Analyst Calls Across Block, Equifax, Lockheed, Spotify—Signal or Noise?
Monday’s analyst calls are a familiar ritual: a dozen price-target adjustments delivered with the statistical rigor of a coin flip wearing a suit. Block, Boise Cascade, Equifax, FirstEnergy, Lockheed Martin, Lumentum, NetApp, Roper, Spotify — all re-rated by sell-side desks whose aggregate accuracy, per every study of analyst revisions I’ve ever backtested, hovers suspiciously close to the unconditional drift.
That’s not to say the calls are worthless. They’re just mispriced as information. Analyst upgrades contain signal, but mostly as a function of the *revision pattern*, not the target itself. A firm’s third consecutive upward revision to estimates — as opposed to a one-off target bump — has historically been the more durable predictor, consistent with the post-earnings-announcement drift literature going back to Ball and Brown. Price targets, by contrast, are essentially narrative with decimals. If you want to trade them, code the revision frequency and direction into a factor, not the headline number.
A few of the names deserve quantitative scrutiny rather than reaction. Lockheed Martin sits in the low-beta defense factor bucket, which means an analyst upgrade tends to move it less than the headline suggests — defense multiples compress and expand on budget cycles and geopolitical beta, not on a research note. Lumentum, meanwhile, is a high-volatility optical name where analyst sentiment shifts coincide disproportionately with AI-infrastructure flows; trading the call alone means paying for momentum that factor models already priced days earlier. That’s a classic case of the signal arriving after the alpha.
For anyone building this systematically: pull the calls into a pipeline (yfinance plus a scrape of the headline metadata, or a paid feed if you’re serious), tag each event by firm, direction, and whether it’s target-only or estimate-revision, and run an event-study framework — `eventstudy` on GitHub does the heavy lifting for abnormal-return windows. Measure CARs over 1-day, 5-day, and 20-day horizons, control for sector beta, and see for yourself whether the calls clear transaction costs. Most of the time, they don’t. The edge, when it exists, is in the second-order names — suppliers and competitors that haven’t been re-rated yet.
The honest summary of Monday’s research calls is that they’re a coincident indicator of what the market already thinks, dressed as a leading one. Trade the revisions, ignore the targets, and let the backtest arbitrate.