Risk Models Ch. 5: Factor Investing — Momentum, Value, and Low Volatility

Investment Background

This library already has a book that covers stock-level factors in detail: What Works on Wall Street dissects price-to-sales, momentum, and other cross-sectional characteristics with decades of backtested data. This chapter does not repeat that work. Instead, it asks a more foundational question that sits underneath any individual factor: what is a "factor," why do factor premia plausibly exist and persist, and what does institutional-grade factor investing actually look like — including the parts of it, like most of this book's subject matter, that are not appropriate to replicate directly in an individual retirement account.

A factor, in this context, is a stock characteristic that has historically been associated with different average returns than the broad market — not a single company's story, but a systematic tilt applied across hundreds or thousands of stocks at once.

The Wall Street Translation

The Three Factors This Chapter Covers

Factor	Definition	Historical premise
Value	Stocks cheap relative to fundamentals (low price-to-book, price-to-earnings, price-to-sales)	Cheap stocks have historically outperformed expensive stocks over long horizons
Momentum	Stocks that have recently outperformed tend to continue outperforming over the following months	Recent relative winners, on average, keep winning for a period before the effect fades
Low Volatility	Stocks with lower historical price volatility have delivered returns roughly comparable to, or in some periods better than, higher-volatility stocks	The opposite of what standard theory (more risk, more return) predicts — this is sometimes called the "low-volatility anomaly"

Each of these has decades of academic documentation across many markets and time periods, which is a meaningfully higher bar of evidence than most trading ideas ever clear — Fama and French's original work on value dates to the early 1990s, momentum research (Jegadeesh and Titman) to a similar period, and the low-volatility anomaly has been documented since the 1970s.

Why Factor Premia Plausibly Exist — Two Competing Explanations

The evidence that these patterns existed historically is not seriously disputed. Why they exist, and whether they will persist, is genuinely debated. Two broad camps:

1. Risk-based explanations: factor premia are compensation for bearing some risk that is not fully captured by market beta alone. Value stocks may be cheap because they carry genuine distress or cyclical risk; the premium is fair compensation, not a free lunch.

2. Behavioral explanations: factor premia exist because of persistent, difficult-to-arbitrage-away investor biases. Momentum may work because investors underreact to news initially, then overreact. Low-volatility may work because many investors chase excitement (lottery-like, high-volatility stocks) irrationally, bidding those up and leaving low-volatility stocks relatively cheap.

Both explanations likely hold partial truth for different factors, and the distinction matters practically: a risk-based premium should persist as long as the underlying risk is real, while a behavioral premium could shrink if enough capital arrives to exploit it (or persist if the bias driving it is deeply rooted in how humans evaluate risk and reward).

How Institutional Factor Investing Actually Works

Real factor strategies are more involved than "buy cheap stocks":

• Long-short construction: many institutional factor strategies go long the favorable side of a factor (cheap, high-momentum, low-vol stocks) and short the unfavorable side, isolating the factor's return from the broad market's return entirely.
• Multi-factor combination: rather than betting on one factor alone, sophisticated approaches blend several factors, since they are imperfectly correlated with each other and tend to underperform at different times — value often struggles when momentum thrives, and vice versa.
• Careful transaction cost management: factor premia, especially momentum, involve significant portfolio turnover, and trading costs can consume a meaningful fraction of the theoretical premium if not managed carefully.

This infrastructure — shorting, leverage in some implementations, sophisticated multi-factor weighting, transaction cost optimization — is exactly the kind of thing this book has repeatedly flagged as institutional machinery, not a retail replication project.

Executable Trading Rules

1. Understand that a single factor can underperform the broad market for a long time — sometimes over a decade — even if the long-run premium is real. Value underperformed growth for roughly the 2007-2020 period, testing the patience of even long-term believers. A factor tilt requires the same discipline this library repeatedly emphasizes for any strategy: judge the process over decades, not the outcome over quarters.

2. Do not chase a factor because it performed well recently. This is the same mistake as chasing any recently-hot asset — recent outperformance of a factor is weak evidence about its near-term future, and factor rotations are notoriously hard to time.

3. If you access factor exposure, prefer low-cost, broad, rules-based index products (factor ETFs) over attempting to construct long-short factor exposure yourself. The long-short, leveraged, actively-rebalanced version of factor investing belongs to institutions with the infrastructure to manage its costs and risks.

4. Recognize that a global market-cap-weighted index fund already captures a diversified blend of all stocks, including some exposure to whichever factors are currently in favor. Adding a deliberate factor tilt is a decision to deviate from that neutral starting point, and should be sized and justified accordingly, not adopted by default.

Relevance to a Retirement Portfolio

Factor investing's evidence base is stronger than almost any other "edge" discussed in this library, and it is also one of the more reasonable tactical tilts an individual investor could consider — within limits.

What the evidence supports	What it does not support
A long-run historical premium for value, momentum, and low-volatility tilts exists across markets and time periods	A guarantee that any specific factor outperforms over your specific investing horizon, which could easily span a period of factor underperformance
Low-cost, broad factor index funds are a reasonable way to express a modest tilt	Attempting to replicate institutional long-short, leveraged factor construction in a personal account
Combining factors reduces the risk of any single factor's prolonged underperformance dominating your results	Concentrating in one factor because it has outperformed recently

Consistent with our standard position: a factor tilt, if used at all, belongs as a modest supplement to a low-cost, globally diversified core — not a replacement for it. A reasonable, disciplined version of this might mean allocating a modest fraction of the equity sleeve to a broad, low-cost value or multi-factor index fund, sized small enough that a decade of factor underperformance (which has happened before and will happen again) does not meaningfully derail your plan.

The version of factor investing that does not belong in a retirement account is the institutional one: leveraged, long-short, actively timed factor rotation. That machinery requires infrastructure, cost management, and risk tolerance that individual retirement savers should not attempt to replicate.

Chapter 6 closes this book by asking the harder question directly: given everything in Chapters 1 through 5 — Kelly sizing, risk parity, correlation instability, and factor premia — what, specifically, should an individual retirement investor actually do with any of it?