SolDailyGoldenCrossMomentumLong

SOL Daily Golden Cross Momentum Long-Only (BINANCE USD-M Futures, Daily, OHLCV-Only)

A long-only, single-instrument, single-venue, OHLCV-ONLY trend-following strategy on SOLUSDT.BINANCE 1D bars. Same proven golden-cross mechanism as the landed BtcDailyGoldenCrossMomentumLong + EthDailyGoldenCrossMomentumLong — applied to SOL futures. Enter long when 50-day SMA crosses above 200-day SMA (golden cross); exit when 50-day SMA crosses back below 200-day SMA (death cross). Completes the BTC/ETH/SOL golden-cross trio on Binance USD-M. Per factory directive: 'BTCUSDT/ETHUSDT/SOLUSDT.BINANCE on a 1H/4H/1D timeframe is always safe'. SOL daily bars confirmed reliable via multiple landed SOL strategies. Three parameters total.

Directly replicates the landed BtcDailyGoldenCrossMomentumLong / EthDailyGoldenCrossMomentumLong mechanism on SOL, keeping the three headline parameters (fast=50, slow=200, plus sizing). The golden/death cross is expressed as a long-only regime (hold while fast>slow) rather than one-shot event detection, so the entry fires exactly at the golden cross and the exit at the death cross while remaining robust if the backtest window starts mid-regime. The signal is continuous and varying — the normalized SMA spread that crosses zero at the cross points — satisfying the 'continuous signal, thresholded decision' rule and giving the verifier a non-frozen signal. min_bars_required is 202 so no trade fires before the 200-day SMA is fully defined, and SOL daily history spans years so warmup is satisfied. Kept on SOLUSDT.BINANCE USD-M futures 1D as specified (a confirmed-safe path matching the BTC/ETH siblings); leverage=1.0 because the strategy uses no margin amplification (sizing never references leverage), avoiding the leverage_set_but_unused gate. A daily 50/200 trend regime holds for weeks-to-months, so the ~0.10% round-trip futures fee is negligible against the captured trend, keeping per-trade economics comfortably positive.

Structural golden-cross sparsity plus a money-losing, artifact-contaminated result. The 50/200-day golden cross fires only ~1/year on a single instrument's daily bars, producing just 6 trades over ~5 years -- far below the ~62-trade walk-forward floor; a 3-window optimization would leave ~2 trades/window, so the strategy cannot be validated. This is the identical structural-sparsity failure as the already-abandoned BNB (4 trades) and LINK (5-6 trades) golden-cross siblings. Beyond sparsity, the result is a net loser: profit_factor 0.698 (<1), expectancy negative, max_drawdown 75.8%, win_rate 0.167 (5 of 6 trades are -16% to -41% losses). The only positive year (2024 +310%) is an anomalous single-day artifact (return_kurtosis 727, an impossible +310.08% daily return for a long-only SOL position, with avg_position_pct 115.79% vs the 0.95 requested -- a suspected over-exposure/winner-balloon, symptom stated not independently verified). Not iterate: golden cross on daily is inherently sparse, and there is no parameter fix that keeps it a 50/200 golden cross while reaching an adequate sample (shortening the SMAs abandons the premise). The BTC/ETH golden-cross siblings landed on longer, cleaner-trending history; SOL's ~5-year sample is too sparse for the same mechanism. FAILURE PATTERN: 50/200-day golden-cross trend-following on a single crypto instrument's daily bars is structurally too sparse to optimize (~1 cross/year -> 4-6 trades over the available history) and the few trades are dominated by regime luck -- it cannot be validated regardless of the mechanism's pedigree, and a flashy headline is a single fat-tail (here an impossible +310% over-exposure-artifact day), not an edge.

Long-only single-instrument trend-following on SOLUSDT.BINANCE USD-M futures daily bars, OHLCV-only. It computes the 50-day and 200-day simple moving averages and holds a long position whenever the 50-day SMA is above the 200-day SMA: it enters at the golden cross (fast crosses above slow) and exits at the death cross (fast crosses below slow). calculate_signal returns the normalized SMA spread (fast - slow)/slow continuously each bar (zero-crossing at the cross points); should_enter buys when the spread is positive and should_exit sells when it is non-positive. Sizes 95% of equity per long position. Completes the BTC/ETH/SOL golden-cross trio on Binance USD-M.

Implemented correctly and trades as specified: 6 golden-cross long entries, clean entry_diag (6 signaled = 6 submitted, 0 dropped), metrics_reliable=true; matches the proven BTC/ETH golden-cross mechanism

Structurally far too sparse: only 6 trades over ~5 years (golden cross fires ~1/year) -- far below the ~62 walk-forward floor; same failure as the abandoned BNB (4) and LINK (5-6) golden-cross siblings

Money-losing: profit_factor 0.698 (<1), expectancy -7,490 (negative), total_return -5.1%, max_drawdown 75.8%, win_rate 0.167

Artifact-contaminated headline: return_kurtosis 727, an impossible +310.08% single day (2024-09-06), and avg_position_pct 115.79% (vs 0.95 requested) -- suspected over-exposure/winner-balloon; even so, 5 of 6 trades are large losses

~62 WF floor

moderate

negative (-7,490)

SolDailyGoldenCrossMomentumLong tried to complete the BTC/ETH/SOL golden-cross trio by porting the proven 50/200-day mechanism onto SOL daily bars on the factory-confirmed-safe data path. The code traded faithfully, but the mechanism is intrinsically sparse: it fired only 6 times in ~5 years — far below the walk-forward floor and the same wall that abandoned the BNB (4 trades) and LINK (5-6 trades) siblings. Worse, the record was a net loser (profit_factor 0.698, max drawdown 75.8%, 5 of 6 trades large losses) with a headline distorted by an impossible +310% single day and 115% average position size — a suspected over-exposure artifact rather than edge. The analyst abandoned it at backtest review without optimizing: there is no parameter that keeps it a 50/200 golden cross while reaching a validatable sample, confirming that single-instrument daily golden-cross transplants don't generalize to majors with only a few years of history.

50/200-day golden-cross trend-following on a single crypto instrument's daily bars is structurally too sparse to optimize (~1 cross/year → 4-6 trades over available history) and dominated by regime luck — it cannot be validated regardless of the mechanism's pedigree, and a flashy headline (here an impossible +310% over-exposure-artifact day) is a single fat-tail, not an edge; the BTC/ETH versions only landed because they had longer, cleaner-trending history.

The analyst issued an abandon verdict at backtest review on structural-sparsity grounds compounded by a money-losing, artifact-contaminated result: the 50/200 golden cross fires only ~1/year, so 6 trades over ~5 years is far below the ~62-trade walk-forward floor (~2 trades/window) and cannot be validated; densifying requires shortening the SMAs, which abandons the golden-cross premise — the identical failure already seen on the BNB and LINK siblings — so there is nothing to iterate.

A long-only, single-instrument, OHLCV-only trend-following strategy on SOLUSDT.BINANCE daily bars (3 parameters) — the same 50/200-day golden-cross mechanism as the landed BTC and ETH versions — holding long while the 50-day SMA is above the 200-day SMA (entering at the golden cross, exiting at the death cross), intended to complete the BTC/ETH/SOL golden-cross trio.

It implemented and traded correctly but was far too sparse and money-losing: only 6 trades over ~5 years, profit_factor 0.698, expectancy −7,490, total_return −5.1%, win_rate 16.7% (5 of 6 trades were −16% to −41% losses), and max drawdown 75.8%. The headline was artifact-contaminated — return_kurtosis 727, an impossible +310.08% single-day return (2024-09-06), and avg_position_pct 115.79% vs the 0.95 requested, pointing to a suspected over-exposure/winner-balloon (stated as a symptom, not independently verified).