Daily Alpha Brief — 2026-09-04 — BULL regime across majors

Bayesian crypto brief for 2026-09-04: 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: ema_cross on BNB raised 0.274 (0.39 → 0.67).
• Shadow portfolio: $732 (-26.83%), 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 72/100; BTC dominance 59.6% signals balanced.

🌐 Market Overview

Economic (FRED):

Stablecoin dominance 9.8% (risk-on); yield curve regime Flat.

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

Economic Calendar: Next FOMC: in 12 days | Next CPI: in 7 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): XRP/DOGE: r=0.89 ✅ | DOGE/AVAX: r=0.87 ✅ | BTC/ETH: r=0.87 ✅ | DOGE/SUI: r=0.87 ✅ | XRP/SUI: r=0.87 ✅ | SUI/AVAX: r=0.87 ✅

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

BNB/USDT (BULL):

walk_forward_ema is the clear standout with 54% win-rate, a Sharpe of 10.83, and +161 pnl on 13 7d trades (just above the low-sample threshold), so I raised it further toward 0.80 in line with rule 4 and BULL regime alignment. rsi_macd and volume_spike both have 7d win-rates below 30% with deeply negative Sharpes (-4.97 and -9.91 respectively) and negative pnl, triggering rule 2 cuts toward 0.10-0.20 despite the BULL regime — a clear data-over-prior conflict per rule 6. tsmom shows a near-zero but negative Sharpe and sub-30% win-rate, so it is nudged down modestly; ema_cross has a worrying 33% 7d win-rate that contradicts its strong 24h numbers, keeping it only slightly above current; sentiment_strategy earns a small bump on solid 7d coverage (4 sources), 75% confluence, and mild positive composite; correlation_strategy and macro_strategy report no signals and are held close to their current weights per the sparse-signal guidance.

SOL/USDT (BULL):

rsi_macd is cut sharply from 0.648: 7d win-rate of 32% and Sharpe of -2.11 with negative pnl triggers rules 2-3, and this is a conflict with the BULL regime (rule 6 — data wins). ema_cross and walk_forward_ema are both deeply negative (7d Sharpe -10.31, win ~25-26%, large negative pnl), so they are cut toward the floor per rules 2-3 despite the BULL prior — a clear data-beats-regime conflict. volume_spike shows a deceptively positive 24h but its 7d Sharpe is -7.75 with -539 pnl and only 28% win-rate, so it is also cut to near-floor. bollinger is the standout anomaly: 7d Sharpe of 7.18 with +462 pnl and 42% win-rate in a BULL market warrants a raise per rule 4 even though mean-reversion is regime-disfavoured; post_shock_reversal similarly shows positive 7d Sharpe (4.36) and pnl (+307) so it receives a modest lift. sentiment_strategy earns a small bump given 100% confluence, 4-source coverage, and consistent positive composite. tsmom, correlation_strategy, and macro_strategy have zero trades and are held near current weights per rule 1.

BTC/USDT (BULL):

In a confirmed BULL regime, tsmom is the standout performer with 7d Sharpe of 2.62 and positive pnl (+62.47) on adequate sample (14 trades), earning a modest raise from 0.642 to 0.68 per rules 4 and 5. Bollinger and volume_spike are cut hard — both show 7d win-rates below 30% with deeply negative Sharpes (-10.29 and -7.80) and large negative pnl, triggering rules 2 and 3; bollinger is particularly penalised as a mean-reversion strategy in a BULL trend (band-walk risk). rsi_macd similarly has a 7d Sharpe of -4.70 with a 33% win-rate so it is reduced; ema_cross and walk_forward_ema have negative 7d Sharpes despite 24h noise, so they are trimmed but kept above floor given regime alignment and the 24h flash suggesting some momentum. Sentiment_strategy is nudged down slightly from 0.656 to 0.62 because 24h confluence was only 33% and the composite signal is nearly flat (+0.07), while the 7d 100% confluence lends moderate trust; correlation_strategy and macro_strategy had zero activity on both windows so they are held near current weight with minimal change per the no-data guidance.

📡 Signal Outlook

💼 Shadow Portfolio

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

📡 Discourse Radar

Crowd-vs-quant sentiment from CryptoCompare news, crypto RSS, the macro feeds (FT / Fed / CNBC), and the Fear & Greed index. (Reddit retired — Responsible Builder Policy restrictions, 2026.)

• Fear & Greed: +0.50 (bullish crowd).
• Macro signal (FT/Fed/CNBC): +0.11 (bullish) — “Federal Reserve Board announces approval of the application by Banco Santander, S.A. and Santander H”.

Most-talked alt: BNB/USDT at +0.50 (bullish crowd lean).

📰 Today’s Key Reads

Curated from 3 sources monitored today:

⚠️ Hargreaves Lansdown offers crypto ETNs nearly a year after ban was lifted — ft_markets — Score: -0.99

🔴 Solana, ether, xrp lead majors slide as Iran strikes drive a broad risk selloff — coindesk — Score: -0.56

🔴 Dogecoin becomes only losing bet for Japan-listed firm as it sells altcoins for bitcoin — coindesk — Score: -0.38

🟢 Hyperscale Data shares hit all-time low as company flips Michigan site from bitcoin mining into AI — theblock — Score: +0.32

🟢 Japan-listed Remixpoint sells all ETH, SOL, XRP and DOGE holdings in shift to bitcoin-only crypto strategy — theblock — Score: +0.23

Academic Footnote

Krauss (2017) — statistical arbitrage pairs trading.

Disclaimer

This is algorithmic analysis only, not financial advice.