Meridian Indices Research — Profitability Factor Robustness Dashboard

Historical research, not live data. Every number on this page comes from external long-short research portfolios built on the Tidy Finance Factor Library (CRSP/Compustat, 1960–2024) plus a MarketBrain pilot aggregation. This is not Meridian index performance, not a live trading signal, and not a backtest of any Meridian product. Coverage ends December 2024 and nothing here updates going forward. See the companion article for the full write-up, or the definitions & caveats below.

Dispersion by era, all five factors

Each marker is one factor’s median annualized return across every tested specification in that era; the line spans the 10th–90th percentile of outcomes. A short, high line means the definition worked reliably across construction choices in that period — a long line means results were sensitive to how the factor was built. Factors are always listed in the same fixed order below; nothing is sorted by performance.

Return weighting Capped value-weighted is the closest available analogue to Meridian’s own concentration controls, so it is the default.

Median and 10th–90th percentile annualized return, by factor and era, for the selected return-weighting method. Data: MarketBrain factor-library pilot summary, 86,400 specifications.

Data table (accessible alternative to the chart above)

Median, 10th percentile, and 90th percentile annualized return by factor and era for the selected return-weighting method.
Era	Factor	Median return	P10	P90	% specs positive	Series count

Factor spotlight

Pick one factor, era, and return-weighting method to see its full robustness profile.

Factor Era

Methodology sensitivity

Full-history, capped value-weighted specifications only. Median annualized return by one construction choice at a time, holding the rest at their observed mix. Construction choices change results by more than the choice of factor sometimes does — a reason not to read any single backtest as the definitive number for a factor.

By information lag

Factor	3-month lag	6-month lag	Fama–French lag

By portfolio count

Factor	3 portfolios	5 portfolios	10 portfolios

Definitions, sources & caveats

Return on assets (ROA): Long-short portfolios sorted on operating income scaled by total assets.
Return on equity (ROE): Long-short portfolios sorted on net income scaled by book equity; the most leverage-sensitive of the five.
Gross profitability (GPA): Sales minus cost of goods sold, scaled by total assets. The most stable definition across eras in this pilot.
Operating profitability (OPE): Revenue minus cost of goods sold and operating expenses, scaled by book equity.
Cash-flow-to-market (CFM): Operating cash flow scaled by market equity. Strongest full-history result, weakest post-2010 result — its denominator brings market valuation into the signal.
Series / specification: One tested combination of factor, information lag, sorting method, and portfolio count (86,400 total). Not an individual company.
Not a Meridian backtest: These are third-party long-short research portfolios, not issuer-level signals and not a reconstruction of any Meridian index. None of Meridian’s quality score, ROIC formula, or weighting rules changed because of this pilot.

Sources: Tidy Finance Factor Library (CRSP/Compustat-derived U.S. long-short portfolio returns, 1960–2024); MarketBrain factor-library pilot (jeremyj2e/factor-library, 86,400 series). See the companion research article for the narrative read of the cash-flow-vs-gross-profitability regime split.