THE EQUITY FACTOR BRIEFING

What Works in U.S. Equities?

Choose a time frame

THE RESULTS, FIRST

Loading the leaders and laggards.

This briefing compares U.S. factor returns over the same dates, then looks inside the signals that led and lagged.

Loading published daily research.

Core and extended factors are ranked together by Sharpe. These are historical synthetic returns, before trading, borrowing and financing costs; a high rank is not a forecast.

What the numbers are saying

FOLLOW THE EVIDENCE

Look inside the drivers of return.

Explore quality, value or credit risk. Compare their ingredients, select any combination of core and extended factors, and follow the evidence through returns, Sharpe and correlations. All views use the time frame selected above.

Choose individual factors

Cumulative factor returns

Return scale

Cumulative factor returnsLoading the historical factor data.

Compounded from daily research returns. Market and style factors have different risk profiles; the definitions explain each exposure.

Explore the other side of a factor

Flip individual factors with the reverse-arrow button in the chart legend or factor details. Changing the overall direction clears manual flips.

The selected factors, side by side.

Compounded return, risk and coverage over the selected dates. Open a factor to read its definition and inspect its latest composition.

A LEADER TODAY. WHAT ABOUT THE LONGER RUN?

Does the edge survive a different window?

Compare risk-adjusted returns across time frames. A signal that leads one window may tell a different story in another.

Absolute Sharpe shows magnitude; color still shows the signed result. In the historically stronger view, each lookback chooses its own sign using hindsight. A manual flip applies to every lookback.

DIFFERENT LABELS. DIFFERENT EXPOSURES?

See which factors move together.

Correlation helps reveal shared drivers behind the names. Explore the selected factors as a heatmap, then open any pair to see the daily-return relationship.

−1 · Move in opposite directions0 · No linear relationship+1 · Move together

Pearson correlation uses dates with a return for both factors, with at least 20 overlapping observations. It includes the disclosed zero fills and reflects the displayed factor directions. Correlation is historical and does not establish causation.

BEYOND THE LEADERBOARD

Three findings.
More useful questions.

THE FULL PICTURE

Every loaded factor.
One comparable period.

The leaders and laggards are useful starting points. The full ranking lets you test the story. Click any column header again to reverse the order.

All loaded factors and their coverage over the same dates
#DirectionCompare

Rankings require at least 21 complete observations. All loaded factors remain visible; incomplete histories are marked unranked. Government bonds, Low volatility, Low volatility beyond beta and U.S. dollar use 0 returns for missing sessions, including dates outside their published histories. Filled counts are shown; other gaps remain missing.

KNOW WHAT YOU ARE LOOKING AT

What’s actually in a factor?

Read the definition, see what is long and short, and inspect the latest stored weights. Current compositions describe the latest published weights, rather than the historical holdings behind the entire return chart.

HOW TO READ THIS BRIEFING

Evidence you can inspect.
Conclusions you can question.

These are research exposures.

Most stock-style factors compare long and short baskets, adjust for the broad market, and scale risk using lagged information. Market and macro factors have different constructions. These are synthetic histories, not fund returns or a cost-free investment.

A ranking needs a common yardstick.

Sharpe measures mean daily return relative to daily volatility, annualized using 252 sessions. Every ranked factor uses exactly the same dates. Short or incomplete histories remain visible as unranked, with an explanation. Related factors can share exposures; they are not independent discoveries.

Direction changes the question.

The briefing starts in the published directions so losses remain visible. The optional stronger-direction view selects the better compounded result between the published series and its daily sign reversal. That uses hindsight, can change with the dates, and is not a trading rule.

The source stays visible.

Newly published daily data can take up to 12 hours to appear here. The page checks for updates every five minutes. If the initial request fails, it shows an unavailable message. If a later refresh fails, the last successfully loaded data remain visible with their date and a refresh notice. Selection, survivorship, revised inputs, trading costs and financing can change how much of a historical result an investor could realize.

Exact calculation notes and downloads

Cumulative return is the product of one plus each displayed daily return, minus one. Annualized return uses a 252-session year. Volatility is sample daily standard deviation times the square root of 252. Sharpe is daily mean divided by daily sample standard deviation, times the square root of 252; no risk-free series is subtracted. Positive days are the share of sessions with a strictly positive return.

Drawdown is the loss from the running high, including starting capital immediately before the first observation. For Government bonds, Low volatility, Low volatility beyond beta and U.S. dollar, missing daily returns are filled with 0 before calculating paths, rankings, Sharpe, volatility and correlations. This includes pre-inception and trailing missing sessions; it is an assumption, not an observed return. The next published bond return is retained. Zero filling changes the daily sample and can change risk statistics. Other missing or invalid returns break the path and suppress full-period statistics. Displayed-return downloads include the fills; all source histories retain the original gaps. Findings and rankings require at least 21 sessions and complete data over the displayed interval. Full history uses the overlap of all loaded inception and latest-observation dates, instead of requiring every factor to start with the market. Internal gaps in other factors still leave them unranked. Use View full factor history in a factor’s details to see its earlier individual observations.

Inversion negates each daily return before compounding and reverses instrument and hedge weights. It does not negate the final cumulative return. Both directions can lose through volatility drag. Five complete observations are required to choose a historical sign; ties keep the published sign.

Stored weights include scaling and can exceed 100% gross exposure. Their timestamp is separate from the latest return date. They are current exposures, not historical holdings or recommendations.

Latest stored composition

Positive weights are long; negative weights are short. Scaling and hedges can make gross exposure exceed 100%.