Book-to-market (B/M, also written BE/ME for "book equity to market equity") is the ratio of a firm's accounting book value of common equity to its stock-market capitalization. It is the canonical academic measure of the value characteristic: a high B/M stock is "cheap" relative to its balance-sheet net worth, a low B/M stock is "expensive" (a growth stock). Its enduring importance comes from being the sorting variable Eugene Fama and Kenneth French chose to define the HML ("High Minus Low") value factor in their 1993 three-factor model. The core tension B/M embodies is whether the historically higher returns of high-B/M stocks are compensation for risk (the efficient-markets reading) or the correction of mispricing caused by investors over-extrapolating growth — a debate that is still unresolved.

How it's calculated / formed

The ratio itself is simple:

> B/M = Book value of common equity / Market capitalization

where book equity is the shareholders'-equity figure from the balance sheet (broadly, total assets minus total liabilities and preferred stock, with various researcher adjustments). It is the reciprocal of the more common price-to-book (P/B) multiple; academics prefer B/M because it stays well-defined when book equity is small or negative-adjacent, and it scales linearly.

The HML factor is built by sorting, not by the raw ratio. Fama–French (1993), as documented by Ken French's data library and summarized by Quantpedia, construct it via a 2×3 double sort:

• Sort the NYSE/AMEX/NASDAQ universe by size (small/big) using the NYSE median.
• Independently sort by B/M into three buckets using NYSE 30th/70th percentile breakpoints: Value (top 30%), Neutral, Growth (bottom 30%).
• Form six value-weighted portfolios; HML = ½(Small Value + Big Value) − ½(Small Growth + Big Growth).

This averaging of small and large legs deliberately strips out the size effect so HML is a "pure" value signal. Book equity is lagged (fiscal year-end book against market cap typically the following June) to avoid look-ahead bias, and portfolios are rebalanced annually.

How it's used in practice

Three distinct user groups consume B/M differently:

1. Quant factor portfolios. B/M is one of the standard inputs to a multi-factor value composite. Most practitioners do not use it alone — they blend it with earnings yield (E/P), cash-flow yield (CF/P), sales/EV, and shareholder yield, because no single value metric dominates across regimes (AQR's Fact, Fiction and Value Investing makes this point explicitly). B/M serves as the academic "reference" value definition against which others are benchmarked.

2. Academic asset pricing. B/M is a regressor, not a trade. Researchers run returns on the market, SMB (size) and HML factors to measure a manager's loadings and compute factor-adjusted alpha. Here B/M's role is descriptive: how much of a strategy's return is just exposure to cheap stocks.

3. Fundamental value investors use P/B (the inverse) as a screen, most meaningfully for asset-heavy, mark-to-market businesses — banks, insurers, REITs, shipping — where book value tracks economic value reasonably well. P/B is far less informative for asset-light firms.

Adoption, debate & evidence

The value premium tied to B/M is one of the most-studied phenomena in finance. Early evidence (Fama–French 1992/1993; Lakonishok, Shleifer & Vishny 1994) found high-B/M stocks earned materially higher average returns. Quantpedia reports the long/short HML factor returned roughly 3.6% annualized with a ~0.30 Sharpe over 1926–2014 — a modest but persistent premium, with a worst drawdown near −56% in the 1930s. The premium also appeared internationally and across asset classes, which strengthened the case it was not data-mined.

But the picture has darkened, and honesty requires emphasizing this:

• The "lost decade." Per Arnott, Harvey, Kalesnik & Linnainmaa (Reports of Value's Death May Be Greatly Exaggerated, 2020), the HML factor suffered a ~55% drawdown from 2007 to mid-2020 — its largest since 1963. They argue the entire drawdown is explained by the widening valuation spread (growth getting relatively more expensive), not by the premium disappearing, implying mean-reversion potential. Others (notably GMO, Research Affiliates) broadly agree value got cheaper, not dead.

• The intangibles critique. A growing literature (CFA Institute 2020; Alpha Architect; Bongaerts et al. in Financial Management, 2025) argues book equity has become a poor proxy for economic capital because R&D, brand and software are expensed, not capitalized under GAAP. This systematically understates the book value of knowledge-economy firms, making B/M misclassify them. Proposed fixes include capitalizing intangibles into adjusted book value, or replacing B/M with fundamental-equity or intangibles factors.

• Factor redundancy. Fama & French's own 2015 five-factor model found HML becomes largely redundant once profitability (RMW) and investment (CMA) factors are added — high-B/M stocks tend to load on those. This is a serious internal challenge: the original authors concluded their flagship value factor adds little incremental explanatory power.

So: the academic value premium (especially as a multi-metric, valuation-spread-aware strategy) retains credible support, but raw single-metric B/M specifically is the weakest and most contested member of the value family.

Strengths & limitations

Strengths: transparent, hard to game (book value is audited), stable through time, available cross-sectionally for decades, and the academic lingua franca for measuring value exposure. It works best where book value is economically meaningful — financials, real estate, capital-intensive industrials — and as one input within a blended value score.

Limitations / #1 misuse: treating B/M as a standalone buy signal. It is most distorted exactly where modern equity value is concentrated (tech, pharma, brands), is contaminated by buybacks and write-offs that depress book equity, and on its own has delivered long, painful drawdowns and arguably redundant signal post-2015. A high B/M can equally flag a deeply impaired "value trap" as a bargain — the ratio cannot tell the difference, which is why practitioners pair it with quality/profitability screens.

Sources

• Fama & French (1993), "Common Risk Factors in the Returns on Stocks and Bonds"; Fama & French (2015) five-factor model — summarized via Wikipedia: Fama–French three-factor model and Corporate Finance Institute.
• Quantpedia — Value (Book-to-Market) Factor (construction, breakpoints, ~3.6% return / 0.30 Sharpe figures).
• Arnott, Harvey, Kalesnik & Linnainmaa (2020), Reports of Value's Death May Be Greatly Exaggerated (SSRN) (the ~55% 2007–2020 drawdown, valuation-spread explanation).
• CFA Institute — The Value Factor's Pain: Are Intangibles to Blame? (2020) and Alpha Architect — Intangibles and the Performance of the Value Factor.
• Bongaerts et al., Revisiting Asset Pricing Models: The Case for an Intangibles Factor (Financial Management, 2025).

Flagged disputes: (1) risk-compensation vs. mispricing explanation of the premium is unresolved; (2) whether the value premium is "dead" or merely cheap is actively contested (Arnott et al. / GMO say cheap; skeptics say structurally impaired by intangibles); (3) Fama–French's own 2015 finding that HML is redundant is a genuine internal challenge to single-metric B/M.