Value Spread

How many times more investors pay for growth stocks than for value stocks, per dollar of book value, for U.S. small and large companies since 1926. Built from Ken French’s data and updated monthly. Free, no signup, and not a buy or sell signal.

Small caps

6.5x

Higher than in 82% of months since 1926

Estimate for September 2026

Large caps

10.1x

Higher than in 97% of months since 1926

Estimate for September 2026

All stocks, on earnings

4.3x

Higher than in 87% of years since 1950

Reading for December 2025 prices

A reading of 6.5 means the most expensive 30% of small companies cost 6.5 times as much per dollar of book value as the cheapest 30%. Source: Kenneth R. French Data Library (September 2026 data); Summitward calculations.

Where the spread stands

  • Small caps: 6.33 when French formed his portfolios in June 2026, against 8.09 a year earlier, and an estimated 6.52 in September 2026. That is higher than in 82% of months since 1926 and 92% since 1963. The median month since 1926 read 4.68. Dot-com peak 7.33 (February 2000); 2020 peak 9.42 (December 2020). The all-time highs came in the 1930s and 1940.
  • Large caps: 9.71 in June 2026, an estimated 10.06 in September 2026, higher than in 97% of months since 1926 and 99% since 1963. Dot-com peak 9.83 (June 2000); 2020 peak 11.41 (September 2020).
  • Same profitability: comparing cheap and expensive small companies within each of four operating-profitability groups gives 6.45 for June 2026, higher than in 86% of Junes since 1963, against 6.33 for all small caps. For large caps the same adjustment gives 8.33 against 9.71.
  • On earnings and book value, all stocks: at December 2025 prices the earnings-to-price spread was 4.35 (peaks 6.48 in December 1999 and 6.75 in December 2019), and the book-to-market spread was 9.90 (peaks 9.89 and 11.03). Book value leaves out internally built intangibles, which are concentrated in growth companies and can make them look pricier on book value than on earnings. The two measures have often disagreed since 1950.

Returns over the 12 months to September 2026 in French’s portfolios: large value +33.0%, large growth +10.4%, small value +18.5%, small growth +16.8%.

What a wide spread has predicted

For every June since 1926 with enough later data, the table records the starting spread and how much French’s value portfolio then beat or trailed growth, within the same size half, over the next five or ten years. The out-of-sample column asks whether a forecast built only from data available at each date beat the running historical average; a negative number means it did worse.

SizeHorizonNarrowest thirdWidest thirdValue trailed (widest)Out-of-sample R²
Small caps5 years+5.3% a year+4.8% a year38%-1.7%
Small caps10 years+6.0% a year+4.9% a year23%-14.3%
Large caps5 years+0.8% a year+3.9% a year38%-4.2%
Large caps10 years+1.0% a year+3.2% a year37%-6.1%

Average annualized value-minus-growth total return after starting points in the narrowest and widest third of spreads, June 1926 to June 2021 for five years. Windows overlap, so there are only 19 independent five-year periods; the thirds are ranked with hindsight. Source: Kenneth R. French Data Library; Summitward calculations.

Starting from a wide spread, small value’s later edge over small growth has been about the same as from a narrow one. Large caps show a relationship in the full sample (Newey-West t-statistic 2.15 at ten years) that did not survive out of sample. Li, Ng and Swaminathan find the plain value spread fails out of sample in month-ahead forecasts in their 2026 JFQA paper; a measure built from analysts’ earnings forecasts did better. The spread tells you how value is priced. It has not told investors when value would pay.

Method

Definition. Value spread = book-to-market of the value portfolio divided by book-to-market of the growth portfolio, which equals growth’s price-to-book divided by value’s. Small and large caps use French’s six portfolios formed on size and book-to-market: value is the top 30% of book-to-market within each size half, growth the bottom 30%, value-weighted.

Monthly readings. French re-forms the portfolios each June and reports their book-to-market at June prices. Between Junes, Summitward rolls each portfolio’s book-to-market forward on its value-weighted return without dividends, holding book value per share fixed. At each re-sort the reading moves as new book values arrive and stocks change groups: by a median of 7.8% for small caps and 6.2% for large caps across 100 re-sorts. The 2026 re-sort moved the large-cap reading from 7.74 (last year’s portfolios, rolled forward) to 9.71.

Other series. The profitability-controlled spread uses the 32 portfolios formed on size, book-to-market and operating profitability (from 1963): the highest over the lowest book-to-market quarter within each profitability quarter, averaged geometrically. The all-stock series use the univariate sorts on book-to-market, earnings-to-price and cash-flow-to-price (top over bottom 30%, December prices).

Limits. U.S. stocks only; French’s international files do not include valuation ratios. Book value omits most intangible capital. These are portfolio averages, not a fund’s holdings, and French’s monthly updates can revise history. Percentiles are empirical ranks over the full history, which includes the current reading.

Data: Kenneth R. French Data Library, mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html (September 2026 data, generated 2026-11-15). Script and method: summitward-research/value-spread.

Frequently asked questions

What is the value spread?

The price-to-book multiple of growth stocks divided by that of value stocks. A higher reading means value stocks are cheaper relative to growth stocks than usual.

Is a high value spread a signal to buy value funds?

Not on the evidence above. Across nearly a century of June readings, wide small-cap spreads were followed by about the same small-cap value edge as narrow ones. Wide large-cap spreads were followed by a larger edge in the full sample, but forecasts built only from data available at the time did worse than the historical average for both sizes.

Why do the book-value and earnings spreads disagree?

Book value excludes internally developed intangibles such as research and brands. Growth companies hold more of them, so their price-to-book looks higher than their price-to-earnings would suggest.

How often is this updated?

Monthly, after Ken French posts new data. The current figures use his September 2026 data.

Does this cover international stocks?

No. French’s developed and emerging market files publish returns but not the valuation ratios this page needs.

Read the analysis

Is the Value Premium Back? uses this data to explain the 2025–26 run in AVUV, AVDV and DFSV, what drove it, and what it does and does not imply.

Disclaimer: This tool is for educational and informational purposes only and is not financial, investment, tax, or legal advice. Summitward is not a registered investment adviser, broker-dealer, or financial planner, and no fiduciary relationship is created by your use of it. Consult a qualified professional before acting. Past performance and model projections do not guarantee future results. Provided as is, without warranty of any kind; see our Terms of Service for limitations of liability.