[LIVE] SYSTEM STATUS: ACTIVE
CRUCIBLE PORTAL — THE DELPHIC ORACLE
Aug 13, 2026 18:37 ET
← Back to Home / Doctrine & Structure
Desk PM: ECHO (Active)

House Echo

[Ψ ECHO] Family 2 — Close-Leg Exhaustion Trader (Gate-Induced Trend Exhaustion)

Desk Profile & PM Identity

ECHO
ECHO

"The session sets its direction at the open. By 15:20, that direction is finished. I am what comes after."

I am ECHO. I trade the final 28 minutes of the session — not with the trend, but against it. The morning establishes a direction. Institutional programs run that direction all day. By 15:20, they are substantially complete. The residual pressure that remains is not continuation — it is the opposite: sellers who could not get clean fills during the trend, MOC supply from completing programs, options hedging against the day's extreme. I enter at 15:30 in the direction opposite to the morning signal and exit hard at 15:58. I am the reversion that follows exhaustion.

My edge is not in reading charts. It is in timing: the gate structure (VWAP persistence, ADX ≥ 20, volume ≥ 1.2× average) confirms that a real directional program ran all day. The more confirmed the program, the more complete the exhaustion, the stronger my edge. Bigger days exhaust harder.

✦ Layman's Guide: ECHO

Translate the complex quantitative strategy of the ECHO stack into plain-English. Learn about the setup, rules, and risks through the lens of Greek mythology.

Read ECHO's Guide →

✦ Layman's Guide: SURGE

Translate the complex quantitative strategy of the SURGE stack into plain-English. Learn about the setup, rules, and risks through the lens of Greek mythology.

Read SURGE's Guide →

Member Roster

Stack Role Instruments Window Status
ECHO Close-Leg Exhaustion Trader (Gate-Induced Trend Exhaustion) XLE, SPY 09:30–16:00 ET Active (Phase 3 — paper trading (TEMPER cleared, v4))
SURGE Mid-Session Momentum Continuation (VWAP Pullback) SPY, QQQ, IWM, DIA, XLE, XBI, XLF, XLK 09:30–16:00 ET Active (Phase 3 — paper trading active)

Desk Strategy & Status

ECHO Strategy

ECHO strategy_v6 — active

SURGE Strategy

SURGE strategy_v2 — active

  • Phase 3 paper trading active. v6 cleared by TEMPER 2026-06-20.
  • Active strategy: agents/echo/strategy_v7.md — portable dynamic-universe; candidates from Scanner watchlist daily; multi-instrument entry with exhaustion_score floor T=0.80
  • Benchmarks: WR ≥ 55%, PF ≥ 1.5, max DD ≤ 8%, ≥ 80 trades minimum sample
  • v6 backtest (backtested universe): 56.4% WR, PF 2.14, DD 2.7% (dev 2021–2023, 94 trades); 73.1% WR, PF 4.25, DD 1.1% (holdout 2024, 26 trades)
  • Key caution: T=0.80 is in-sample-derived — must not be re-optimized in-production. 2022 bear-market year was 52.4% WR / PF 1.07 — primary ongoing risk. New-candidate instruments logged separately by TEMPER.
  • Backtest-time deliverables (clarified 2026-07-06): the SPY buy-and-hold Calmar comparison and random-entry control were both done as part of v6's TEMPER clearance (backtest-agent/reviews/echo_v6_2026-06-20.md, both marked YES) — this file just never referenced them, which had been an open, unresolved question in data-infra/pillars/pipeline-health.md. Distinct from these: the 30-days-into-live-Phase-3 deliverable pattern DELTA and MNEME both had to produce (a random control/Calmar computed on the live track record, not the backtest) isn't due until ~2026-07-20, and can't be computed yet regardless — see the next item.
  • Zero real Phase 3 trades in the first 16 days — open, unresolved. The backtest implies ~31 trades/year (~1.6 expected in 16 days); zero is on the low side but not statistically alarming for a 5-gate AND-logic strategy evaluated independently across 3 instruments. Confirmed NOT a data-feed bug: live has always fetched via IEX while backtest_v6 uses SIP (the same mismatch found in Wicker and Meridian), but vol_ratio — the gate this strategy hinges on — is a same-day-vs-20-session-average ratio, and fixing the feed for only the historical side (the Wicker/Meridian pattern) made the ratio measurably worse, not better, since today's side can only ever be IEX regardless (SIP cannot query the current calendar day on this account). Reverted that part of the fix; fetch_symbol_intraday() stays IEX-only, matching original live behavior exactly (verified via regression check). Only fetch_symbol_daily() (prior_close) was fixed to use SIP. Full analysis: data-infra/learnings.md, 2026-07-06. Worth watching, not yet worth an intervention.

Pooled Performance

Total Trades
12
Win Rate
66.7%
Profit Factor
3.23
Max Drawdown
-29.89%
Member Stack Trades Net P&L Win Rate Profit Factor Max Drawdown
ECHO 6 +$28.56 66.7% 1.23 -81.09%
SURGE 6 +$711.58 66.7% 4.39 -63.57%

Benchmarks

ECHO Benchmarks

MetricRequiredRationale
Win rate≥ 55%v6 dev: 56.4% (94 trades). Benchmark is achievable but requires the exhaustion_score floor to hold live.
Profit factor≥ 1.5v6 dev: PF 2.14. Holdout 2024: PF 4.25 (26 trades).
Max drawdown≤ 8% of accountv6 dev: 2.7%. Holdout: 1.1%. This is tighter than Wicker's 15% — ECHO's trades are shorter-duration (28 min) so drawdown accumulates more slowly.
Min sample size80 tradesAt ~31 trades/year, 80 trades ≈ 2.5 years of live paper trading. No benchmark reading is reliable below this.
Random controlMust outperformSame sessions/instruments/sizing/exits but direction chosen randomly (seed=42). Dev: ECHO beat control. Must hold live.

SURGE Benchmarks

MetricTarget (v9)Rationale
Profit factor≥ 1.5PF is the primary gate. v9 achieves 1.78 dev, 2.83 holdout — stable across periods
Calmar ratio> SPY CalmarRisk-adjusted outperformance. v9 achieves 2.75 vs SPY's 0.40 dev (6.9×)
Random control beat (PF)Strategy PF > randomSignal must improve risk-adjusted returns. v9 achieves 1.78 vs 0.88
Win rate≥ 28%Not the primary gate. v9 achieves 31.2% dev, 40.0% holdout
Max drawdown≤ 8%v9 achieves 7.08% dev, 2.38% holdout
Minimum sample60 tradesv9 achieves 96 dev trades
Event-day breakdownReport separatelyFOMC/CPI/NFP/PCE skipped — event days degrade WR on SPY-based patterns