This article explains factor investing as a concept. It is educational and general, not personalised investment advice, and nothing here is a recommendation to buy or sell any security or to tilt a portfolio toward any factor. Where it references research findings, those describe widely-observed patterns; past performance does not predict future results.
For most of the twentieth century, a good return was assumed to be the mark of a good stock-picker. Then a body of research asked an awkward question: how much of any portfolio’s return is really just exposure to a handful of systematic traits, size, cheapness, recent strength, that any investor could have captured mechanically? The answer reshaped how the profession thinks about performance. Here is what factor investing is, the classic factors and the models that named them, what a factor premium actually is, and why every one of these factors will disappoint investors for years at a time.
What factor investing is
Factor investing is the practice of building a portfolio around systematic, measurable characteristics (“factors”) that have historically explained a large share of returns, rather than around bets on individual companies. The core insight is that much of what looks like stock-picking skill is really exposure to a few broad, repeatable traits, and that those traits can be isolated, measured, and harvested on purpose. Instead of asking “is this a good company,” a factor approach asks “which underlying drivers is this portfolio exposed to, and are those the ones I meant to hold.”
The word to hold onto is systematic. A factor is a characteristic that spans the whole market, small companies as a group, cheap ones as a group, recent winners as a group, and that has been associated with a persistent difference in return. The promise of the framework is decomposition: take a portfolio, and instead of one opaque number, express its return as the sum of its exposures to known factors plus whatever is left over. That leftover, the part no factor explains, is the only piece that can honestly be called selection skill.
How the idea grew: from one factor to five (plus momentum)
The lineage is worth knowing because the models are named after the people who built them. The starting point was the Capital Asset Pricing Model, which said a stock’s expected return was explained by a single factor: its sensitivity to the overall market (its “beta”). Everything else was supposed to be noise. The trouble was that the noise had structure.
In 1992 and 1993, Eugene Fama and Kenneth French showed that two more characteristics carried explanatory power the market factor missed: size (small companies tended to outperform large ones) and value (cheap companies, measured by a high book-to-price ratio, tended to outperform expensive ones). Their three-factor model, market, size, value, explained the cross-section of returns far better than the market alone. Then in 1997 Mark Carhart added a fourth: momentum, the persistence-of-winners effect documented by Jegadeesh and Titman, giving the widely-used four-factor model. Two decades later Fama and French extended their own work to a five-factor model, keeping market, size and value and adding profitability (a proxy for quality) and investment (firms that grow their assets conservatively). Each step was the same move: find a trait that explains returns the existing factors do not.
The classic factors, at a glance
Practitioners have converged on a handful of factors that show up again and again across markets and eras. Each is a plain idea with a rough-and-ready way to measure it (the “proxy”):
| Factor | The idea | A common proxy |
|---|---|---|
| Size | Smaller companies have tended to out-earn larger ones over the long run. | Market capitalisation (small minus big). |
| Value | Cheap companies have tended to out-earn expensive ones. | Book-to-price, earnings or cash-flow yield (high minus low). |
| Momentum | Recent winners have tended to keep outperforming over intermediate horizons. | Trailing return, conventionally the last three to twelve months. |
| Quality | Profitable, stable, low-debt firms have tended to out-earn weak ones. | Return on equity, gross profitability, earnings stability. |
| Low volatility | Calmer stocks have delivered surprisingly strong risk-adjusted returns. | Trailing volatility or market beta (low minus high). |
Two of these deserve a note. Low volatility is the awkward one, because a simple risk-and-reward view says calmer stocks should return less, not more; that they have historically delivered strong returns per unit of risk is a genuine anomaly, often attributed to a preference among some investors for lottery-like, high-volatility names, which leaves the quiet ones underpriced. Momentum is the one most likely to be familiar and is covered in depth in what is momentum investing; it stands out from the others because it is a price-based trait that rotates quickly, while size, value and quality are slower, fundamentals-based tilts.
What a “factor premium” actually is
A factor premium is the extra return, on average and over long windows, that has historically accrued to holding one end of a factor rather than the other, cheap over expensive, winners over losers, small over large. The academic way to measure it is a long-short spread: notionally buy the stocks with the most of a trait and sell the ones with the least, and the return on that spread is the premium. It is important to read that as a statement about a large population over a long horizon, not a property of any single stock. A value premium means only that, averaged across thousands of names and many years, the cheap basket has tended to come out ahead.
Why the premia exist is contested, and the honest framing matters. In one camp, factors are compensation for risk: value stocks are cheap because they are genuinely more fragile, and the premium is the fair reward for bearing that fragility. In the other, factors are mispricing: behavioural biases and institutional frictions push prices away from fair value, and the premium is what an investor earns for leaning against the crowd. As with momentum, the safest assumption is that most factors are part risk and part mispricing, which has a sharp consequence: a premium that is partly risk can and will hurt, and a premium that is partly mispricing can shrink as more capital chases it.
Why factors are cyclical (the part the brochures rush past)
Here is the single most important thing to understand about factor investing, and the one most likely to be soft-pedalled: every factor goes through long stretches of underperformance. These are not brief wobbles. Value spent much of the 2010s badly lagging the broad market, a drought long enough that serious people asked in print whether the value premium was dead. Size has been inconsistent enough that its very existence is debated. Momentum, as its own literature stresses, suffers rare but violent crashes at market turning points. A factor is a bet that a particular characteristic will be rewarded, and characteristics fall in and out of favour with the regime.
This cyclicality is the reason the premia can persist at all, which makes it a poor candidate for engineering away. If a factor paid off smoothly and reliably, capital would flood in until the premium vanished. The discomfort of the lean years is, in part, what stops everyone from crowding the trade, which is why factor investing is often described as requiring a tolerance for looking wrong for uncomfortably long periods. It also explains the appeal of holding several lowly-correlated factors together: value and momentum, for instance, have historically zigged and zagged at different times, so a blend can smooth the ride that any single factor delivers. The cyclicality never disappears; it can only be diversified.
Factors versus stock-picking: the real contribution
The lasting contribution of factor research is a ruler. Once a portfolio’s returns are regressed on the known factors, one can ask a demanding question of any manager or any system: after accounting for cheap, mechanical exposure to size, value, momentum, quality and the rest, is there any return left over? That leftover, the “alpha” in the regression, is the only part that is not simply factor exposure repackaged as skill. Plenty of strategies that looked brilliant turned out, under this lens, to be a value tilt or a momentum tilt an index fund could have delivered more cheaply. Factor analysis is how the profession tells genuine selection from expensive beta.
How a systematic process relates to factors
A rules-based publisher sits in an interesting spot relative to this framework, because the same regression that judges a fund can be turned on the system itself. Shishin’s published track-record analysis does exactly that: it regresses the book’s returns on the standard factors to ask what is really driving them. Two findings from that “edge-versus-beta” work are worth stating at a high level. First, the loadings on the classic factors come out near zero, and the market beta in particular is close to nil, which is a way of saying the returns are not simply a repackaged tilt toward size, value or the broad market. Second, the one place a relationship does appear is momentum, but a faster and smaller momentum than the textbook factor: the system leans on shorter-horizon relative strength than the conventional three-to-twelve-month definition, and in more concentrated names.
That is the honest connection, and it is a descriptive one, not a pitch. Factor analysis is the neutral ruler; running the book through it is simply the transparent thing to do, and the result places the system nearer the momentum anomaly than the slower value and size premia, while showing that a near-zero market beta means it is not just riding the index. How that record is assembled and independently verified is the subject of the public attestation, and the broader logic of harvesting relative strength is the subject of what is momentum investing.
The limits: proxies, crowding, and data-mining
Factor investing carries real caveats a careful reader should hold in mind. The factors are only as good as their proxies: book value, the classic yardstick for “value,” means less in an economy dominated by intangible assets it does not capture well, and a clumsy proxy can miss the very thing it is meant to measure. Crowding is a live worry: once a factor is famous and packaged into cheap products, more capital chases the same names, and a premium that was once the reward for discomfort can compress. And the research literature has a well-known data-mining problem, the so-called “factor zoo” of hundreds of published factors, most of which are probably statistical accidents that will not survive out of sample. The five or so classic factors earn their standing precisely because they have been replicated across countries, asset classes and time; the long tail of exotic factors mostly has not.
So, what is factor investing?
It is the recognition that a large share of returns comes from systematic exposures, size, value, momentum, quality, low volatility, rather than from individual stock picks, and the practice of measuring and harvesting those exposures deliberately. Fama and French named the first of them and built the three- and five-factor models; Carhart added momentum. The premia are real and widely replicated, but they are statistical tendencies, not promises, and every one of them will underperform for years at a stretch. The framework’s deepest use may not be as a strategy at all but as a lens: a way to separate genuine skill from cheap exposure dressed up as skill, which is exactly why a transparent system is happy to be held up to it.
Sources & further reading
- Fama, E. F. & French, K. R. (1993). “Common Risk Factors in the Returns on Stocks and Bonds.” Journal of Financial Economics, 33(1), 3 to 56 (the three-factor model: market, size, value).
- Fama, E. F. & French, K. R. (2015). “A Five-Factor Asset Pricing Model.” Journal of Financial Economics, 116(1), 1 to 22 (adds profitability and investment).
- Carhart, M. M. (1997). “On Persistence in Mutual Fund Performance.” Journal of Finance, 52(1), 57 to 82 (the momentum four-factor extension).
- Jegadeesh, N. & Titman, S. (1993). “Returns to Buying Winners and Selling Losers.” Journal of Finance, 48(1), 65 to 91 (the basis for the momentum factor). See also what is momentum investing.
- Shishin’s track-record and public attestation, the book regressed on these factors, with near-zero loadings and a faster, smaller momentum than the standard factor.