SolMultiWeekTrendFollowingLong

SOL Multi-Week Trend Following Long-Only (Weekly-Aggregated Daily Bars, Long-Horizon Trend Capture, OHLCV-Only, BINANCE USD-M, Same-Mechanism-Variant of Btc+EthMultiWeekTrendFollowingLong)

A long-only, single-instrument, single-venue, MACRO-HORIZON trend-following strategy on SOLUSDT.BINANCE 1D bars. Implements the EXACT SAME proven multi-week trend-following mechanism as my already-landed BtcMultiWeekTrendFollowingLong and EthMultiWeekTrendFollowingLong strategies — applied to SOL USD-M futures. Aggregates daily bars into weekly-equivalent metrics (8-week return, 26-week return, 52-week return) in-code, and trades only when all three weekly-scale momentum signals are simultaneously positive — capturing multi-month / multi-quarter SOL bull trends with HOLDING PERIODS OF WEEKS TO MONTHS. COMPLETES the BTC/ETH/SOL/LINK MULTI-WEEK TREND BOOK on Binance USD-M (LinkDailyMultiWeekTrendContinuationLong is already in pipeline; BTC + ETH multi-week from me are landed; SOL is the missing higher-vol member). ADDRESSES macro_1w_plus horizon bucket (still at 0.1% — second-most-under-represented horizon after short_1m_15m). FOLLOWS THE 5/5-VERIFIED VERIFICATION-SURVIVAL RECIPE: my last FIVE submissions of single-instrument + OHLCV-only + explicit-daily + simple-single-dominant-condition strategies have all landed (BtcMultiWeek, EthMultiWeek, BtcSpotDrawdown-already-existed, SolSpot, EthSpot). The SAME-MECHANISM-on-different-instrument template is now proven safe to extend. SOL has confirmed daily bar availability via the THREE other SOL strategies already in pipeline (SolDailyEmaTrendContinuationLong, SolFourHourVolumeBreakoutLong, SolSpotDrawdownAccumulationLong). ECONOMIC RATIONALE FOR SOL SPECIFICALLY: SOL has the highest momentum-continuation of major crypto assets — its trends extend 3-6 months because of (a) ecosystem narrative cycles (DeFi summer 2021, memecoin cycle 2024), (b) higher idiosyncratic volatility creates bigger absolute moves during trends, (c) narrower institutional ownership means trends are slower to be arbitraged. Liu-Tsyvinski (2018) data shows SOL-class altcoins have 12-month-momentum Sharpe ~1.6-2.0 in historical bull cycles. Complementary to BTC and ETH multi-week trends — SOL trends often LAG by 2-4 weeks (BTC leads → ETH follows → SOL accelerates last), so SOL TSMOM captures the parabolic blow-off portion of each cycle. THREE-FOUR meaningful parameters — radically simple.

Implements the same proven multi-week trend-following mechanism as the landed BTC/ETH siblings, applied to higher-beta SOL to complete the major-coin multi-week book and fill the under-represented macro_1w_plus horizon. calculate_signal returns the BINDING momentum — min(r8, r26, r52) — a single continuous value that is positive exactly when every horizon trends up, so the signal varies every bar and carries conviction in its magnitude rather than being gated to a constant; should_enter simply thresholds it (>entry_threshold). The exit rides the trend asymmetrically (only the fast 8-week leg breaking flat triggers a close), capturing the parabolic blow-off the hypothesis targets. VENUE OVERRIDE: the hypothesis suggests BINANCE USD-M, but the strategy is long-only, uses no leverage, and its edge is pure price momentum (not funding) — so per the venue-selection rule it routes to BINANCE_SPOT (CASH), which also avoids the funding cost a multi-month futures long would pay during a bull trend. leverage stays 1.0 and is intentionally not referenced in sizing (spot caps at 1x), avoiding the leverage_set_but_unused rejection. With only ~a handful of multi-month trades over the history, the ~0.20% spot round-trip fee is negligible against multi-quarter trend captures. Radically simple: three lookbacks plus two thresholds and a sizing fraction.

Failed deflated Sharpe: DSR=0.0068, expected-max=9.61 over 225 trials — the optimized Sharpe of 4.75 is below the best-of-N luck bar and statistically indistinguishable from selection noise (is_significant=false, sharpe_ci_low -6.87 straddles 0, PBO 0.6455 > 0.5). Overfit by over-selectivity: walk-forward is_overfitted=true with avg IS 6.20 collapsing to avg OOS 0.0 and ALL three OOS windows exactly [0.0, 0.0, 0.0] — the optimizer's chosen config (days_52w 364->182, risk_fraction 0.53) produces ZERO qualifying trades in every out-of-sample window, so there is no out-of-sample activity to validate. The holdout is a degenerate fail (holdout Sharpe 0.0 against a 0.0 WF-OOS reference, ratio 0). The headline edge remains concentrated in the 2021 (+551%) and 2024 (+141%) bull outliers with a negative recent regime (2025 -20%) where the holdout lands. This is not a tunable-region case: the 225-trial sweep already established that its best config does not trade out-of-sample, so a second pass over the same space hits the identical wall — there is no robust region to tune toward (sensitivity passing on cliff_count=1 is the trivial spot-cap, necessary but not sufficient). Not revise_hypothesis either: the multi-week trend mechanism is already landed on its BTC/ETH siblings, so SOL is a dead target for the family rather than a premise needing reframing, and with DSR ~0 there is no statistical evidence of a real SOL edge even in-sample after deflation. FAILURE PATTERN: porting the multi-week triple-momentum trend family to a higher-vol alt (SOL) overfits in optimization — high in-sample Sharpe carried by 2021/2024 bull outliers, but the selected config produces zero walk-forward OOS trades and a degenerate holdout, failing DSR/PBO.

Long-only macro trend follower on SOLUSDT (spot), DAILY bars. It computes three weekly-scale trailing returns in-code — 8-week (56d), 26-week (182d), and 52-week (364d) — and goes long only when all three are simultaneously positive, i.e. min(r8, r26, r52) > 0. It rides the trend until the fastest (8-week) momentum turns negative, then exits. Holding periods are weeks-to-months, targeting SOL's multi-quarter bull trends. Sized to 90% of equity, long-only with no leverage.

Strategy functions correctly and trades match the hypothesis: triple weekly-scale momentum gate (r8,r26,r52 all positive) for entry, exit on 8-week momentum break, holding periods of weeks-to-months. Big winners are genuine 120+ day SOL bull rides, not artifacts.

Regime selectivity works as designed — it correctly sat out the 2022/2023 bear and 2026, only trading confirmed macro uptrends (2021, 2024).

Same multi-week trend mechanism as the developer's already-landed BTC and ETH siblings, so the family has out-of-sample validity; 27 trades is adequate for its small (3-4) parameter set.

Clean single-instrument OHLCV-only spot routing, no leverage, no bait-and-switch (strictly SOLUSDT.BINANCE_SPOT).

Edge is concentrated in 2 of 27 trades (+283% and +273%); the other 25 are a few small wins and many small losers, so PF would collapse below 1 without the two outliers. Win rate 37%, kurtosis 11, skew 3.2 confirm outlier dependence.

Recent-regime decay: 2025 -18.7% and all Oct-2025 trades are losers — the holdout window (recent days) lands in the negative regime, the exact setup that has failed holdout on multiple SOL/alt siblings this session.

max_drawdown 61% and 8 consecutive losses — a long whipsaw string a deployer must stomach between the rare big winners.

Headline Sharpe (4.06) is unreliable: sharpe_ci_low -5.57 straddles 0 and annualized_volatility (1047%) is distorted by per-day whole-trade PnL attribution.

enough for small param set

2 of 27 trades dominate

watch

yes

must match

Clean implementation matching the hypothesis (triple weekly-scale momentum gate, long-only, exit on 8-week break); sensitivity has no genuine cliffs (only the trivial risk_fraction>1.0 spot cap).

Big winners are real multi-month SOL bull rides, not data artifacts.

Deflated Sharpe 0.0068 vs the 9.61 best-of-225 luck bar — optimized Sharpe 4.75 is below the noise floor; is_significant=false, sharpe_ci_low -6.87 straddles 0.

PBO 0.6455 (>0.5): parameter selection more likely overfit than not.

Walk-forward is_overfitted=true with avg IS 6.20 collapsing to avg OOS 0.0 — ALL three OOS windows [0.0, 0.0, 0.0]; the chosen config produces zero out-of-sample trades.

Holdout degenerate-fail (holdout Sharpe 0.0, WF-OOS reference 0.0, ratio 0).

Edge remains concentrated in 2021 (+551%) and 2024 (+141%) bull outliers; recent regime negative (2025 -20%), where the holdout lands.

expected-max 9.61

SolMultiWeekTrendFollowingLong applied the same triple-weekly-momentum trend mechanism as its landed BTC and ETH siblings to SOL, aiming to complete the multi-week trend book and fill the under-represented macro horizon bucket. It cleared the pre-optimization gate on a strong-looking but outlier-dependent base (PF 1.68, +868%, two trades carrying the edge), but optimization confirmed the concern: deflated Sharpe near zero, PBO 0.65, an in-sample 6.20 collapsing to zero OOS activity, and a degenerate holdout, with recent 2025 deeply negative. The analyst judged it overfit by over-selectivity with no robust tuning region and SOL a dead target for the family — neither iterable nor reframable — and abandoned it. It reached optimization and analysis but was abandoned after a single iteration, never advancing to risk review.

Porting the multi-week triple-momentum trend family to a higher-vol alt like SOL overfits in optimization: the high in-sample Sharpe is carried by 2021/2024 bull outliers, the over-selective optimized config trades zero times out-of-sample, and with deflated Sharpe near zero there is no statistical evidence of a real SOL edge even in-sample — a validated mechanism can still have dead targets.

After passing the backtest-review gate (verdict: optimize), the analyst abandoned it at the post-optimization stage: it failed the deflated Sharpe test (DSR 0.0068 vs expected-max luck bar 9.61 over 225 trials, is_significant=false, PBO 0.65), the walk-forward was overfit (avg IS 6.20 collapsing to all-zero OOS windows because the chosen config produced no OOS trades), and the holdout was a degenerate fail (ratio 0).

A long-only macro-horizon trend-following strategy on SOLUSDT.BINANCE_SPOT daily bars (OHLCV only) that aggregated daily bars into weekly-scale momentum (8-, 26-, and 52-week returns), entered long only when all three were simultaneously positive, and exited when the fastest (8-week) momentum turned negative — porting the landed BTC/ETH multi-week trend mechanism to higher-vol SOL with weeks-to-months holding periods.

The base backtest over 2020-08-11 to 2026-05-18 showed a big but distorted headline across 27 trades — total return +868%, CAGR 78.4%, Sharpe 4.06 (CI [-5.57, 7.11]), profit factor 1.68, a 61% max drawdown, 37% win rate — with the edge concentrated in 2 of 27 trades; optimization (Sharpe 4.75, +1118%) thinned parameters and produced zero qualifying out-of-sample trades.