Factor investing

Here is a disquieting possibility you must confront before you commit your life to picking stocks: much of what feels like skilled stock selection may be a systematic tilt toward a handful of well-documented "factors": tilts you could rent through a cheap ETF for a fraction of the effort. Understanding factors is therefore not a detour into quant finance; it is how you find out whether your edge is real. What follows explains the five best-documented factors, the honest evidence for each, and, most important for you, how to diagnose your own portfolio's implicit factor tilts.

What is a factor, and which five matter?

What a factor is. A factor is a characteristic shared across many stocks that has historically been associated with a return premium: a systematic reason a group of stocks tends to outperform (or underperform), distinct from anything specific to one company. The research tradition runs from the Capital Asset Pricing Model's single factor (market beta, M3.02) through Fama & French's three-factor model (1993: market, size, value), Carhart's addition of momentum (1997), to Fama & French's five-factor model (2015: adding profitability and investment). So the modern "factor zoo" has a well-established core of roughly five or six tilts that explain most of the cross-section of returns. The five you must know:

• Value. Cheap stocks (low price relative to book value, earnings, or cash flow) have historically outperformed expensive ones. This is Graham's whole edifice expressed as a factor (the academic proxy is HML, "high minus low" book-to-market). Evidence and caveat: the long-run US value premium was historically ~3–5%/year (Fama-French data back to 1927), but value endured a brutal lost decade from roughly 2007 to 2020 when growth dominated, prompting a genuine debate (is the premium dead, arbitraged away after publication, or just cyclical?) before value came roaring back in 2021–2022. The honest read: value is real over very long horizons but painfully cyclical, and its magnitude is contested.
• Quality. Profitable, stable, low-debt companies with high and durable returns on capital have outperformed "junk." This is the Buffett/Akre/Terry Smith compounder philosophy (M6.02) expressed as a factor (proxies: Novy-Marx's gross profitability; the QMJ, "quality minus junk," factor of Asness-Frazzini-Pedersen; Fama-French's RMW, "robust minus weak" profitability). Evidence and caveat: quality has been one of the more robust and intuitive premia (good businesses that are also not wildly overpriced tend to do well), but "quality" has many definitions and some of its return is really low-volatility or value in disguise. Definition matters enormously.
• Momentum. Stocks that have outperformed over the past ~3–12 months tend to keep outperforming over the next few months (Jegadeesh & Titman 1993; the academic proxy is UMD, "up minus down," typically the 12-month return skipping the most recent month). Evidence and caveat: momentum is statistically the strongest and most pervasive factor, showing up in nearly every market and asset class studied, but it is treacherous to harvest: it has high turnover (costly), and it suffers rare, savage momentum crashes (e.g., 2009, when beaten-down stocks violently rebounded and momentum strategies were long the wrong names). It is also the factor most at odds with a fundamental investor's temperament, since it is pure price behavior with no valuation anchor.
• Size. Small-cap stocks have historically outperformed large-caps (Banz 1981; the SMB, "small minus big," factor). Evidence and caveat: this is the weakest and most fragile of the five. Much of the historical small-cap premium disappears once you control for quality (small caps are disproportionately junky) and much of it was concentrated in tiny, illiquid micro-caps and in January; post-publication it has been faint. Treat a raw "small-cap premium" with real skepticism. But note (relevant to active vs passive investing) that small caps are less efficiently priced, which is a different and more durable source of edge than a mechanical size premium.
• Low-volatility (low-beta). Stocks with lower volatility and lower beta have delivered higher risk-adjusted returns than the CAPM predicts, and sometimes higher absolute returns than high-beta stocks (the "low-volatility anomaly"; Frazzini & Pedersen's "betting against beta," BAB). Evidence and caveat: this directly contradicts the CAPM's "more beta, more return" (M3.02) and is one of the more robust anomalies, plausibly driven by leverage-constrained investors bidding up exciting high-beta names and neglecting boring low-beta ones. Much of "quality" and "low-vol" overlap.

Why do factors go through long droughts?

Factor cyclicality: why this matters for your temperament. No factor works all the time; each endures long stretches of underperformance that would get a human manager fired. Value's 2007–2020 drought, momentum's 2009 crash, size's multi-decade fade: these are not bugs, they are the reason the premia might persist (a factor that never hurt would be arbitraged to zero; the pain is what stops everyone from harvesting it). The practical consequence: when your portfolio lags, you must be able to tell "my stock selection was wrong" from "my factor tilt is in a cyclical headwind", because the responses are opposite. The first calls for changing your process; the second calls for holding your process through the pain that most people can't. Which brings us to the sharpest use of factors for a bottom-up investor.

How do you find your own factor tilts?

Diagnosing your own implicit tilts: "are you just a factor?" Here is the question that separates a self-aware analyst from a deluded one. Take your portfolio and ask what factors it is implicitly loaded on. A classic Buffett-style book of high-ROIC, wide-moat, reasonably-priced compounders (HDFC Bank, Asian Paints, TCS, Titan; Costco, Visa, Moody's) is, whether you call it that or not, heavily long the quality and low-volatility factors, short the value factor (you're paying up for quality), and short the size factor (they're large-caps). That diagnosis has two uncomfortable, essential consequences:

1. Much of your return may be rentable. If a quality + low-vol factor ETF captures most of what your carefully-researched compounder book delivers, then your active effort is only earning its keep to the extent you beat that factor blend, not to the extent you beat the broad index. The right benchmark for a quality-tilted stock-picker is not the Nifty 50 or the S&P 500; it is a quality/low-vol factor benchmark, and beating it is the honest bar for "do I have stock-selection alpha beyond my tilt?"
2. You will have factor-driven bad years that are not your fault, and factor-driven good years that are not your skill. When value screams back (2021–2022) your quality book will lag for reasons that have nothing to do with your company analysis; if you fire yourself or abandon the approach at that moment, you convert a factor headwind into a permanent loss. Conversely, in a quality-led melt-up, don't mistake the tailwind for genius. Knowing your tilts is what lets you keep score honestly and hold your process when it matters most.

What should you do with the factor evidence?

The mature stance is not to abandon stock-picking for factors, nor to ignore factors and pretend all your return is skill. It is to know your tilts, benchmark against them, and locate your true edge in what remains: the business-specific judgment, the moat durability, the management read, the mispricing, that no mechanical factor captures. If nothing honestly remains, if your returns are fully explained by a cheap factor blend, then intellectual honesty (and active vs passive investing) says buy the factor blend and reclaim your time.