Solana regime monitor with the MA10/30/50 classifier used by the HYPE index. The 180-day backtest found NO TA signal currently passes the verify bar (3% mean fwd / 55% win / n≥5). This page documents the regime + the failure mode so the absence of signal is itself informative.
research/sol-signals.md for the full failure table.No PASS-grade SELL signal found for SOL. The buy/sell sweep tested 8 VWAP-vs-MA window pairs, RSI extremes (25/30/35 oversold; 65/70/75 overbought), and MACD signal-cross — in BOTH directions, across multiple regime-gate variants (ungated, TOPPING+BEAR, BULL+TOPPING). None of the short variants cleared the verify bar (mean ≤−3%, win ≥55%, n≥5).
Full rejection table: research/sol-buy-sell.md
Classifier: MA10/MA30/MA50 applied to daily close prices (same as the HYPE index page). BULL = all three rising in order; BEAR = reverse; RECOVERY = short-term up but medium still down; TOPPING = the reverse split.
Signal verify bar: mean forward 7d return ≥ 3%, win rate ≥ 55%, n ≥ 5 trigger days in the 180-day backtest window. Each PASS signal listed above cleared this bar.
Regime gate: Per refinement R2 (research/refinement.md), signals only register when the classifier is in BULL or RECOVERY (BULL-only for ZEC T5). This is the "trade only with the trend" filter — historically improves win rate without losing meaningful trigger count.
Source data: CoinGecko free-tier /coins/{id}/market_chart, daily candles, 240-day fetch (so MA50 has warmup before the 180-day backtest window). Server-cached for 30 minutes.
research/sol-signals.md, research/NOTES.md, research/refinement.md