Daily Alpha Brief — 2026-09-01 — BULL regime across majors

Bayesian crypto brief for 2026-09-01: BULL regime, strategy performance, weight shifts, and quant-vs-crowd sentiment across 9 symbols.

TL;DR

• Dominant regime: BULL (5 bullish / 4 bearish of 9 symbols).
• Top signal: tsmom on BTC raised 0.334 (0.32 → 0.65).
• Shadow portfolio: $727 (-27.32%), 0 open, 35% win rate.

Shadow portfolio value since inception (dashed line = $1,000 starting capital).

Current Bayesian reliability weights, averaged across the universe.

Detected regime per symbol (EMA200 + ADX); % = lead-strategy conviction.

🌐 Market Indices

Crowd reads greed at 75/100; BTC dominance 59.8% signals balanced.

🌐 Market Overview

Economic (FRED):

Stablecoin dominance 9.6% (risk-on); yield curve regime Flat.

Funding Rates (contrarian — heavy positive = crowded longs to fade):

Economic Calendar: Next FOMC: in 15 days | Next CPI: in 10 days | position-size modifier 1.00×.

📊 Regime Analysis

Regime per symbol from EMA200 + ADX on daily candles, with the fusion signal.

The book is broadly bullish: 5 BULL / 4 BEAR of 9 symbols.

🔗 Correlation Signals

Stable pairs (r > 0.70): DOGE/SUI: r=0.89 ✅ | DOGE/SUI: r=0.89 ✅ | SUI/AVAX: r=0.89 ✅ | XRP/DOGE: r=0.89 ✅ | SUI/AVAX: r=0.89 ✅ | DOGE/AVAX: r=0.89 ✅

Index signals: BTC dominance NEUTRAL → no rotation; Fear & Greed contrarian NONE.

⚙️ Strategy Performance

Aggregate 7-day performance per strategy (replayed across the universe). Edge = win-rate × avg P&L per trade — the table is sorted by it.

All P&L figures normalized to a 100 USDT position size for comparability (not summed across symbols).

🧠 Bayesian Weight Shifts

BTC/USDT (BULL):

Nearly every price-based strategy has terrible 7d metrics: bollinger (29% win, -6.41 Sharpe, -238 pnl), post_shock_reversal (30% win, -5.91 Sharpe, -242 pnl), volume_spike (34% win, -4.06 Sharpe, -110 pnl), and walk_forward_ema (25% win, -3.43 Sharpe, -79 pnl) all trigger rules 2-3 and are cut sharply; ema_cross similarly underperforms on 7d (-3.67 Sharpe, 31% win) despite a noisy 24h spike, so it is reduced from 0.488 though less aggressively given its BULL-regime alignment. tsmom is the standout exception with positive 7d Sharpe (5.66) and pnl (+130), consistent with Shen et al. momentum findings in a confirmed BULL regime, earning a meaningful raise. rsi_macd and the three non-price voters (sentiment, correlation, macro) have zero trades or no new signal; sentiment_strategy is nudged down slightly given zero 24h source coverage and a near-neutral composite, while correlation and macro are held at current weight per the low-signal hold rule.

BNB/USDT (BULL):

In this BULL regime nearly every price-based strategy is losing money badly on 7d: ema_cross (29% win, -3.63 Sharpe, -87 pnl), volume_spike (27% win, -13.13 Sharpe, -413 pnl), tsmom (18% win, -9.32 Sharpe, -168 pnl), post_shock_reversal (28% win, -6.16 Sharpe, -221 pnl), and walk_forward_ema (33% win, -0.59 Sharpe, -13 pnl) all trigger rules 2-3 and are cut hard — this is a clear data-over-regime conflict (rule 6) for ema_cross and tsmom especially, which should thrive in BULL but are losing. Bollinger hits rule 2 with 30% win and -5.67 Sharpe (-165 pnl) so it is cut to near-floor; rsi_macd has the only survivable 7d Sharpe (3.73) despite a 40% win rate and positive pnl so it is modestly reduced from its inflated current weight rather than slashed. Sentiment_strategy had zero sources in 24h and near-zero composite signal on 7d so it is held close to current per the sparse-data guidance; correlation_strategy and macro_strategy report zero activity on both windows and are held near their current weights with only a token rounding.

ETH/USDT (BULL):

rsi_macd and walk_forward_ema are cut hard: both show 7d win rates well below 30% with deeply negative Sharpe and negative pnl (rules 2-3), and ema_cross similarly posts a 34% win rate with negative 7d Sharpe despite the BULL regime — a clear data-overrides-regime conflict (rule 6) warranting a meaningful cut. bollinger is the outlier performer with 48 trades, positive 7d Sharpe of 6.75 and +286 pnl; although a BULL regime is not its natural habitat, the data is strong enough to hold it near current weight with a small lift. post_shock_reversal’s 24h looks good but the 7d tells the real story: 33% win, negative Sharpe, negative pnl, so it is cut per rules 2-3. tsmom and correlation_strategy have zero trades across both windows so the low-sample guard (rule 1) keeps them near current weight; sentiment_strategy shows 7d coverage of 3 sources with 100% confluence and a mildly positive composite, justifying a small hold near its current level rather than a cut despite the 24h blackout.

📡 Signal Outlook

💼 Shadow Portfolio

$1,000 USDT paper portfolio, 1% risk/trade, MEXC fees modelled. No real capital is deployed.

📡 Discourse Radar

Sentiment analysis pipeline launching soon.

Academic Footnote

Fang et al. (2020) — a survey of cryptocurrency trading strategies.

Disclaimer

This is algorithmic analysis only, not financial advice.