Equity Research Performance

Weekly Performance Article Archive

A new weekly performance article is published each Friday; this dashboard stays as the live daily-close scorecard.

Monthly Performance Article Archive

A new monthly performance article reviews each completed calendar month by rating tier, with a blended overall figure; showing the trailing 6 months.

Cautious on U.S. Companies, Not Calling a Crash

The mix is deliberately defensive: most covered companies are Hold, with more Sell/Strong Sell calls than Buy/Strong Buy calls. That is a company-level valuation and fundamentals read, not a macro recession forecast. Hold calls are monitored separately because they are not directional portfolio positions.

Performance by Rating

Call Hit Rate = Correct for Buy/Sell-style ratings, or stayed within the +/-10% band for Hold.

What We Got Right And Wrong

Correctness Summary

Rating vs Actual Stock Outcome

The first table is the clean score: right versus wrong by LF0 rating. The matrix below shows what actually happened to the stock after entry. Strong Buy outcome means the stock rose 20% or more; Buy outcome is +5% to +20%; Hold outcome is -5% to +5%; Sell outcome is -20% to -5%; Strong Sell outcome is -20% or worse. Green cells are aligned with the original LF0 rating; red cells moved against it.

Cumulative Performance Through Latest Close

Monthly Cohort Cumulative Performance

This chart is grouped by publication month. The current-month point is month-to-date through the latest available close, not a stale month-end value.

Monthly Rating Portfolios

Blended (long/short pooled)

By Rating

Statistical Analysis

This section is new and the sample is very thin: our tracked history spans only a few months, and Ken French’s public factor library itself currently reports data with roughly a two-month lag (this uses the daily-frequency factor file, which gives far more data points per week of overlap than a monthly or weekly file would — “daily” here means one return per trading day, not that the file itself is refreshed daily; its own reporting lag is unaffected by that choice). The usable overlap between the two is often still just a few weeks of trading days per tier today — far short of the years of daily history real factor studies use, and a daily alpha estimated over only a few weeks gets compounded into an enormous (and largely meaningless) annualized figure. CAPM is shown down to a low observation floor “for reference”; the 3-factor model is hidden entirely below its own higher floor. Treat every number below as indicative, not conclusive, and never as a basis for calling any tier “the best” or “the worst.” Sell and Strong Sell rows (marked ^† below) are regressed on a direction-flipped return series, so their Alpha/Beta/SMB/HML describe that short-signal, not the covered stocks directly — see the footnote under the table.

CAPM & Fama-French 3-Factor, by Rating

Factor Obs. is the count of LF0 rating-tier return days that overlap the Ken French factor file, not the age of the full article archive. CAPM appears once a rating tier has 10+ overlapping factor days; Fama-French 3-Factor appears at 20+. The Ken French factor file can lag the current market date, so this table uses the latest overlapping factor/price window available through 2026-08-31. “Alpha p” is the two-sided p-value on the alpha (intercept) estimate; * = p<0.05, ** = p<0.01. With as few as 7-37 daily observations and likely-autocorrelated daily returns, treat even a starred alpha as weak evidence, not proof the true alpha is nonzero. ^† This tier’s return series is direction-flipped (a Sell/Strong Sell call profits when the stock falls, so its tracked “return” is the inverse of the stock’s own return — see the Signal Return note above). Alpha, Beta, SMB, and HML for this row are therefore the regression of that inverted series, which is the mirror image (same magnitude, opposite sign) of the underlying stocks’ own factor loadings — R² is unaffected by the sign flip. Do not read this row’s beta/SMB/HML as describing the covered stocks directly.

Style Box (Size × Value/Growth tilt)

Stylized: places each tier using its own Fama-French SMB/HML regression coefficients, not Morningstar’s actual holdings-based percentile methodology.

Sector Exposure by Rating

Best Active Calls So Far

All Equity Research Calls

Click a column header to sort. 175 calls tracked in total, one row per ticker per rating call (re-covered tickers show every call).

Methodology