Forward EPS revisions: the breadth tell
When more sectors are getting cut than raised in the last 4 weeks, forward returns flip negative. The current reading is one we've only seen 6 times since 2010.
The signal
Forward EPS revision breadth — the share of S&P 500 sectors with rising 12-month forward EPS, minus the share with falling — is one of the cleanest mid-cycle tells we have. The current reading is -3 out of 11 (5 up, 8 down). That's not a crisis number, but it's a number that historically precedes a stall.
The historical record
Since 2010, we've had six readings at or below -3 while the index was within 5% of an all-time high. Forward 6-month returns:
- 2011: -8% (debt ceiling)
- 2015: -5% (China devaluation, oil)
- 2018: -10% (Fed mistake + tariffs)
- 2020: -1% (COVID, then pulled forward)
- 2022: -16% (everything)
- 2024: +2% (the dovish pivot exception)
Median: -6.5%. The 2024 outlier matters — the Fed pivot bailed it out — but it's the exception, not the pattern.
What's getting cut
The breadth is concentrated. Energy (oil rolling), Materials (China), and Health Care (Medicare pricing) are the three biggest contributors to the downside. Tech and Communication Services are still net-up — but the magnitude of upgrades is smaller than three months ago, which is the second derivative we watch.
The full sector revisions table is at /sp-earnings. The Yardeni-derived series feeding it lives in yardeni-forward-metrics.json.
The trade
Two ways to play this:
- Quality factor over equal-weight. When breadth narrows, the safest names outperform the average — even if the average outperforms the index.
- Long vol on a 3-month horizon. Five of the six historical analogs saw a >25 vol print within 90 days.
Bottom line: the breadth tell isn't "sell" — it's "tighten." The 2024 escape hatch was a Fed pivot. Without one, the historical median wins.