AlphaBrief — crypto algorithmic analysis

Daily Alpha Brief — 2026-07-21 — SIDEWAYS regime across majors

Bayesian crypto brief for 2026-07-21: SIDEWAYS regime, strategy performance, weight shifts, and quant-vs-crowd sentiment across 9 symbols.

TL;DR

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

Current Bayesian strategy weights Current Bayesian reliability weights, averaged across the universe.

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

🌐 Market Indices

Fear & GreedBTC DominanceTotal Market Cap24h Volume
37/100 — Fear58.8% (balanced)$2.23T (+0.88%)$66B

Crowd reads fear at 37/100; BTC dominance 58.8% signals balanced.

🌐 Market Overview

Economic (FRED):

Yield CurveFed RateM2 YoY
🟡 +0.37% (Flat)3.63%+5.6%

Stablecoin dominance 11.1% (risk-off, money fleeing crypto); yield curve regime Flat.

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

BTCETHSOLTop Mover
+0.005%+0.001%+0.010%BTC (+0.005%, balanced)

Economic Calendar: Next FOMC: in 8 days | Next CPI: in 22 days | position-size modifier 1.00×.

📊 Regime Analysis

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

SymbolRegimeSignal
AVAX/USDTBEAREXECUTE
BNB/USDTSIDEWAYSWATCH
BTC/USDTSIDEWAYSWATCH
DOGE/USDTBEAREXECUTE
ETH/USDTSIDEWAYSWATCH
SOL/USDTSIDEWAYSWATCH
SUI/USDTSIDEWAYSWATCH
WIF/USDTSIDEWAYSWATCH
XRP/USDTSIDEWAYSWATCH

The book is broadly bearish: 0 BULL / 2 BEAR of 9 symbols.

🔗 Correlation Signals

Stable pairs (r > 0.70): BTC/SOL: r=0.87 ✅ | ETH/SOL: r=0.87 ✅ | SOL/DOGE: r=0.86 ✅ | BTC/ETH: r=0.85 ✅ | SOL/XRP: r=0.85 ✅ | BTC/DOGE: r=0.84 ✅

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.

StrategyAvg P&L / tradeWin RateAvg SharpeEdgeTrades
bollinger+3.0431%-1.01+0.95122
post_shock_reversal+2.4628%4.17+0.70398
tsmom+1.4926%0.53+0.3957
volume_spike+0.7926%0.99+0.21604
rsi_macd+0.000%0.00+0.000
ema_cross+0.000%0.00+0.000
sentiment_strategy+0.000%0.00+0.000
correlation_strategy+0.000%0.00+0.000
macro_strategy+0.000%0.00+0.000
walk_forward_ema-5.2026%-4.62-1.3719

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

🧠 Bayesian Weight Shifts

SymbolTop moverWeight ΔLead strategy (conf)
SUI/USDTbollinger-0.259volume_spike (64%)
ETH/USDTpost_shock_reversal-0.243sentiment_strategy (64%)
BNB/USDTtsmom-0.208bollinger (77%)

SUI/USDT (SIDEWAYS):

post_shock_reversal is the standout performer with 70 trades, 34% win-rate, Sharpe 12.86, and +483.93 pnl on 7d — though win-rate is modest, the strongly positive risk-adjusted return and large pnl justify a meaningful raise, and it aligns with the SIDEWAYS/mean-reversion regime; raised to 0.65. bollinger has 7d win-rate of 25% and Sharpe of -2.48 with negative pnl, triggering rule 2 (persistent underperformance) and rule 3 (negative Sharpe), so it is cut despite nominal SIDEWAYS regime alignment — the data overrides the prior per rule 6. walk_forward_ema has adequate 7d sample (9 trades, borderline rule 1) with positive Sharpe 7.00 and +128.75 pnl, earning a modest raise to 0.53; however volume_spike shows a contradictory pattern — very high Sharpe but only 28% win-rate on 102 trades suggesting large-loss tail risk, so it is trimmed slightly to 0.45 from 0.567. All zero-trade strategies (rsi_macd, ema_cross, tsmom, correlation_strategy, macro_strategy) are held near their current weight per rule 1; sentiment is nudged down slightly to 0.45 reflecting sparse coverage (2 sources) and weak 50% confluence.

ETH/USDT (SIDEWAYS):

volume_spike is the dominant data signal: 7d win-rate is exactly 30% (borderline rule 2) but Sharpe of 6.59 and +277 pnl are strongly positive, so I keep it elevated but trim modestly from 0.701 given the 24h collapse (-85 pnl, 11% win) signals recent deterioration and the SIDEWAYS regime gets it a small bump per rule 5. post_shock_reversal is cut hard: 7d win-rate 22% and Sharpe -4.72 with -170 pnl triggers rule 2 (persistent underperformance), and despite being a mean-reversion strategy nominally suited to SIDEWAYS the data clearly overrides the regime prior (rule 6). bollinger also carries a deeply negative 7d Sharpe (-3.65) and negative pnl, so despite SIDEWAYS regime alignment the data win applies and it is reduced per rule 3; walk_forward_ema has only 1 7d trade with a catastrophic Sharpe so rule 1 keeps it near current with a minimal trim. rsi_macd, ema_cross, tsmom, correlation_strategy, and macro_strategy all have zero trades on 7d, so rule 1 holds them within +/-0.05 of current; sentiment_strategy shows moderate coverage (3 sources, 67% confluence, mildly positive composite) warranting a small trim from its elevated 0.648 to reflect only modest signal strength.

BNB/USDT (SIDEWAYS):

rsi_macd, ema_cross, walk_forward_ema, correlation_strategy, and macro_strategy all have zero trades across both windows, so rule 1 applies and weights are held near their current 0.263 level. bollinger has 18 7d trades with a 33% win-rate which breaches the persistent underperformance threshold (rule 2), so despite the SIDEWAYS regime alignment I reduce it from 0.624 toward 0.55 — the positive pnl and anomalously high Sharpe (likely artefact of small pnl/low volatility scaling) prevent a deeper cut but cannot override the weak win-rate. volume_spike and tsmom both show 7d win-rates near 22% with negative 7d Sharpe and negative pnl, triggering rules 2-3 for cuts; tsmom’s 24h Sharpe of -14.50 further reinforces the reduction. post_shock_reversal has abundant trades (70 on 7d) and positive pnl (+189) in a SIDEWAYS regime that suits mean-reversion, but a 31% win-rate prevents a raise and keeps it near current; sentiment shows 3-source coverage with 67% 7d confluence warranting a modest hold near current weight.

📡 Signal Outlook

SymbolDecisionConfidenceLeadConfluence
AVAX/USDTEXECUTE61%sentiment_strategy (0.61)⚠️ Divergence
BNB/USDTWATCH77%bollinger (0.77)
BTC/USDTWATCH61%sentiment_strategy (0.61)
DOGE/USDTEXECUTE63%sentiment_strategy (0.63)⚡ Confluence
ETH/USDTWATCH64%sentiment_strategy (0.64)
SOL/USDTWATCH64%sentiment_strategy (0.64)
SUI/USDTWATCH64%volume_spike (0.64)
WIF/USDTWATCH64%bollinger (0.64)
XRP/USDTWATCH65%sentiment_strategy (0.65)

💼 Shadow Portfolio

MetricValue
Total Value$745.97
P&L since 2026-06-14-254.03 (-25.40%)
vs BTC Buy&Hold-24.69% (B&H -0.71%)
vs ETH Buy&Hold-35.99% (B&H +10.59%)
Open Positions1
Win Rate34.6% (1398 closed)

Open positions:

SymbolSideEntryMarkUnreal. P&L
DOGE/USDTshort0.072250.07231-0.12

$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.)

SymbolSentimentCrowdConfluence
AVAX/USDT-1.00bearish🔴 crowd bearish
BNB/USDT-0.51bearish🔴 crowd bearish
BTC/USDT-0.26bearish🔴 crowd bearish
DOGE/USDT+0.36bullish🟢 crowd bullish
ETH/USDT-0.31bearish🔴 crowd bearish
SOL/USDT-0.38bearish🔴 crowd bearish
SUI/USDT-1.00bearish🔴 crowd bearish
WIF/USDT-1.00bearish🔴 crowd bearish
XRP/USDT-0.16bearish🔴 crowd bearish

Most-talked alt: WIF/USDT at -1.00 (bearish crowd lean).

📰 Today’s Key Reads

Curated from 3 sources monitored today:

🔴 Allbridge Core pauses protocol after $1.65 million flash loan exploit: onchain analyststheblock — Score: -0.81

🟢 Hut 8 fully commercializes 1 GW Texas AI campus with second $9.8B lease as IREN signs $2.8B in contracts; shares climb double digitstheblock — Score: +0.39

⚠️ These underperforming trades could yield big returns over next six monthscnbc_finance — Score: +0.24

AI frenzy losing steam leaves bitcoin less volatile than South Korean stockscoindesk — Score: -0.19

DOG Mode explains Bitcoin’s next governance fightcoindesk — Score: -0.11

Academic Footnote

Miralles-Quirós (2022) — post-shock reversals in digital assets.

Disclaimer

This is algorithmic analysis only, not financial advice.