, no one knows when those returns will materialise. A premium can be positive over several decades whilst remaining negative over a period long enough to test an investor’s patience.
This creates a practical problem.
An investor who relies heavily on one factor may abandon it after years of disappointment. The premium may subsequently recover, but the investor no longer participates.
Diversifying across factors may make the journey more tolerable. When one premium disappoints, another may provide better relative results. This can reduce the temptation to identify a recently weak factor as ‘broken’ or to chase whichever investment style has performed best.
There is also an important distinction between expected return and reliability.
Adding more exposure to one factor may raise the portfolio’s expected return, but it also increases dependence on that factor. Combining several credible premia may produce a more balanced portfolio, even where each individual exposure is less pronounced.
The objective is not necessarily to maximise the theoretical return from whichever factor appears strongest. It is to build a portfolio that does not require a single uncertain source of return to succeed.
Risk Premia Must Sometimes Be Uncomfortable
A genuine risk premium cannot be expected to deliver smooth and consistent outperformance.
If value always beat growth, investors would rapidly bid up value companies until the opportunity disappeared. The same applies to profitability and the market premium.
Periods of disappointing performance are not necessarily evidence that a premium has stopped existing. They may be part of the reason the premium can persist.
This is where diversification becomes especially valuable.
An investor does not know which premium will underperform next, how severe that underperformance will be or how long it will last. Holding several economically distinct sources of expected return acknowledges that uncertainty.
Factor investing is therefore not simply a bet that value and profitable companies will outperform.
It is a decision to avoid relying solely on the market premium and to diversify exposure across several return drivers that have historically been rewarded at different times.
Conclusion
Factor investing is normally discussed as a way of pursuing higher expected returns. Its diversification benefit receives less attention.
The available Fama–French evidence shows that the market, value and profitability premia have not all been negative in the same calendar year across 62 complete years in the United States or 35 complete years in developed markets excluding the United States.
This does not guarantee positive portfolio returns. Value and profitability are relative factors, and a factor-tilted portfolio remains heavily exposed to the overall equity market.
Nor does the result prove that the factors cannot all underperform together in the future.
What it does demonstrate is that different equity premia have historically experienced their difficult periods at different times.
A diversified factor portfolio therefore offers more than exposure to several potential sources of higher expected return. It reduces reliance on any one of those sources being rewarded immediately.
That is the central role of diversification: not to eliminate uncertainty, but to avoid making the success of the entire portfolio dependent on a single uncertain outcome.
References
Fama, Eugene F., and Kenneth R. French. 2015. ‘A Five-Factor Asset Pricing Model’. Journal of Financial Economics 116 (1): 1–22.
French, Kenneth R. 2026a. ‘Description of Fama/French 5 Factors (2x3)’. Kenneth R. French Data Library.
French, Kenneth R. 2026b. ‘Description of Fama/French 5 Factors for Developed Markets’. Kenneth R. French Data Library.
French, Kenneth R. 2026c. ‘Description of 6 Portfolios Formed on Size and Book-to-Market for Developed Markets’. Kenneth R. French Data Library.
French, Kenneth R. 2026d. ‘Description of 6 Portfolios Formed on Size and Profitability for Developed Markets’. Kenneth R. French Data Library.