Fama-French Dashboard
Analyzing equity value spreads across multiple definitions
Last updated: August 31, 2026
HML FF3
High Minus Low factor from Fama-French 3-Factor model
All Value Spread Measures at a Glance
The table compares the latest monthly return for every series as of August 31, 2026. Each percentile ranks that return within the same series' recorded history. A low rank means a relatively low monthly return. The z-score measures its distance from the historical mean in standard deviations. Neither statistic measures the underlying stocks' valuation.
| Measure | Latest | Percentile | Z-Score | Mean | Std Dev | Min | Max |
|---|---|---|---|---|---|---|---|
| HML (FF3)High Minus Low factor from Fama-French 3-Factor model | -3.54% | 8.0Rank within recorded months: 8.0 out of 100 | -1.10 | 0.35% | 3.55% | -13.83% | 35.52% |
| HML (FF5)High Minus Low factor from Fama-French 5-Factor model | -3.54% | 7.2Rank within recorded months: 7.2 out of 100 | -1.29 | 0.29% | 2.97% | -13.83% | 12.86% |
| B/M Portfolio SpreadValue minus Growth spread from B/M sorted portfolios | -3.99% | 12.3Rank within recorded months: 12.3 out of 100 | -0.96 | 0.35% | 4.52% | -18.75% | 36.55% |
| Small Value - Small GrowthSmall Value minus Small Growth (from 25 Size-BM portfolios) | -2.00% | 23.8Rank within recorded months: 23.8 out of 100 | -0.38 | 0.78% | 7.29% | -93.79% | 50.14% |
| Big Value - Big GrowthBig Value minus Big Growth (from 25 Size-BM portfolios) | -1.36% | 34.9Rank within recorded months: 34.9 out of 100 | -0.28 | 0.32% | 5.94% | -27.81% | 71.45% |
| Small Value - Big ValueSmall Value minus Big Value (size effect within value stocks) | 2.84% | 75.0Rank within recorded months: 75.0 out of 100 | 0.41 | 0.34% | 6.06% | -43.99% | 69.18% |
| E/P SpreadValue minus Growth spread based on Earnings/Price ratio | -4.36% | 9.5Rank within recorded months: 9.5 out of 100 | -1.16 | 0.30% | 4.02% | -19.06% | 17.22% |
| CF/P SpreadValue minus Growth spread based on Cash Flow/Price ratio | -3.22% | 14.6Rank within recorded months: 14.6 out of 100 | -0.88 | 0.23% | 3.92% | -19.27% | 18.57% |
| D/P SpreadValue minus Growth spread based on Dividend/Price ratio | 0.03% | 48.8Rank within recorded months: 48.8 out of 100 | -0.02 | 0.10% | 4.44% | -18.63% | 31.93% |
| Momentum (UMD)Momentum factor (winners minus losers based on prior 2-12 month returns) | -5.70% | 5.5Rank within recorded months: 5.5 out of 100 | -1.34 | 0.61% | 4.70% | -52.61% | 18.05% |
How to read these numbers
Each row compares portfolio returns using a different sort. High Minus Low (HML) from the Fama-French three- and five-factor models uses book-to-market. Earnings-to-price (E/P), cash-flow-to-price (CF/P), and dividend-to-price (D/P) series use those ratios to form portfolios, then compare their returns. The size comparisons distinguish smaller and larger companies. Momentum compares recent winners with recent losers.
All of these readings are monthly returns, including the series named after valuation ratios. A percentile and a z-score put a month in context; they do not make that month a more reliable forecast. The historical chart shows how returns varied, including periods of repeated underperformance. Low ranks describe the lower end of the observed return distribution and do not establish that a reversal is due. This is historical analysis, not investment advice.
Historical Time Series
HML FF3
Last observation: August 31, 2026.
Source: Kenneth R. French Data Library and famafrenchdashboard.com
Methodology: Monthly returns from Fama-French factor model (High Minus Low)
12-Month Rolling Returns
HML FF3 - 12-Month Rolling Annualized Returns
Last observation: August 31, 2026.
Source: Kenneth R. French Data Library and famafrenchdashboard.com
Methodology: 12-month rolling annualized returns calculated from monthly returns using compound annual growth rate
Key Insights: HML FF3
Understanding Value Spread Analytics
What does a value spread measure?
On this dashboard, a value spread measures the return difference between portfolios of value and growth stocks during a month. Value portfolios hold stocks with higher fundamentals relative to their prices, such as higher book equity per dollar of market value. Growth portfolios hold stocks at the other end of that sort. Subtracting one portfolio return from the other shows how those groups performed relative to each other. The result is a monthly return, expressed as a percentage.
A positive value spread means the value portfolio earned a higher return than the growth portfolio that month. A negative spread means growth did better. Both portfolios can rise, or both can fall, while producing either sign. The difference alone does not reveal the return of either portfolio separately. This dashboard follows several portfolio definitions so readers can compare how the choice of sorting characteristic changes the measured return difference.
Reading a return spread in context helps explain when a factor contributed to performance and when it detracted. It does not establish what the underlying stocks are worth today. A strong month can reflect many events, and a weak month does not establish that a recovery is due. Use the history to study the range and variability of realized returns, alongside research on portfolio construction and the limits of applying academic results to investments.
Understanding the Different Definitions of Value Spread
The dashboard shows several return series because portfolio construction affects the result. High Minus Low (HML) subtracts growth portfolio returns from value portfolio returns within size groups. The Fama-French three-factor (FF3) and five-factor (FF5) datasets supply their respective HML series. Comparing them shows the readings available in each source history. The model names describe the datasets; a displayed HML observation remains a monthly return difference in either model.
Other series sort portfolios using book-to-market (B/M), earnings-to-price (E/P), cash-flow-to-price (CF/P), and dividend-to-price (D/P). These ratios determine which stocks enter each portfolio. The plotted series then subtracts the low-ratio portfolio return from the high-ratio portfolio return. The ratios themselves are not plotted. A company can rank differently on earnings and dividends, so these portfolio sorts need not produce the same monthly return or historical percentile.
The size comparisons include Small Value minus Small Growth (SV-SG) and Big Value minus Big Growth (BV-BG). They measure value-versus-growth returns separately within smaller and larger companies. Small Value minus Big Value compares company sizes within the value group. Momentum, also called Up Minus Down (UMD), compares recent winners with recent losers. Its percentile ranks that return difference within momentum history, so it should not be read as a statement about value stocks.
Each sorting characteristic has limits. Book equity can omit parts of a business built on intangible assets. Earnings and cash flow can change with the business cycle and accounting choices, while dividends depend on payout policy. Those differences affect portfolio membership and, in turn, portfolio returns. Comparing the series can show whether a month was strong across several definitions or only one. Agreement among them still does not supply a measure of current stock valuations.
How to read high and low monthly returns
Percentile ranks put the latest monthly return in the context of the same series' recorded months. A reading near the 90th percentile places that return toward the high end of its history; a reading near the 10th places it toward the low end. The ranking describes realized performance. It does not measure the price of the underlying stocks relative to their fundamentals, and it does not predict which group will perform better next month.
The z-score measures the distance between the latest return and the historical mean in standard deviations. A positive score means the return is above that mean; a negative score means it is below. Larger absolute scores describe a greater departure from the average. Monthly factor returns need not follow a normal distribution, so the score alone does not establish a probability or a statistically significant forecast. It also does not turn a return observation into a valuation measure.
The reference sample matters when comparing percentiles. This dashboard uses each series' full available history for its headline statistics, and source histories begin at different dates. Changing the chart range changes the visible dates; it does not redefine the headline percentile. A separately calculated ranking over a recent decade could differ from the full-history ranking. Always identify the sample before comparing two readings, especially when the series cover different periods.
Rolling returns combine monthly observations across a selected window. They help distinguish a strong individual month from performance sustained over a longer period. Positive rolling returns describe a gain in the compounded series over that window, while negative returns describe a loss. Nearby windows share many observations, so a run of positive readings is not a set of independent findings. Use the rolling chart alongside monthly history to examine how performance varied over time.
Important Limitations and Considerations When Using This Dashboard
The figures describe academic portfolio returns. They do not represent an account that a reader could have held with identical results. Trading costs, taxes, financing, portfolio access, and implementation choices can change realized investment outcomes. Source data can also be revised. Before reproducing a calculation, check its dates, portfolio definition, and return units against the methodology and source files. A close historical match does not establish that the same result would be available in practice.
These observations look backward. A high return spread records a strong month for one portfolio relative to another; a low reading records a weaker month. Neither identifies a future turning point. Repeated losses can occur, and an average calculated across the entire history can conceal long stretches of different performance. Describing a return as unusual is useful for historical research, but a trading rule would need its own evidence and an evaluation that avoids using future information.
The dashboard draws on United States equity research portfolios with specific sorting and weighting rules. Results can differ for other markets, different definitions, or portfolios constrained by liquidity and trading costs. The spread between two groups also leaves out much of their underlying exposure. A value portfolio can have a positive relative return while losing money in absolute terms. Understanding an investment therefore requires more than the return difference displayed here.
Businesses, accounting practices, and market conditions change over time. A long historical sample provides context but does not make each month directly comparable with every earlier month. Read the percentile as a summary of the available observations, and use the source and methodology pages to examine what those observations contain. The dashboard supports research into realized performance. Decisions about future returns require evidence beyond a historical rank and attention to the risks and costs of an actual portfolio.