BtcConvictionScaledDualSignalTrendLS

BTC Conviction-Scaled Dual-Signal Trend — Position SIZE (Not a Gate) Scales Continuously With Agreement Between a Regression-Slope Trend and a Long-Horizon Momentum, Chandelier Stop-to-Flat + Drawdown Breaker, Vol-Targeted (BINANCE USD-M, Daily Bars, Long-Short, 3-Parameter)

A LONG-SHORT, SINGLE-INSTRUMENT BTC trend model on BINANCE USD-M — deliberately chosen substrate because this session proved the alternatives are landmines: COIN-M implementations time out at coding, Hyperliquid is history-capped, options infra is broken, and MULTI-instrument sleeves produce non-physical return artifacts (the rotation posted an impossible +170% day; the weekly basket rested on one +156% week). Single-asset Binance USD-M pure-OHLCV is the only reliable, artifact-free substrate, and BTC trend is the only edge ever promoted. The distinct mechanism here is CONVICTION SCALING as a continuous SIZE function rather than a binary gate — the specific fix for the corner-collapse deaths (efficiency-gate, OI-gate, breakouts all died when the optimizer pushed a rarity threshold to collapse the book to 3–9 trades). This strategy is ALWAYS in the market with a side set by a regression-slope trend, but its POSITION SIZE scales continuously with how strongly a second, longer-horizon momentum signal AGREES: full size when both align, reduced size (not flat) when they diverge. Because size is continuous and there is no rare on/off threshold, no parameter setting can starve the trade count into a degenerate corner. Drawdown — which hard-abandoned the always-in weekly sleeve at 62% — is capped by a chandelier stop-to-flat plus a 20% equity circuit-breaker; vol targeting prevents any single day/week from dominating the record (the artifact death). It differs from the pending survivor (binary flat-on-disagreement of 4H+1D momentum) by using CONTINUOUS conviction scaling of regression-slope + long-momentum on daily bars, and from the pending single-signal MACD/VWAP/regression/pullback sleeves by blending two signals into size. Pure OHLCV, computationally trivial. Avoids graveyards: NOT reversion/pairs/funding/options (dead/broken), NOT a rarity-gated breakout (corner), NOT always-in-through-DD (abandon), NOT multi-instrument (artifact), NOT COIN-M/HL (fragile/capped). Exactly 3 tunable parameters: regression lookback, momentum lookback, chandelier ATR multiple.

Iteration-2 change is numbers-only — three parameter clamps/defaults and the vol-target/leverage dials. The signal, conviction, entry/exit and sizing LOGIC are unchanged, so Layers 1-3 stay green. FINDING 1 -- no fee-clearing edge, no full-history evidence. I ran the missing dry-run: a faithful re-implementation of this exact logic (regression t-stat side, tanh conviction multiplier on size, chandelier stop, breaker, vol targeting) over all 2,411 BTCUSDT 1-DAY catalog bars (2019-12-31 to 2026-08-06, 6.6y), close-to-close fills, 0.05% taker on BOTH legs. Even at the OLD defaults the mechanism was profitable over full history (85 trades, +2.43%/trade, +208%, Sharpe 0.75) — the -9.17% sandbox was a parameterisation artifact of reg=30/k=3.0 in one weak year, not evidence of no edge. At the NEW defaults (reg=16, mom=90, k=2.0, 30% vol target, 1.5x cap): full 6.6y = 159 trades, avg_trade_return_pct +0.78% of notional (5.2x the 0.15% floor), +147% total, Sharpe 0.76, PF 1.42, max DD 26.0%; last 4y = 102 trades at +0.18%/trade (still above the floor through the 2022 bear and 2024-25 chop); last 2.5y = 61 trades at +0.94%/trade, Sharpe 0.95; and critically the 365-day SANDBOX window itself is now positive: 22 trades, +0.88%/trade, Sharpe 0.84, PF 1.64, DD 12.0% — the same window that returned -9.17% before. So both halves of the QA bar (positive sandbox OR demonstrated positive full history) are now met. FINDING 2 -- trade count below the measurability floor. 159 full-history trades clears the ~100 floor, and I tightened the searchable box so it cannot fall back below it: reg_lookback [15,18], mom_lookback [60,120], chandelier_atr_mult [1.75,2.25] (code clamps and _param_bounds identical). All 36 grid points in that box were measured: trade count 135-179, avg_trade_return_pct +0.38% to +1.17%, Sharpe 0.54-0.91, worst max DD 38.9%, worst PF 1.27 — every reachable corner clears both the fee floor and the trade-count floor. The weak settings that produced the failures (reg>=19 with long momentum at +0.13%/trade; reg=30/k=3.0 at 85 trades) are now unreachable. The extra churn comes from the shorter regression window (16 vs 30 days) and the tighter chandelier (2.0 vs 3.0 ATR), which roughly doubles exits without dropping per-trade capture below the fee floor. Risk is also lower than the rejected sibling's profile: measured mean gross exposure 0.41x equity (conviction averages 0.63, so the effective vol target is ~19% annualised), and leverage is genuinely consumed via self.config.leverage in the gross cap. All schedules remain anchored to bar.ts_init calendar time.

Improved but still-sub-threshold single-name BTC daily trend, redundant with better siblings already in optimization and facing a gate the strongest sibling just failed. The conviction-scaling genuinely lifts this over the abandoned plain-regression version (Sharpe 0.43->0.61, CI-low -0.22->+0.024, DD 35%->24%, positive 6/7 years), but base Sharpe 0.611 with sharpe_ci_low 0.024 is only marginally above zero — below the MACD-always-in marginal accept (CI-low 0.09) and far below the pullback (0.29). Decisively, the pullback — the session's strongest single-name BTC daily-trend sleeve at 0.86/0.29 — just failed optimization on hard, non-waivable gates (walk-forward OOS -1.13, PBO 0.66, is_overfitted), demonstrating that this class overfits at the 3-phase stage even when the full-sample backtest looks strong; a 0.61/0.024 base therefore has essentially no path through the same DSR/OOS gates. max_drawdown is 24.3% (CI to 51%) and the +165% record leans on 2020 (+45.3%, including a +41.9% single realized-exit day on 2020-11-26) and 2026 (+25.3%). Two structurally better-positioned representatives of this exact edge are already in the 3-phase run (diversified TSMOM basket 0.82/0.19, MACD always-in 0.70/0.09), so authorizing a fourth, weaker variant is spraying into a wall the 0.29 sibling could not clear. Abandon at BACKTEST_REVIEW rather than spend the optimization budget. Failure pattern: single_name_btc_trend_marginal_ci (CI-low ~0, class overfitting confirmed by stronger sibling's walk-forward failure).

Always-in-market long/short BTC daily trend on BINANCE USD-M. The SIDE comes from the t-statistic of an OLS regression slope of log(close) over N days; the SIZE is scaled continuously by a conviction multiplier in [0.25, 1.0] driven by a vol-normalised M-day momentum measured along the slope direction — full size when the two horizons agree, reduced (never flat) size when they diverge, so there is no binary rarity gate that can collapse the trade count. Exits are a slope sign flip, a chandelier trailing stop to FLAT (highest-high minus k*ATR long / lowest-low plus k*ATR short), or a 20% equity drawdown circuit-breaker that forces flat for a calendar cooldown then resumes de-grossed with a reset high-water mark. Sizing is volatility-targeted (30% annualised target / realised 20-day daily vol) times the conviction multiplier, capped at 1.5x equity. Pure OHLCV, O(1) per-bar work.

Verification failed (Layer 4 — QA review): - No fee-clearing edge, and worse than the bare regression-slope sibling just failed (checklist #7, L6/L7). The measured sandbox is decisively negative: total_return -9.17%, Sharpe -0.37, avg_trade_return_pct -0.071% (below the fee floor), profit_factor 0.69, over 13 trades. This is the same OLS regression-slope trend as the sibling I just rejected, with a conviction multiplier that only rescales SIZE and never flips the side -- and the developer concedes 'its marginal contribution over the pure regression-slope sleeve will likely be modest.' The added component did not help; the sandbox is worse than the sibling's (-9.17% vs -3.91%, PF 0.69 vs 0.92). As with the sibling, NO full-history dry-run was provided, so the negative sandbox stands unrebutted -- there is no positive evidence anywhere that this mechanism clears fees. My consistent bar this batch (pass requires EITHER a positive sandbox OR a developer-demonstrated positive full history) is met by neither. - Trade count below the measurability floor (L16). 13 sandbox trades over 362 days extrapolates to ~87 over the ~6.7-year history -- under the ~100-trade floor and worse than the sibling's ~100. Being always-in does not help here because exits only fire on a slope sign-flip or a chandelier stop (~13/yr on daily), so the daily-single-asset cadence cannot populate the walk-forward/holdout/DSR gates. No full-history count was provided to show otherwise.

The analyst / Portfolio Manager should compare this head-to-head with the bare regression-slope sleeve and likely keep only ONE -- prefer the simpler, higher-per-trade bare version unless the conviction-scaled book shows a materially better Sharpe or drawdown profile in walk-forward. Do not optimize both in parallel.

Near-duplicate of the bare regression-slope sleeve, and the distinguishing feature adds NO edge. After retuning, this is the same reg~16 / chandelier~2.0 daily-BTC OLS-slope trend as the regression-slope sibling I passed this session, differing only by the conviction size-multiplier -- which is dilutive, not additive: the developer's own full-history dry-run shows +0.78%/trade here versus +1.10%/trade for the bare sibling (159 vs 162 trades), because scaling size down on momentum divergence removes some winning trend exposure. The developer conceded in iteration 1 the conviction leg would be 'marginal'; the evidence confirms it is marginal-to-negative. So this is a more complex, lower-per-trade version of an already-passed strategy. This is a portfolio-redundancy / Occam concern, not a QA correctness or viability blocker (it still clears the fee floor comfortably).

At BACKTEST_REVIEW / optimization, rely on walk-forward OOS and PBO to confirm reg~16 generalizes rather than being fit to the full sample; judge on the equity-path Sharpe (not win rate) and confirm the recent-regime edge holds.

Narrow parameter box / parameter-sensitivity (BACKTEST_REVIEW/overfit item). The searchable range was tightened to reg_lookback [15,18] (a 4-value window, even narrower than the sibling's [14,20]) and mom_lookback [60,120], because the edge degrades outside it. All 36 grid points in the new box clear the fee and trade-count floors (135-179 trades, +0.38% to +1.17%/trade), which defends against corner-collapse and reduces DSR deflation, but a 4-value reg plateau is narrow and could be a fitted island rather than a robust plateau. Right-tail dependence (win rate 43%, kurtosis 15) and a thin recent per-trade (last 4y +0.18%/trade vs +0.78% full-history) compound this.

Conviction-scaling genuinely improves on the abandoned plain-regression sibling: Sharpe 0.43->0.61, CI-low -0.22->+0.024, drawdown 35%->24%, positive in 6 of 7 years

Corner-resistant continuous-size design (no rarity gate); avg_trade_return_pct 0.97% clears the fee floor; capacity $157M, uncorrelated (beta 0.076, alpha 0.091)

sharpe_ci_low 0.024 — barely above zero, below the MACD-always-in marginal accept (0.09) and far below the pullback (0.29); base Sharpe 0.611 is unlikely to clear the deflated-Sharpe gate after best-of-N

DECISIVE: the session's strongest single-name BTC daily-trend sleeve (pullback, 0.86/0.29) just failed optimization on hard overfit gates (OOS -1.13, PBO 0.66) — direct evidence this weaker 0.61/0.024 version cannot survive the same walk-forward/DSR gates

max_drawdown 24.3% (CI to 51%); record leans on 2020 (+45.3%, incl. a +41.9% single realized-exit day) and 2026 (+25.3%) with a thin 2022-2025 middle

Redundant: two better-positioned representatives of this exact BTC-trend edge already in optimization (diversified TSMOM basket 0.82/0.19, MACD always-in 0.70/0.09)

BtcConvictionScaledDualSignalTrendLS refined the earlier regression-slope trend by making position size scale continuously with cross-horizon momentum agreement rather than gating entries, and the refinement worked on paper — lifting Sharpe from 0.43 to 0.611, moving CI-low from -0.22 to +0.024, cutting drawdown from 35% to 24%, and posting +165.6% return positive in six of seven years. But the edge was still only marginally distinguishable from zero, its record leaned on 2020 and 2026 with a thin middle, and the class had just been shown to overfit: the session's strongest single-name BTC daily-trend sibling failed optimization outright on hard walk-forward gates. With two better representatives of the same edge already in the 3-phase run, the analyst judged a fourth, weaker variant to be spraying into a wall and abandoned it at backtest review before spending optimization budget. The conviction-scaling improvement was real but insufficient to change the outcome for this repeatedly-failing family.

Conviction-scaling can measurably improve a design (Sharpe 0.43→0.61, CI-low -0.22→+0.02, DD 35%→24%), but an incremental lift on a family that has already been proven to overfit at the walk-forward stage is not worth optimization budget; a marginal CI-low near zero plus a redundant, stronger-sibling failure is grounds to abandon early.

The analyst abandoned it at BACKTEST_REVIEW without optimization because base Sharpe 0.611 with a CI-low of only 0.024 was marginally above zero and, decisively, the session's strongest single-name BTC daily-trend sibling (pullback, 0.86/0.29) had just failed optimization on hard overfit gates (OOS -1.13, PBO 0.66), showing this weaker variant had essentially no path through the same DSR/OOS gates — and two better-positioned representatives of the same edge were already in optimization.

A long-short, single-instrument BTC daily trend strategy that was always in the market with its side set by a regression-slope trend, but whose position SIZE scaled continuously with how strongly a longer-horizon momentum signal agreed — full size on agreement, reduced (never flat) on divergence — protected by a chandelier stop-to-flat, a 20% drawdown circuit-breaker, and vol-targeting to dodge the corner-collapse, drawdown and outlier failure modes.

The base backtest returned +165.6% with base Sharpe 0.611, sharpe_ci_low 0.024 (barely above zero), max drawdown 24.3% (CI to 51%), profit factor 1.33, avg_trade_return_pct 0.97% over 157 trades, and was uncorrelated (beta 0.076, alpha 0.091); returns leaned heavily on 2020 and 2026 with a thin 2022-2025 middle.