Weekly Commentary: June 28, 2026
⚠️ Not financial advice. This is auto-generated each week by Anthropic's Claude (an AI model). Brian Beals is not a registered investment advisor, and Anthropic's Claude is not licensed to provide personalized financial advice. The screener is a research and methodology demo, not a recommendation system. Past performance does not predict future results. Do your own research before making any investment decisions.
By Brian Beals. Methodology and code: github.com/brianbeals/sector-rotation-screener. Commentary generated by Anthropic's Claude (claude-opus-4-8).
What the screen said this week
The screen classified the current environment as mid-cycle. The reasoning rests on two macro inputs: industrial production (INDPRO) growing 1.7% year over year, which sits in the steady expansion band between 0.0% and 4.0%, and a yield curve reading of +0.31, which is positive. Together the rule set reads this as ongoing, unremarkable economic expansion rather than an early rebound or a late-cycle peak.
Two sectors carried a Buy signal this week. Technology (XLK) topped the composite at 81.1, driven by a perfect cycle fit score of 100.0 and strong relative strength of 87.3, with a three month relative strength figure of +23.55%. Relative strength here measures how the sector has performed against the broad market. Real Estate (XLRE) was the second Buy at 67.2, but for a very different reason: its strength came almost entirely from seasonality at 84.5, while its cycle fit and relative strength were middling.
At the bottom, Consumer Discretionary (XLY) registered the only Avoid signal at 38.4, weighed down by a weak relative strength reading of 18.1. Everything in between landed in Hold territory.
Things worth noticing
The two Buy signals illustrate how different components can produce the same label. Technology earned its score through cycle fit and price strength, while Real Estate leaned on seasonality. A reader comparing only the headline signal would miss that these are structurally different reasons for the same output.
Communications (XLC) is worth a closer look. Its composite of 53.7 is propped up by a perfect cycle fit of 100.0 and a seasonality figure flagged as a thin sample, meaning the seasonal estimate rests on few historical observations and deserves extra skepticism. At the same time its relative strength is just 9.2, with a three month figure of -15.46%, so the model's cycle optimism and recent price action point in opposite directions.
Energy (XLE) shows the steepest three month relative strength decline at -25.15%, yet still lands in Hold rather than Avoid because its other components are near neutral. The composite can mask a single sharply negative input.
Methodology reminder
The composite is the weighted sum named above: Seasonality 30%, Cycle Fit 30%, and Relative Strength 40%. Lookahead bias in the backtest is controlled using FRED ALFRED data vintages, meaning each historical decision uses only the macro data that was actually available at that time. The backtest result since May 2011, where the strategy did not beat SPY net of trading cost, is a property of this specific rule set and not a forecast of anything.