Bitcoin 2026-08-16 daily brief — AI stance: bearish

Funding is cool, not crowded, and this is the most commonly misread number on the board. The open-interest-weighted rate across exchanges — the right figure for any market-wide claim — annualizes to 6.11%, against a structural baseline of roughly 11% per year that corresponds to the exchange-default 0.01% per eight hours. Longs are paying shorts, but at a little over half the rate they pay in an unremarkable market. Across the past two weeks the weighted rate peaked at roughly 10.5% annualized on 08-13 and has cooled to current levels; even that peak only reached the neutral baseline. Nobody is crowded here. Separately, one major venue prints 5.14% annualized, running roughly 0.96 percentage points cooler than the all-venue weighted figure — that venue simply carries a milder long lean than the aggregate does; it says nothing about a spread between two market-wide measures, and the venue is unnamed in this data so no geographic inference follows from it.

De-risked and unlevered, which shapes the character of the move more than its direction. Market-wide futures open interest stands at $47.43bn, down roughly 4% from $49.50bn on 08-05, while price fell 2.7% over the same span — positions closing into weakness, not new shorts stacking. Options open interest is $25.81bn, essentially flat across the fortnight. With weighted funding at half the neutral baseline and open interest contracting, there is no leverage overhang to cascade, so the expected path is a grind lower toward and through the range floor rather than an air-pocket. The corollary matters for anyone positioned short: without a crowded long book there is also less fuel for the violent flush that usually marks a tradeable low. The single venue's $1.456bn of open interest is roughly a thirtieth of the aggregate and is cited here only as a venue-level observation; it has been flat between $1.426bn and $1.457bn all fortnight. On the volatility surface, 30-day at-the-money implied sits at 33.99% in the 25.6th percentile of the past 90 days, against 30-day realized of 21.73% — a premium of roughly 12 percentage points, so the options market is pricing meaningfully more movement than has occurred, and the term structure is in contango with 90-day implied 3.66 points above 30-day, which is the normal shape and shows no near-dated stress bid.

The 08-16 print — $10.35m of long liquidations against $2.58m of short, a 0.25 ratio, total at the 7.95th percentile of the past year — is a Sunday reading and understates the week. The informative pattern is the down days: $35.10m of long liquidations on 08-10 as price broke below 64,000, and $29.88m on 08-13, versus short liquidations of $6.38m and $8.21m on those same sessions. The long side has repeatedly been the one paying, in size, on every downward impulse — five of the last seven sessions liquidated more longs than shorts. That is a market where each decline is being met by long stop-outs rather than short covering, which is confirmation of the direction rather than exhaustion of it.

The US-versus-offshore spot spread is at -10.66 bps, through the practical bottom-decile threshold of -10 bps, having made a 14-day low of -10.85 bps on 08-15. What matters more than the level is the path: the spread has widened steadily from -8.42 bps on 08-04, with the only meaningful narrowing being a brief -4.87 bps on 08-08. Twelve of thirteen sessions sit below -8 bps. A persistent US discount of this depth means the Coinbase-side book is consistently offered relative to offshore, and it corroborates the ETF tape directly — roughly 6,060 BTC of US spot ETF outflow over 08-10 to 08-14, with IBIT alone at about $55.5m on 08-14. Two independent measurements of the same thing: the American institutional channel is a net seller, and the offshore bid is what has been absorbing it near 62,800.

Options volume of $769.9m on 08-16 is a Sunday print against a $2.0-2.9bn weekday range and is not evidence of a demand collapse.

ALIGNED

Supply is the story and it is arriving through four channels at once. US spot ETFs shed roughly 6,060 BTC over the 08-10 to 08-14 sessions — about $390m, reversing the prior week's $853.54m of inflows — with three consecutive negative sessions through 08-14 (-964 BTC, -2,065 BTC, -894 BTC), led by IBIT at roughly $55.5m on 08-14 alone. That week's outflow is roughly 13 days of new issuance at the ~466 BTC/day rate implied by the flow-to-issuance data. The 7-day flow average flipped negative to -336 BTC on 08-14, having been +1,834 BTC on 08-07; the 30-day average is still positive at +354 BTC, so the deterioration is recent rather than structural. Miner economics are stressed: the Hash Ribbons ratio at 0.988487 sits in the 10th percentile with the 30-day hash average beneath the 60-day, and the Puell Multiple at 0.834323 is in the 31st percentile. Long-term holders are realizing losses persistently — aggregate SOPR has printed below 1.0 in 11 of the last 13 sessions, and long-term-holder SOPR has printed below 1.0 in 11 of 13 with a 0.708318 reading on 08-16, meaning coins moved at roughly 71 cents on the dollar of cost basis. Corporate treasury supply is a narrative input rather than a co-equal channel: Strategy disclosed 1,690 BTC sold in early August, about $106m at current spot, roughly a quarter of the week's ETF outflow, but it dragged the miner equity complex down 5-6% on 08-10. Two regulatory catalysts were pushed out of reach: the SEC cancelled its August 14 open meeting to propose Regulation Crypto with no new date, and the CLARITY Act slipped to a September 15 Senate cloture vote requiring 60 votes. On the anomaly side, the 08-16 transaction count of 831,035 is the 99.87th percentile of a 6,435-day history — a record-tier print on a Sunday, at the 10th percentile for fee-per-transaction (263.8 sats) and only the 46th percentile for unique output addresses (475,195). High in count, low in economic weight, narrow in breadth: that is not organic demand. The eCash hard fork targeted at block 964,000 around 08-21, which credits BTC holders a 1:1 balance on a separate chain, is one candidate explanation for pre-positioning activity, but the data does not establish causation.

Neutral with a residual hawkish tilt, and critically not easing. The policy rate stands at 3.63%, held at 3.50-3.75% at the July 28-29 meeting. Against July headline CPI of 3.4% year over year that is a real policy rate of roughly +0.2% — barely restrictive — yet no cut is priced anywhere on the visible curve. The relevant detail is what faded last week: CME FedWatch showed roughly a 71% probability of a hold at the next meeting as traders trimmed bets on a September rate *hike*, not a cut. Money supply is growing 5.53% year over year, only modestly above nominal GDP of roughly 4.9% (1.5% real Q2 growth plus 3.4% inflation), so there is no liquidity impulse either. The real-economy data is deteriorating without a policy response — University of Michigan consumer sentiment fell to 51.0 in August from 55.2 in July, a 7.6% monthly drop that missed the 54.5 consensus, and initial claims rose to 209k from 200k. That combination is stagflation-lite and it is not a backdrop that funds long-duration risk. In-horizon catalyst: FOMC minutes from the July meeting on 08-19. The live sentiment gauge is unusable this week — the Fear & Greed reading of 25 carries an as-of date of 2026-07-18, roughly a month stale, so it describes July positioning and should not be read as current.

The 10-year yield sits at 4.63%, off a 4.72% close on 08-11, with press reporting an intraday test of 19-month highs near 4.75% earlier that week. Against 3.4% CPI that is a real long rate of roughly +1.2% — the long end, not the policy rate, is where the tightening actually lives, and it is the single most direct macro headwind for a zero-cash-flow asset. There is no credit-spread feed in this dataset, so no credit read is offered; that gap matters because spreads are where a risk-appetite deterioration would show up first and it is currently unobservable here.

DXY at 99.499 as of 08-16, easing from a 100.02 print on 08-12 and down roughly 0.34% across the 14-day window, with the softening driven by weaker US data pushing front-end yields down and fading September hike bets. This is a mild dollar tailwind but a small one — a third of a percent over two weeks is noise relative to the 3.4% drawdown in spot over the same period. At this stage of a descending consolidation, a dollar drifting sideways in the high-99s is neither the constraint nor the catalyst; it is simply not the marginal variable.

Risk-on, and pointedly not transmitting. The S&P 500 closed at 7,785.76 on 08-14, within 0.2% of its 7,798.99 high on 08-13, with VIX at 14.25 on 08-14 versus 16.50 on 08-04. Equities have been at or near highs with volatility compressing throughout the exact two-week window in which spot fell from 65,029.73 on 08-09 to 62,837 on 08-16. No S&P or NDX data is available for 08-15 or 08-16, so the equity read is anchored on 08-14.

The immediate shelf is 62,780, the 30-day low, 0.09% below spot — a level that has already held once and is effectively being retested now. Below it the chart is thin down to 58,519, the 60-day and 90-day low, 6.87% away. Sizing that gap against volatility: at 30-day realized of 21.73% annualized a one-sigma seven-day move is about 3.01%, roughly ±1,890 points, giving a 60,950-64,730 band — which places 58,519 at about 2.3 sigma, a real but not base-case outcome. Priced off 30-day implied at 33.99%, one weekly sigma is about 4.71% and that same shelf sits at roughly 1.5 sigma, which is a materially higher probability. The honest reading is that the options market thinks a test of 58,519 within the week is plausible and the recent tape does not. The base case is neither: a grind between 62,780 and the 50-day at 63,580, with a break of the range floor opening a slide rather than a crash — because with weighted funding at half the neutral baseline and open interest already down 4% from 08-05, there is no leveraged long book whose forced unwind would produce a gap. The tail that argues against complacency is the volatility compression itself: 7-day realized at 13.25% against 90-day at 34.44% is a factor of 2.6, and implied volatility at the 6th percentile of the year means the market is not paying for protection into a week containing the FOMC minutes and a chain fork. Compressed volatility resolves; it just does not announce its direction, and a 4-5% single session in either direction would not be unusual once it does.

low