Monday’s Analyst Upgrades: Signal or Noise? A Quant Screens the Research Calls

Monday’s analyst research calls: Block, Boise Cascade, Equifax, FirstEnergy, Lockheed Martin, Lumentum, NetApp, Roper, Spotify, and more.

Let me be direct about how a systematic trader should treat sell-side price target changes: as a factor, not as advice.

The empirical literature on analyst revisions is actually one of the more robust findings in behavioral finance. Post-earnings-announcement drift has a well-documented cousin in recommendation drift — upgrades and downward target revisions tend to be followed by continued drift in the same direction over subsequent weeks. If you’re building signals in Python, `pandas` plus a clean academic dataset gets you further than any single analyst’s conviction on Lockheed Martin.

The problem is that Monday morning research call roundups present these calls as events rather than data points. One upgrade of NetApp tells you almost nothing. A time series of revision breadth — the ratio of upgrades to downgrades across a sector, aggregated over a rolling window — is where signal actually lives. Revision breadth has been shown to correlate with forward returns in ways that survive transaction-cost scrutiny, at least until crowding erodes the edge, as it inevitably does.

A few things worth noting when consuming today’s list:

– **Distinguish target changes from rating changes.** Price target adjustments often just chase the stock’s own trailing move — a mechanical correlation that inflates the apparent accuracy of analysts who are, in effect, lagging the tape by a quarter. Any backtest treating targets as exogenous signal is contaminated by this endogeneity.
– **Watch the denominator.** “Top analyst calls” is a survivorship-filtered sample. The calls that aged badly don’t get syndicated into roundups.
– **Size matters less than breadth.** A single bank moving Spotify’s target is noise; a coordinated sector-wide revision cycle in semis or defense is something you can actually model.

For the names on today’s list — Block, Equifax, FirstEnergy, Roper — the actionable move isn’t trading the headline. It’s logging the revision, timestamping it, and adding it to a dataset. Stack enough Mondays and you can test whether sell-side momentum in these coverage universes adds incremental alpha beyond price momentum, or whether it’s just expensive redundancy. My prior, based on how this usually backtests: mostly redundancy, with occasional alpha in small- and mid-cap names where coverage is thin and information diffusion is slower.

The market doesn’t need another roundup of targets. It needs fewer people trading them and more people measuring them.