SolDailyMultiWeekBreakoutLong

SOL Daily Multi-Week Trend Continuation Long-Only (BINANCE USD-M Futures, Daily, OHLCV-Only)

Long-only SOL daily strategy on BINANCE USD-M futures. Completes the BTC/ETH/SOL/LINK multi-week trend-continuation family. Enter when SOL is in a confirmed multi-week uptrend with a fresh 20-day high. Single-instrument, OHLCV-only, single-condition entry, factory-approved instrument and timeframe (SOLUSDT.BINANCE 1D).

Completes the BTC/ETH/SOL/LINK multi-week trend-continuation family with a Donchian-style 20-day-high breakout filtered by a 50-day (~7-week) trend SMA, distinct from the three-weekly-return variant. calculate_signal returns the continuous breakout depth (close vs the prior 20-day high) so the decision variable varies every bar and is positive exactly when a fresh high prints, rather than being gated to a constant; should_enter thresholds it (>0) and applies the multi-week-uptrend gate (close > 50-day SMA) by reading self._bars. The breakout reference excludes the current bar so today's candle does not contaminate its own comparison. The trend-break exit (close < 50-day SMA) lets winners run for weeks, matching the macro horizon. VENUE OVERRIDE: the hypothesis names BINANCE USD-M futures, but the strategy is long-only, uses no leverage, and its edge is pure OHLCV momentum (not funding) — so per the venue-selection rule it routes to BINANCE_SPOT (CASH), avoiding the funding cost a multi-week futures long would pay in an uptrend. leverage stays 1.0 and is intentionally unreferenced in sizing (spot caps at 1x), avoiding the leverage_set_but_unused rejection. Low trade frequency makes the ~0.20% spot round-trip fee negligible against multi-week trend captures.

Failed deflated Sharpe: DSR=0.1458, expected-max=7.84 over 225 trials — the optimized Sharpe of 5.81 is below the best-of-N luck bar and not significant (is_significant=false, PBO 0.8704 >> 0.5). This fails every gate: sensitivity FAILED with cliff_count 4 (all four parameters — trend_period, risk_fraction, stop_loss_pct, breakout_lookback — are cliffs, Sharpe swinging from +3.6 to negative on small changes); walk-forward is_overfitted=true with avg IS 7.75 collapsing to avg OOS 0.0 and ALL three OOS windows exactly [0.0, 0.0, 0.0] (the chosen config produces zero out-of-sample trades); and the holdout is catastrophically negative (holdout Sharpe -30.23 against a 0.0 WF-OOS reference, ratio 0). The positive full-period sharpe_ci_low (1.48) is a bootstrap artifact that does not correct for the 225-trial selection and is swamped by these failures. The edge is concentrated in 2021 (+984%) and 2024 (+184%) bull outliers with a deeply negative recent regime (2025 -11.5%, 2026 -12%) where the holdout lands and loses, as flagged at pre-optimization. This is overfit on a cliff-ridden space, not a tunable-region case: the sweep's best config does not trade out-of-sample and loses the holdout, so a second pass over the same space hits the identical wall — there is no robust region to tune toward. Not revise_hypothesis: the multi-week trend-continuation mechanism is validated on its BTC/ETH/LINK siblings, so SOL is a dead target rather than a premise needing reframing — and this is the second SOL daily multi-week variant (after the triple-momentum version, DSR 0.0068) to fail identically. FAILURE PATTERN: SOL daily multi-week breakout overfits in optimization — high in-sample Sharpe carried by 2021/2024 bull outliers, but all four parameters cliff, walk-forward OOS is zero, and the holdout is deeply negative; recent-regime decay (2025/2026 negative) guarantees holdout failure.

Long-only single-instrument trend-continuation on SOLUSDT (spot), DAILY bars, OHLCV-only. Enters long when the close prints a fresh 20-day high (breaks above the prior 20-day high) AND is above the 50-day SMA, confirming a multi-week uptrend. Rides the trend, exiting only when the close falls back below the 50-day SMA (trend break) or a 12% protective stop is hit — no fixed take-profit, so multi-week trends are allowed to run. Sized to 90% of equity, long-only, no leverage.

Strategy functions correctly and matches the hypothesis: fresh 20-day-high breakout + 50-day SMA trend gate, long-only, trend-break/stop exits. Big winners are real multi-month SOL bull rides.

Position sizing is sound (risk_fraction 0.90 of equity) — the >100% risk metrics are compounding/per-day-attribution artifacts, not a sizing bug.

PF 1.61, 25 trades; the multi-week trend family has validated BTC/ETH/LINK siblings.

Recent regime deeply negative (2025 -26%, 2026 -29%) — the worst of this batch; the holdout window lands in this dead zone.

Extreme edge concentration: one 2023-24 trade is 95.9% of dollar P&L; the headline 21653% return is the 2021 (+8016% annual) and 2024 (+435%) bull parabolas compounded.

Essentially a duplicate SOL daily multi-week trend mechanism — its triple-momentum sibling was just abandoned post-optimization (DSR 0.0068, all walk-forward OOS = 0).

Headline metrics nearly uninterpretable: Sharpe 1.27 with sharpe_ci_low -23.5, annualized_volatility 5202%, drawdown CI > 100% — all per-day-attribution/compounding distortion.

adequate

holdout lands here

1 trade = 95.9% of P&L

watch

yes

must match

Clean implementation matching the hypothesis (fresh 20-day-high breakout + multi-week SMA trend gate); big winners are real SOL bull rides.

Optimized full-period bootstrap shows sharpe_ci_low 1.48 (positive) — but it does not survive the multiple-testing, sensitivity, OOS, or holdout gates.

Deflated Sharpe 0.1458 vs the 7.84 best-of-225 luck bar; is_significant=false; PBO 0.8704 (very high).

Sensitivity FAILED with cliff_count 4 — all four parameters are cliffs (Sharpe swings +3.6 to negative on small changes).

Walk-forward is_overfitted=true: avg IS 7.75 collapses to avg OOS 0.0, all three OOS windows [0.0, 0.0, 0.0] (zero out-of-sample trades).

Holdout catastrophically negative (Sharpe -30.23, ratio 0 vs 0.0 WF-OOS reference).

Edge concentrated in 2021 (+984%) / 2024 (+184%) with deeply negative recent regime (2025 -11.5%, 2026 -12%) where the holdout lands.

expected-max 7.84

SolDailyMultiWeekBreakoutLong applied the multi-week trend-continuation template — promoting on its BTC/ETH/LINK siblings — to SOL to complete the family, with a clean 20-day-high-plus-SMA entry. It cleared the pre-optimization gate but reviewers flagged that one trade drove 95.9% of P&L, the recent regime was the worst of the batch, and it nearly duplicated a SOL triple-momentum variant just abandoned. Optimization confirmed total failure: deflated Sharpe 0.15, PBO 0.87, four sensitivity cliffs, an in-sample 7.75 collapsing to zero OOS activity, and a -30.23 holdout Sharpe. The analyst judged it overfit on a cliff-ridden space with no robust region and SOL a dead target, abandoning it; it reached optimization and analysis but was abandoned after a single iteration, never advancing to risk review.

SOL daily multi-week breakout overfits in optimization — its high in-sample Sharpe is carried by 2021/2024 bull outliers (one trade ~96% of P&L), all parameters cliff, walk-forward OOS goes to zero, and recent-regime decay (2025/2026 negative) guarantees holdout failure — making SOL the second daily multi-week variant to fail identically and a dead target for a family validated on BTC/ETH/LINK.

After passing the backtest-review gate (verdict: optimize), the analyst abandoned it at the post-optimization stage: it failed deflated Sharpe (DSR 0.15 vs expected-max luck bar 7.84 over 225 trials, is_significant=false, PBO 0.87), failed sensitivity with all 4 parameters as cliffs, was walk-forward overfit (avg IS 7.75 collapsing to all-zero OOS windows), and failed the holdout catastrophically (Sharpe -30.23).

A long-only multi-week trend-continuation strategy on SOLUSDT.BINANCE daily bars (OHLCV only) that entered on a fresh 20-day high while price held above its 50-day SMA, exiting on a trend break (close below the SMA) or a 12% stop and letting winners run — intended to complete the BTC/ETH/SOL/LINK multi-week trend family.

The base backtest over 2020-08-11 to 2026-05-18 showed an extreme, distorted headline across 25 trades — total return +21,654%, CAGR 189%, profit factor 1.61, Sharpe 1.27 (CI [-23.5, 4.83]), a 55% max drawdown, 32% win rate — with one 2023-24 trade accounting for 95.9% of dollar P&L and deeply negative recent years (2025, 2026); optimization (Sharpe 5.81, +2101%) produced zero out-of-sample trades and a catastrophic holdout.