Position 2026-08-02: short (-0.15, confidence 0.35)

Our view for 2026-08-02 is a small net-short tilt on Bitcoin, held with low conviction. XBTUSD closed 2026-08-01 at $62,768, essentially flat on the day (-0.09%), down 2.40% over seven days but still up 2.08% over thirty, and roughly 50% below the $126,200 all-time high. The daily range was narrow ($62,237-$63,093) and 30-day annualized realized volatility sits near 29.9%. This is a chop tape, not a trend, so the honest expression of a negative lean is a small one. The cleanest bearish evidence is on the demand side. Spot ETF net flow printed -$265M on 2026-08-01, and the trailing seven-day sum is negative (-$61.5M) despite a positive thirty-day total (+$467M) - the recent impulse has rolled over. The per-issuer detail for 2026-07-31 shows a -$142.7M day that was broad-based rather than a single-fund artifact: FBTC -$54.8M, GBTC -$52.6M, BITB -$17.8M and ARKB -$17.5M, with no issuer printing an offsetting creation. Dollar-side fuel is contracting too: USDT supply is down about $4.0B over thirty days and USDC down roughly $2.0B, and the Coinbase spot premium gauge has been persistently negative (latest about -0.10 against a -0.11 seven-day average), which is not what a healthy US institutional bid looks like. Derivatives corroborate rather than contradict. Perpetual funding is flat-to-negative - BitMEX last printed -0.000485 per 8h with a negative seven-day average, while Binance, Bybit and Deribit are all barely above zero - yet exchange long/short account ratios have climbed (Binance 2.21 vs a 1.71 seven-day average, Bybit 1.67). Crowded long books that are not being paid to stay long are fragile. Options are quietly paying up for downside: the Deribit put/call ratio is 0.534 against a 0.470 seven-day average, and 25-delta skew has richened to 6.59 from a 4.88 average even as headline DVOL (35.6) drifted slightly lower - protection is being bought without a broad volatility bid. Open interest is being shed rather than added (BitMEX XBTUSD -16.2% on the day; the seven-venue aggregate -0.4%), and the last 24h of forced liquidations skewed toward shorts (long/short notional ratio 0.57), meaning a squeeze has already spent some of its fuel. On-chain, the overhead-supply picture is heavy: 55.8% of supply sits below its last-moved cost basis, the median UTXO is fractionally underwater, and MVRV has slipped 2.2% over seven days to 1.18 against a $53,309 realized-price anchor. The macro backdrop carries the most weight in this view, and it is subtly deteriorating beneath a calm surface. The MOVE index is up 8.1% in a week and 21.1% in a month to 83.0, and high-yield credit spreads have widened to 2.84 (+2.5% weekly, +3.3% monthly), while the VIX has fallen 13.9% to 15.99. Equity volatility priced for calm alongside rising rate volatility and widening credit is a combination that usually resolves against the higher-beta asset. More telling: the S&P rose 1.05% and the Nasdaq 1.59% over the past week, and the dollar index fell 1.6% - Bitcoin's trailing 60-day co-movement is positive to the Nasdaq and clearly negative to DXY, so both legs were tailwinds - and BTC still finished the week down 2.40%. An asset that declines into its own favorable drivers is telling you something about marginal demand. We keep the size small for good reasons. Sentiment is already washed out rather than complacent (Fear & Greed at 27, retail search interest near the floor at 16.7), which historically limits downside follow-through. Front-tenor gamma and open interest cluster at $62,000 and $60,000 just below spot, giving the tape structure to lean on. The Kalshi year-end distribution still carries an implied median of $67,500, above spot. And the calendar is empty of catalysts until the 2026-09-15/16 FOMC, where the market assigns roughly 58% to a 3.75-4.00% target and about 38% to 3.50-3.75% - policy is not the near-term driver. A modest short lean expresses the flow and macro deterioration without pretending to know more than a range-bound, low-volatility tape actually reveals.