Cycle Top and Bottom Identification Signals

Tops and bottoms are processes, not events. They leave a sequence of measurable signals. No single signal confirms a turn; a cluster of three or more shifts the probability meaningfully.

Top Signals

• Euphoric sentiment: AAII bulls >50% with bears <25%; CNN Fear & Greed >80 for weeks; IPO calendar floods.
• Margin debt peak: FINRA margin debt year-over-year change turning negative (not the absolute level) preceded the 2000, 2007, and 2021 tops.
• 200-day MA loss: a weekly close below the 200-day MA after an extended uptrend; a failed test back from below is a secondary distribution signal.
• Distribution days: William O'Neil's session with ≥0.2% decline on higher volume than the prior session. Four or five within 2–3 weeks on a major index precede corrections.
• Narrow leadership: the S&P 500 makes a new high but the equal-weight version does not — breadth is deteriorating.
• Credit divergence: high-yield OAS widening while equities make new highs. Credit peaked months before equities in 2000 and 2007.

Bottom Signals

• Capitulation volume: a down session with 2–3x the 50-day average volume, often a gap-down that recovers into the close (March 2009, March 2020).
• Sentiment extremes: AAII bears >50% with bulls <25%; VIX above 35–40; put/call ratio above 1.2 on a 10-day average.
• Failed breakdown (spring): price breaks support on high volume, then snaps back within 1–3 sessions — the last shakeout before accumulation.
• Base formation: multi-week range with declining downside volume; first close above on volume >150% of average is the breakout.
• Breadth thrust: 10-day EMA of NYSE advancing issues / total issues rising from below 40% to above 61.5% within 10 sessions (1974, 1982, 2009, 2020).
• Credit tightening: HY OAS stops widening and tightens 50–100 bps over 2–3 weeks; credit bottomed ahead of equities in 2009.

Signal Clustering

Three or more matching signals raise the probability enough to act. Fewer than three is noise.

Action Points

1. Build a top/bottom scorecard and update weekly.
2. Require three matching signals before reducing risk at a top or adding at a bottom.
3. Confirm with a weekly trend break (close beyond the 200-day MA or prior swing) before committing.
4. Scale exposure in thirds as signals confirm, not all-in on the first.

Tops and bottoms cannot be called with certainty in real time. Probabilistic action on a cluster of signals is the edge.