The Value Factor
The tendency for cheap securities to outperform expensive ones over long periods. One of the most studied effects in finance, and one that spent a decade not working.
MadStockAlerts Research · Updated August 29, 2026
What to take away
- Value ranks securities by a price-to-fundamentals measure and buys the cheap end.
- The effect is documented across markets and across long histories.
- The measure used changes the result, and book value has weakened as a metric.
- It underperformed for an extended period after 2007, which tested every explanation.
- The two competing explanations are risk compensation and behavioural error.
MAD Academy Training Video · 0:46
Cheap for a Reason, or Cheap by Mistake
The value premium is one of the most studied results in finance, and it has spent long stretches not working at all.
This lesson is part of a Stock Alerts + Tools plan.
The construction
A value factor sorts a universe by a ratio of price to some fundamental measure, then compares the returns of the cheapest group against the most expensive. The original academic construction used book value to market value, and many others are in use.
| Measure | Character |
|---|---|
| Book to price | The academic standard, and increasingly criticised |
| Earnings to price | Intuitive, and undefined for loss-making companies |
| Free cash flow to price | Harder to manipulate, and lumpier |
| Sales to price | Defined for everyone, and blind to profitability |
| EV to EBITDA | Accounts for capital structure, and standard in practice |
The choice is not a detail. The same universe sorted by book value and by free cash flow produces substantially different portfolios, and studies of the factor's performance depend heavily on which was used.
The evidence, and the drawdown
Value's long-run premium is documented across decades, across countries and across asset classes, which is unusual among claimed anomalies. It is also one of the clearest cases of an effect that can disappear for long enough to end most people's patience.
From roughly 2007, value underperformed growth for well over a decade in developed markets, in what was among the deepest and longest drawdowns the factor has recorded. Explanations offered range from the effect having been arbitraged away to book value having become a poor measure of a company's assets.
The episode is worth knowing for a reason beyond value itself. A factor with a century of supporting evidence spent more than a decade not working, which is the practical meaning of the phrase long horizon.
Scroll the chart sideways to see all of it.
The two explanations
| Risk-based | Behavioural | |
|---|---|---|
| The claim | Cheap companies are riskier, and the premium is compensation | Investors extrapolate and overpay for growth |
| Implication | The premium should persist, since the risk does | It may shrink as it becomes known |
| Supporting evidence | Value stocks are often distressed and cyclical | Documented extrapolation in expectations data |
| Difficulty | Identifying the risk being compensated | Explaining why arbitrage does not remove it |
The debate is unresolved and both mechanisms probably operate. The practical relevance is the second row: if the premium is compensation for risk, it should survive being known, and if it is an error, publication should erode it.
Why book value weakened
The most substantive criticism of the classic measure is the one made in the valuation pillar: internally generated intangibles are expensed rather than capitalised, so a company that built its assets shows little book value while one that bought them shows a lot.
As the listed universe shifted toward businesses whose value lies in software, brands and data, the measure classified a growing number of genuinely profitable companies as expensive and a growing number of asset-heavy ones as cheap.
Research adjusting book value for capitalised intangibles has found improved results, which is evidence that at least part of the drawdown was a measurement problem rather than the disappearance of an effect.
Implementing it changes it
Between a documented factor and a portfolio sits a set of construction choices, each of which materially changes the result. Two funds tracking the same factor can hold substantially different portfolios.
| Choice | Effect |
|---|---|
| The measure | Book value and cash flow produce different portfolios |
| The universe | Whether small companies are included changes the result substantially |
| Weighting | Capitalisation-weighted against equal-weighted within the selection |
| Rebalancing frequency | More frequent means more turnover and more cost |
| Sector constraints | Unconstrained value concentrates in whichever sectors are cheap |
| Quality screens | Removing the weakest companies changes the factor's character |
The fifth row is the one with the largest effect. Unconstrained value regularly becomes a concentrated bet on one or two sectors, because sectors become cheap together, and that concentration is a different risk from the one the factor was documented on.