AvaxDailyTrendPullbackVolumeContractionLong

AVAX 1D Trend-Pullback Long with Volume Contraction Filter

Long-only daily strategy on AVAXUSDT perpetual futures using only OHLCV data. The hypothesis is that during confirmed multi-month uptrends (50-day SMA above 200-day SMA), pullbacks to the 20-day SMA accompanied by volume contraction represent high-probability re-entry points for trend continuation. AVAX is a high-beta L1 with deep futures liquidity, demonstrated multi-week trend persistence, and is not currently represented as a standalone strategy in the portfolio (only paired with SOL). The strategy uses three simple, well-established indicators (SMAs, ATR, volume average) with no exotic supplementary data dependencies, no contrarian premise to falsify, and a daily timeframe that keeps fee drag negligible relative to expected per-trade moves. Holds typically range 7-45 days; expected ~6-12 trades per year.

Implements the trend-pullback-with-volume-contraction mechanism exactly as specified, using only well-established OHLCV indicators (SMAs, ATR, volume average) with no contrarian premise and no supplementary-data dependency. VENUE OVERRIDE (noted per the developer venue rules): the hypothesis says 'perpetual futures', but the strategy is long-only, unleveraged, and holds 7-45 days, and on a perp a multi-week long pays substantial funding (~0.9%/month at typical positive rates) that would exceed the ~0.10% futures-vs-spot round-trip fee saving — so I route to BINANCE_SPOT (CASH, leverage 1.0), which is the correct venue for a long-only no-leverage multi-week-hold trend strategy and avoids the funding drag; the price action / edge is identical on spot. AVAX is an asset with prior sparse-signal abandonments, so I verified entry frequency on real AVAXUSDT daily history (2,065 days, 2020-2026): the regime+pullback structure fires ~6.0/year, squarely in the hypothesis's empirically-validated 6-12/year success zone and well clear of the 2-3/year sparsity-failure zone. The volume-contraction filter is deliberately lenient (1.2x average) because should_enter fires on the FIRST qualifying flat bar of a pullback, so the filter mostly shifts entry timing within a pullback rather than eliminating trades — it rejects genuine high-volume distribution dips while preserving signal density (~6/yr, ~34 round-trips over the window, comfortably above the >=5-per-walk-forward-window floor). The ATR chandelier stop lets winners run through the 7-45 day continuation while capping downside, and the regime-break exit prevents holding through a trend reversal. Verified locally: Layer 1 static (clean), Layer 2 synthetic (all 6 scenarios pass with a non-frozen pullback-depth signal), and the real-data entry-frequency check above.

Loses money on a sample too thin to optimize. profit_factor 0.85, Sharpe -0.53, expectancy -$369/trade, total_return -9.2%, negative in three of five years (2022/2023/2025) with the only positives resting on 1-2 outlier days (2024-01-06 +20%, 2024-03-09 +11%, 2021-12-05 +11%). Only 26 trades over ~5.5 years (~4.7/yr, below the hypothesized 6-12), so a 3-window walk-forward would see ~4 OOS trades/window and the optimizer would fit noise across 11 parameters on a negative-expectancy base. The pullback-in-uptrend re-entry premise has no edge on AVAX — it buys continuations that fail — and no parameter change can turn a losing, outlier-dependent 26-trade sample positive. This is the documented pullback-on-alt failure family (AVAX Calm-Regime / DOT Pullback abandons cited in the hypothesis itself): sparse signal + outlier-dependent positives + negative baseline. Code faithfully implements the premise, so this is abandon (premise has no edge here), not iterate. FAILURE PATTERN: long-only daily trend-pullback re-entry on a high-beta alt (AVAX) produces a sparse (~5/yr), negative-expectancy backtest whose few positive years hinge on isolated outlier days — too thin and too unprofitable to optimize.

Long-only AVAX daily trend-pullback re-entry strategy on AVAXUSDT.BINANCE_SPOT, OHLCV-only. In a confirmed multi-month uptrend (50-day SMA > 200-day SMA and close > 200-day SMA), it buys pullbacks to the 20-day SMA (the day's low reaches within touch_band=3% of the 20-SMA) when accompanied by VOLUME CONTRACTION (volume <= 1.2x the 20-day average, i.e. the dip is not a high-volume distribution event). It rides the continuation with an ATR(14) chandelier trailing stop (highest close since entry - 3*ATR) and exits on a trend-regime break (close < 200-SMA or 50-SMA < 200-SMA) or a 45-day time stop, giving the 7-45 day holds the hypothesis describes. Position size is 30% of equity per entry. calculate_signal returns a continuous pullback-depth diagnostic (close vs 20-SMA, non-frozen) while the discrete regime/pullback/volume gate lives in should_enter. Single instrument, long-only, leverage 1.0, three simple indicator families (SMAs, ATR, volume average), zero supplementary data.

Clean execution (26 signaled/submitted, 0 dropped), fully realized (end_unrealized 0.4%), metrics_reliable true

Sound, well-established mechanism design (SMA regime + pullback + ATR chandelier stop) with a hard stop from the start

Negative edge: profit_factor 0.85, Sharpe -0.53, Sortino -1.41, expectancy -$369/trade, total_return -9.2%

Negative in 3 of 5 years (2022 -10.8%, 2023 -8.1%, 2025 -14.4%); positives rest on 1-2 outlier days (2024-01-06 +20%, 2024-03-09 +11%)

Far too thin for a 3-window walk-forward: 26 trades over ~5.5 years (~4.7/yr, ~4 OOS/window) vs 11 parameters — optimization would fit noise on a losing base

AvaxDailyTrendPullbackVolumeContractionLong used a sound, well-established design — SMA regime gate plus a volume-contraction pullback entry and an ATR chandelier stop — to buy trend-continuation re-entries on AVAX. But the premise had no edge there: across just 26 trades over ~5.5 years it returned -9.2% with a -0.53 Sharpe, 0.85 profit factor, and -$369 expectancy, losing in three of five years with positives propped up by one or two outlier days. The analyst flagged it as the documented pullback-on-alt failure family — sparse signal, outlier-dependent positives, negative baseline, and far too thin for a 3-window walk-forward against 11 parameters — and abandoned it before optimization. It ended after one iteration as abandoned, never advancing to optimization or risk review.

Long-only daily trend-pullback re-entry on a high-beta alt (AVAX) produces a sparse (~5/year), negative-expectancy backtest whose few positive years hinge on isolated outlier days — part of the documented pullback-on-alt failure family — and a thin, outlier-dependent, losing sample cannot be made positive by parameter optimization.

It was abandoned at the pre-optimization backtest-review gate (verdict: abandon) because the pullback-in-uptrend premise had negative edge on AVAX and the 26-trade, ~4.7/year sample was far too thin to optimize 11 parameters without fitting noise on a losing base, so optimization and all later stages were never reached.

A long-only daily trend-pullback re-entry strategy on AVAX (OHLCV-only) that, during a confirmed uptrend (50-day SMA above 200-day SMA and close above the 200-day SMA), bought pullbacks to the 20-day SMA accompanied by volume contraction, then rode the continuation with an ATR chandelier trailing stop and exited on a regime break or time-stop.

Over ~5.5 years it took only 26 trades and lost money: total return -9.2%, Sharpe -0.53 (CI [-9.52, 4.76]), Sortino -1.41, profit factor 0.85, negative expectancy of -$369/trade, a 38.5% win rate, and max drawdown 32.7%; it was negative in 3 of 5 years (2022 -10.8%, 2023 -8.1%, 2025 -14.4%) with the few positives resting on 1-2 outlier days.