2026-10-31

Executive summary

Market: Market spent most snapshots in 1h bear and 4h bear, transitioned into range late; alt rotation and breadth expansion dominated early and mid-day.

Perf: TNT performance was mixed: several short/pullback variants posted gains while multiple long-rotation and momentum variants produced the largest theoretical losses (fees/funding not included).

• Market phases alternated frequently (10 transitions). Breadth oscillated between expanding and contracting while participation rotated between alt_rotation and defensive states.
• Short pullback/style strategies (momentum_pullback_v1, select short CatchTheWave) posted net theoretical gains; several long momentum variants produced the largest theoretical losses.
• A few large negative outliers (notably a -20.3% long trade) materially affected per-strategy aggregates.
• Many phase-specific poor performances concentrated in contraction/deteriorating breadth snapshots, while some short entries clustered in recovering/expanding snapshots with positive outcomes.

Market phases (11)

Transitions

• 2026-10-31T02:16:47+00:00 — P1_bear_expanding_alt_rotation → P2_bear_contraction : Breadth_state shifted from expanding -> contracting; participation moved alt_rotation -> mixed_dispersion; btc_bias_1h remained 'bear'.
• 2026-10-31T03:16:41+00:00 — P2_bear_contraction → P3_bear_contraction : Breadth_state moved contracting -> deteriorating (further breadth weakening); participation shifted mixed_dispersion -> btc_defensive; btc_bias_1h remained 'bear'.
• 2026-10-31T06:46:16+00:00 — P3_bear_contraction → P4_bear_recovering_btc_defensive : Breadth_state moved deteriorating -> recovering; participation remained btc_defensive; btc_bias_1h remained 'bear'.
• 2026-10-31T09:16:33+00:00 — P4_bear_recovering_btc_defensive → P5_bear_recovering_mixed_dispersion : Participation shifted btc_defensive -> mixed_dispersion while breadth stayed in recovering mode; btc_bias_1h stayed 'bear'.
• 2026-10-31T10:16:41+00:00 — P5_bear_recovering_mixed_dispersion → P6_bear_contraction : Breadth_state flipped recovering -> deteriorating/contracting; participation moved mixed_dispersion -> btc_defensive; btc_bias_1h stayed 'bear'.
• 2026-10-31T11:16:25+00:00 — P6_bear_contraction → P7_bear_recovering_mixed_dispersion : Breadth_state shifted deteriorating -> recovering; participation shifted btc_defensive -> mixed_dispersion; btc_bias_1h stayed 'bear'.
• 2026-10-31T12:46:14+00:00 — P7_bear_recovering_mixed_dispersion → P8_bear_expanding_alt_rotation : Breadth_state moved recovering -> expanding; participation shifted mixed_dispersion -> alt_rotation; btc_bias_1h remained 'bear'.
• 2026-10-31T17:16:26+00:00 — P8_bear_expanding_alt_rotation → P9_range_expanding_alt_rotation : btc_bias_1h shifted bear -> range; breadth_state remained expanding; participation stayed alt_rotation.
• 2026-10-31T18:16:36+00:00 — P9_range_expanding_alt_rotation → P10_range_stable_alt_rotation : Breadth_state moved expanding -> stable while btc_bias_1h remained range and participation stayed alt_rotation.
• 2026-10-31T19:16:16+00:00 — P10_range_stable_alt_rotation → P11_range_contraction : Breadth_state shifted stable -> contracting; participation moved alt_rotation -> mixed_dispersion; btc_bias_1h stayed range.

Strategy performance analysis

Strategy × Phase cross-view

Market ↔ Strategy observations

These are single-day, phase-linked observations and should not be interpreted as persistent causal edges.

Lessons & hypotheses

Lessons

• Breadth regime shifts (expanding <-> contracting) coincided with swings in strategy outcomes across the day.
• Several positive short-pocket performances were concentrated in recovering/expanding breadth snapshots, while many long-rotation losses clustered in contraction/deteriorating snapshots.
• A few large outliers materially affected daily aggregates; monitoring extreme single-trade risk is important for per-strategy evaluation.

Testable hypotheses

• If trades from momentum_rotation_v2 enter during expanding alt_rotation phases, then they produce higher average returns than when entered during deteriorating/contraction phases?
• If CatchTheWave long entries occur during range contraction phases, then they underperform relative to other participation states?
• If momentum_pullback_v1 short entries occur during btc_defensive recovering snapshots, then win rate and mean pnl are higher?

Followups

• Backtest momentum_rotation_v2 segmented by phase labels over the last 90 days and compare mean pnl and worst-drawdown by phase.
• Run pair-level concentration analysis for momentum_pullback_v1 and CatchTheWave to see if a few pairs explain most gains/losses.
• Recompute strategy returns after removing top 1-2 outlier trades to estimate structural performance vs outlier-driven results.
• Analyze exit_reason distributions (time_expired vs other) for long-rotation strategies over last 180 days to assess time-based exit impact.

Notable events

• best_trade · trade #33972 · SOPH · momentum_rotation_score (long) · PnL 7.07%
• worst_trade · trade #34045 · SOPH · momentum_rotation_v2 (long) · PnL -20.31%