AAC $9.95 ACHR $3.30 +0.76% ▲ AERO $16.37 -0.24% ▼ AIN $64.68 -1.10% ▼ AIR $105.11 -0.36% ▼ AM $20.17 -0.54% ▼ AMP $496.30 -1.14% ▼ AMTM $20.94 +0.10% ▲ APH $76.89 +1.04% ▲ ASB $29.66 -1.98% ▼ ATI $180.71 -0.09% ▼ ATRO $59.12 +0.42% ▲ AVAV $99.31 -3.11% ▼ AVNW $16.63 +1.03% ▲ AXON $317.24 +5.01% ▲ BAB $25.64 +0.08% ▲ BAH $82.40 -1.96% ▼ BBAI $1.79 +1.41% ▲ BDL $52.20 -2.06% ▼ BELFB $224.56 -0.48% ▼ BHE $62.10 +0.10% ▲ BKSY $19.49 -5.30% ▼ BWXT $117.76 -0.65% ▼ CACI $601.47 -1.29% ▼ CAE $27.44 -0.58% ▼ CDRE $30.73 -3.67% ▼ CICN $0.00 CMTL $1.07 -2.73% ▼ CNC $64.56 -5.06% ▼ CODA $7.90 -0.88% ▼ AAC $9.95 ACHR $3.30 +0.76% ▲ AERO $16.37 -0.24% ▼ AIN $64.68 -1.10% ▼ AIR $105.11 -0.36% ▼ AM $20.17 -0.54% ▼ AMP $496.30 -1.14% ▼ AMTM $20.94 +0.10% ▲ APH $76.89 +1.04% ▲ ASB $29.66 -1.98% ▼ ATI $180.71 -0.09% ▼ ATRO $59.12 +0.42% ▲ AVAV $99.31 -3.11% ▼ AVNW $16.63 +1.03% ▲ AXON $317.24 +5.01% ▲ BAB $25.64 +0.08% ▲ BAH $82.40 -1.96% ▼ BBAI $1.79 +1.41% ▲ BDL $52.20 -2.06% ▼ BELFB $224.56 -0.48% ▼ BHE $62.10 +0.10% ▲ BKSY $19.49 -5.30% ▼ BWXT $117.76 -0.65% ▼ CACI $601.47 -1.29% ▼ CAE $27.44 -0.58% ▼ CDRE $30.73 -3.67% ▼ CICN $0.00 CMTL $1.07 -2.73% ▼ CNC $64.56 -5.06% ▼ CODA $7.90 -0.88% ▼

Analyst Calls Pile Up—But Do Sell-Side Ratings Beat the Quant Signals?

October 26, 2026 · by stackcalder

Monday’s sell-side research calls landed with the usual fanfare: Block, Boise Cascade, Equifax, FirstEnergy, Lockheed Martin, Lumentum, NetApp, Roper, Spotify. A tidy basket of tickers, each attached to a rating change and a price target that someone’s compensation depends on defending.

Treat these calls the way a signal processor treats noisy input. Individual analyst upgrades have historically shown weak predictive alpha on their own — the post-publication drift on rating changes is thin, decays fast, and frequently gets arbitraged away in the first session. What matters is aggregation. If you’re serious about it, build the panel yourself: scrape rating distributions, compute the delta in mean target versus current price, normalize by analyst coverage count, and only then look for cross-sectional dispersion worth trading. `pandas` and a decent RSS parser get you 80% of the pipeline in an afternoon.

A few observations on the names themselves. Defense-adjacent names like Lockheed tend to trade on order backlog and budget cycles, not sentiment — a price target move there is mostly narrative dress-up. Equifax and NetApp are cyclical IT/data plays where estimate revisions carry more information than ratings; revisions momentum is a documented factor, ratings momentum much less so. Spotify gets upgraded whenever the market rediscovers subscription economics and downgraded whenever churn prints ugly. Neither event contains new information an attentive quant hadn’t already modeled.

The real tell is the price target mechanics. Analyst targets cluster with embarrassing correlation to the trailing 90-day price — regress target-to-price ratios against prior returns across a large cross-section and you’ll find the “conviction” is largely momentum with extra steps. A target that moved because the stock moved is noise. A target that moved against the stock is a signal, and it’s rare enough to matter.

If you insist on trading these calls, size for the actual edge: event-study your own sample first, estimate the conditional drift with proper confidence intervals, and skip anything where the standard error swamps the effect. Most of Monday’s list will fail that test. The market’s opening auction will have priced the upgrades before your market-data feed finishes its first tick — which is less a commentary on your latency than on the information content of the calls themselves.