Net Fed liquidity rising 3 weeks straight — the tailwind BTC is trading on
Fed balance sheet − TGA − RRP, $T · weekly to 2026-10-14 · last 12 weeks highlighted
FRED (WALCL, WTREGEN, RRPONTSYD) · 2026-10-19
Where we are
weekly · 2y · higher timeframe 3BTC +22% vs the 200W, 10 weeks above it — 26% off the 52w high
weekly close, USD · 104w · anchor: 200W MA
Kraken · 2026-10-19
Alts 21% below the 52w high, but holding up 7 pts better than BTC
TOTAL3 market cap, $B · 52w · top-100 alts ex BTC/ETH scaled to CoinGecko global · anchor: 52w high
CoinGecko · 2026-10-19
Calendar, in context
daily · 90d back · 30d forward 1BTC 13% above the 200D; opex max-pain sits 5% below spot
BTC daily, USD · anchor: 200D MA, Sep opex max pain
- 1 29 JUL FOMC · hold, 2 dissents -3.1% /2d
- 2 12 AUG CPI · cool +4.0% /2d
- 3 28 AUG Jackson Hole · Warsh: hike on the table -3.2% /2d
- 4 02 SEPT ARB unlock · 2.1% supply —
- 5 04 SEPT NFP · 162k vs 53k -1.2% /2d
- 6 09 SEPT NVDA earnings —
- 7 11 SEPT CPI · last before FOMC +0.3% /2d
- 8 15 SEPT CLARITY Act · cloture failed 49–50 -3.3% /1d
- 9 16 SEPT FOMC · 25bp to 3.75–4%, 16 of 18 dots want more +0.7% /1d
- 10 17 SEPT SEC · tokenised-stock exemption issued; 24h roundtable +21% /1d
- 11 18 SEPT BoJ · 25bp to 1.25% on a 7–2 vote, no guidance · yen fell +5.9% /1d
- 12 18 SEPT Triple witching · $6.2T of US options · BTC +5.9% +5.9% /1d
- 13 23 SEPT Flash PMIs 58.7 / 56.7 · 10Y 5.18%, a cycle high −2.1% /2d
- 14 25 SEPT Quarterly opex · 189k BTC expired at 84.1k, max pain 73k missed by 15% −0.3% /1d
- 15 28 SEPT Hormuz offer rejected · $486M liquidated, the largest day since Sep 15, and the wick held 82.6k −1.1% /1d
- 16 29 SEPT Robinhood Chain gas subsidy ends · replaced the same day for Wallet swaps to Dec 31 · volume held $1.6B +2.4% /2d
- 17 30 SEPT Aug core PCE 3.0% vs 3.3% · Oct hike odds 69% → 34%, a day after the 10Y's 5.28% high +1.0% intraday /1d
- 18 02 OCT September payrolls · 29k vs 90k, unemployment 4.2%, −60k revisions · Oct hike odds 34% → 17% +2.0% /2d
- 19 08 OCT $1.1B liquidated, 94% longs · BTC wicked 80.3k and closed 81.7k, $230 above the 78.2–80.1k retest rung −1.9% /1d
- 20 10 OCT Houthi strikes on Riyadh airport · Brent +1.8% to $106.65 Monday, tankers hit off Oman; BTC −1.0% Friday to Monday −1.0% /2d
- 21 14 OCT September CPI · 0.6% m/m, 3.7% y/y; core 0.3%, 2.6% vs 0.2%, 2.5% · Oct hike odds 19% → 54% intraday, the 10Y lower anyway −1.8% /1d
- 22 15 OCT September PPI 1.3% m/m vs 0.6% · $0.95B liquidated, 90% longs; BTC wicked 79,709 and closed 82,134, green on the day +0.7% /1d
- 23 16 OCT Warsh gives no rate signal in Bangkok; Oct hike odds to 15% · Nasdaq −2.5%, VIX 20, HY 3.31% · BTC −0.8% −0.8% /1d
- 24 27 OCT HOOD Q3 · Oct 27 after the close, the first quarter with chain revenue watch
- 25 28 OCT FOMC · a second hike 15% priced after Warsh, from 54% intraday on CPI day; December 80–90% watch
- 26 29 OCT Q3 GDP advance and September PCE · the last prints before the Oct 30 expiry, the day after the FOMC watch
- 27 30 OCT Monthly opex · 149k BTC, max pain 78k, calls 2:1, 72k BTC of them at 90k and over watch
Kraken · CoinGecko · FRED · entities/events · 2026-10-19
Findings
agent investigations · decay unless re-bumped 5Core CPI and credit fired the rates break; Warsh and a 20bp drop in the 10Y unfired it. BTC held the 50D for eleven sessions
US 10Y (yellow) and 2Y yield, %, 90 sessions to Oct 15 · dashed = 5.00% and the Oct 5 high · FRED
US spot BTC ETFs, daily net flow, $M · Oct 5–16 · the week after the flush netted −$228M: +$176M on Oct 13 bought the dip, −$261M on CPI day sold it · Oct 12 derived from the week's total · Farside
Deribit Oct 30 monthly, open interest by strike, BTC · up = calls, yellow = 90k and over · down = puts · 149k BTC, 0.51 puts per call, max pain 78k · 72k BTC of calls need +10% in eight sessions
where BTC trades by Oct 30, per path · bar = the range · spot 81.9k · odds are the board's, re-weighted after the CPI week
- 1 08 OCT $1.1B liquidated, 94% longs · BTC wicked 80.3k and closed 81.7k, $230 above the 78.2–80.1k retest rung −1.9% /1d
- 2 14 OCT September CPI · 0.6% m/m, 3.7% y/y; core 0.3%, 2.6% vs 0.2%, 2.5% · Oct hike odds 19% → 54% intraday, the 10Y lower anyway −1.8% /1d
- 3 15 OCT September PPI 1.3% m/m vs 0.6% · $0.95B liquidated, 90% longs; BTC wicked 79,709 and closed 82,134, green on the day +0.7% /1d
- 4 16 OCT Warsh gives no rate signal in Bangkok; Oct hike odds to 15% · Nasdaq −2.5%, VIX 20, HY 3.31% · BTC −0.8% −0.8% /1d
- 5 27 OCT HOOD Q3 · Oct 27 after the close, the first quarter with chain revenue watch
- 6 28 OCT FOMC · a second hike 15% priced after Warsh, from 54% intraday on CPI day; December 80–90% watch
- 7 29 OCT Q3 GDP advance and September PCE · the last prints before the Oct 30 expiry, the day after the FOMC watch
- 8 30 OCT Monthly opex · 149k BTC, max pain 78k, calls 2:1, 72k BTC of them at 90k and over watch
BASE ~45%: 78–86k into Oct 28, 78k max pain below, 72k BTC of 90k+ calls above · RATES BREAK ~30%: under 79.7k toward the 200D, 72.6k · EXTENSION ~25%: over 87.4k on an Oct 28 hold
a daily close over 87.4k with five sessions of ETF inflows: the extension; a close under 79.7k, or HY past 3.50% with the Nasdaq still falling: the rates break
Scored since Oct 13. The data fired; the price did not.
| tripwire | fired | side |
|---|---|---|
| Oct 14 core CPI ≥ 0.3% | 0.3%, 2.6% y/y; PPI 1.3% the next morning | rates break |
| HY spread > 3.25% | 3.31% on Oct 16, from 3.15% | rates break |
| 10Y > 5.35% | no: 5.19% on CPI day, 5.10% by Oct 15, 5.07% on Oct 19 | — |
| ETF flow, 5 sessions, < −$1.0B | no: −$0.23B, with +$176M on Oct 13 | — |
| funding ≤ 0.01% with OI rising | 0.006–0.010% with OKX OI 29.5k → 30.0k BTC | extension |
| BTC close > 87.4k or < 80.1k | neither: low 79,709 on Oct 15, closes 81.4–83.1k | — |
Tripwires, re-cut. The lower line moves to the Oct 15 wick.
| watch | extension | rates break |
|---|---|---|
| BTC daily close | > 87.4k | < 79.7k |
| 10Y (5.07%) | < 4.90% | > 5.35% |
| HY spread (3.31%) | < 3.10% | > 3.50% |
| ETF flow, 5 sessions (−$0.2B) | > +$1.0B | < −$1.0B |
| Oct 28 FOMC (hike 15%) | hold, no December signal | hike, or December past 90% |
| Oct 29 core PCE | ≤ 0.2% | ≥ 0.4% |
| Oct 30 book (72k BTC of 90k+ calls) | spot through 90k before expiry | rolled down, not out |
Why the odds moved. The rates break got its data and not its price: a 0.3% core and a 1.3% PPI took October’s hike odds to 54% intraday, BTC wicked 79.7k on $0.95B of liquidations, and both were bought. Then Warsh declined to signal, the 10Y fell 20bp, and the break moved from rates to credit, which is slower. Eleven sessions on the 50D with 10-day realized vol at 23% against DVOL 37 is compression into Oct 27–30, when HOOD, the FOMC, GDP, PCE and 149k BTC of expiry land inside four sessions.
FRED · BLS · Kraken · Deribit · OKX · CME FedWatch · Polymarket · Kalshi · CoinGlass · Farside via TFTC · 2026-10-20
Oct 15 was the year's fifth flush and its shallowest: $0.95B, 90% longs, a −2.5% wick to 79.7k that filled rung 1b and closed green
BTC daily close, USD · 2026 YTD · red = flush-day wick, low to close, with the day's liquidations · shaded = ladder rungs, re-cut on today's averages
how deep each flush went: wick low vs the close 3 days before, % · dashed = median · yellow = 2026 · grey = wick held 90d · red = wick revisited
- 1 08 OCT $1.1B liquidated, 94% longs · BTC wicked 80.3k and closed 81.7k, $230 above the 78.2–80.1k retest rung −1.9% /1d
- 2 15 OCT September PPI 1.3% m/m vs 0.6% · $0.95B liquidated, 90% longs; BTC wicked 79,709 and closed 82,134, green on the day +0.7% /1d
LADDER · 1 filled · 1b filled Oct 15 at 79.7k · core 72.6–78.0k (200D, max pain) · deep 62.2–67.2k (200W) · tail 58k · 25% of the budget in, no bid chases 84k
funding over 0.03% with open interest past 33k BTC: the fuel a flush needs, and every rung sits one lower
Fuel for a flush: unchanged since Oct 13, and it was not what fired.
| Sep 30 | Oct 13 | Oct 19 | |
|---|---|---|---|
| OKX BTC open interest, all contracts | $3.06B | $3.29B | $3.33B |
| USDT perp OI, BTC | 27.5k | 29.5k | 30.0k |
| funding, per 8h | +0.005% | +0.002% | +0.007% |
| DVOL · 10d realized | 35.2 · — | 37.8 · — | 37.4 · 23 |
| spot vs 20D · 50D · 200D | +4% · +9% · +18% | −3% · +0% · +13% | −2% · +1% · +13% |
| largest liquidation day | Sep 28, $486M | Oct 8, $1.1B | Oct 15, $0.95B, 90% longs |
Oct 15 was a data flush, not a positioning one: open interest was flat into it, funding under 0.01%, and the trigger was a 1.3% PPI the morning after a 0.3% core CPI. It went −2.5% from the Oct 12 close, the shallowest bar on the chart, and closed +0.7% on the day, the first 2026 flush bought inside its own session. Rung 1b filled at 79.7k; the core stays on the 200D and the 78k max pain. The book was not cleared, so a third flush before Oct 30 needs a trigger rather than a build, and the calendar carries four.
Kraken · Deribit · OKX · CoinGlass totals as reported · 2026-10-20
Rung 1b filled at 79.7k on Oct 15; a quarter of the budget is in, the core waits at 72.6–78.0k on the 200D
BTC, USD · levels from Kraken daily and weekly closes at 2026-10-19 · size = split of the bid budget
| RUNG | LEVEL | FROM SPOT | WHAT SITS THERE | SIZE |
|---|---|---|---|---|
| 1 · light · filled | 76.0–76.6k | in · +7% | filled Sep-15, held Sep-16/17, broke up Sep-18 | 15% · in |
| 1b · retest · filled | 78.2–80.1k | in · +3% | filled Oct-15 at 79,709 on the PPI print, closed 82.1k the same session · Sep-14 close 78.2k still untested | 10% · in |
| 2 · core | 72.6–78.0k | -5% to -11% | 200D 72.6k · Oct-30 max pain 78k · the 50D (81.5k) is at spot, no longer a bid | 35% |
| 3 · deep | 62.2–67.2k | -18% to -24% | 200W 67.2k · Aug lows 62.2–62.5k: every 2026 flush held the 200W | 30% |
| 4 · tail | 57.7–58.1k | -29% | Jul-1 low, the 52w low: a fill is a regime change, not a dip | 10% |
- 1 15 OCT September PPI 1.3% m/m vs 0.6% · $0.95B liquidated, 90% longs; BTC wicked 79,709 and closed 82,134, green on the day +0.7% /1d
Sizes split a bid budget, not a case for being fully invested. Rungs are re-cut when the 200D/200W move, not when spot does. A fill of rung 4 means the regime read is wrong: pull the ladder.
Kraken · OKX · Deribit · 2026-10-20
Week three: the chain's fees fell 11% against Solana's 20%, and the rails bounced (ARB +10%, UNI +5%) while the venue leg sold
daily chain fees, $M · DefiLlama · Jul 1 mainnet → Oct 18 · anchor: Solana · dotted = Sep 29, the 90-day gas subsidy expires
30d, indexed to 100 at Sep 19 · the rails (ARB, UNI dashed, LIT, ENA) vs the venue taking the flow back (SOL, PUMP) and the gas token (ETH, dashed) · OKX daily closes
- 1 29 SEPT Robinhood Chain gas subsidy ends · replaced the same day for Wallet swaps to Dec 31 · volume held $1.6B +2.4% /2d
- 2 27 OCT HOOD Q3 · Oct 27 after the close, the first quarter with chain revenue watch
STILL TRIPPED, SLOWER · fees $17.9M, the lowest week since mid-August, share of Solana 22% from 20% · DEX volume +8% w/w · Q3 lands Oct 27, not Nov 4
a week back over $1.5B a day of DEX volume with no new subsidy; or Robinhood names chain revenue on the Oct 27 call
The chain, week to Oct 18, against the week before.
| 7d | w/w | share of Solana | |
|---|---|---|---|
| Robinhood Chain fees | $17.9M | −11% | 22%, from 20% |
| its DEX volume | $6.9B | +8% | 47%, from 47% |
| fees, Oct 17–19, per day | $1.7–1.9M | the lowest since Aug 18 | |
| Solana fees | $80M | −20% | |
| Solana DEX volume | $14.8B | +9% |
The price, 7d · since Sep 30.
| rails | venue and gas | ||
|---|---|---|---|
| ARB | +10% · −3% | SOL | −1% · −11% |
| UNI | +5% · −18% | RAY | −10% · −1% |
| ENA | −8% · −28% | PUMP | −10% · −24% |
| LIT | −9% · −22% | ETH | −2% · −11% |
The fade slowed because Solana’s did not: the venue’s fees fell twice as fast in a week when the Solana complex (RAY, PUMP, JUP, each −10%) was the board’s worst, so the chain’s share rose while its fees made a new low. Not a recovery. The rails’ bounce is a positioning week in names down 20–30% since Sep 30, on no chain news beyond Tenev’s Oct 13 post calling the chain a top priority. The scored question moves to Oct 27: the Q3 print is eight days earlier than the board had it, and the first with chain revenue in it.
DefiLlama · OKX · Kraken · Robinhood IR · Decrypt · crypto.news · 2026-10-20
BTC flat and breadth 56% → 45%, through 32% on Sunday: the rails led (ARB, UNI, CRV), the Solana venue lagged (RAY, PUMP, JUP)
7d return by name, % · the board's three metas, the displaced venue and the majors · label: 7d · 30d · OKX daily closes to Oct 19; XMR, CAKE from CoinGecko
chains by fees · 7d to Oct 18, $M · label: total and change on the prior 7d · DefiLlama
ROTATION OFF · BTC +0.2% w/w and over half of alts under their 50D · the median name −3% · what led: the names most sold since Sep 30, and XMR; what lagged: September's winners
two green BTC weeks with breadth still under 50%: alts are being sold, not beta; breadth back over 70% on ETF inflows: the rotation resumes
Mean-reversion list, scored.
| candidate | read | this week |
|---|---|---|
| Robinhood Chain’s fee share to Solana | high | wrong this week: 20% → 22% because Solana’s fees fell 20%; both at October lows |
| ZEC against XMR | momentum | wrong again: −2% against +12%; XMR was the board’s best name |
| SOL against ETH | upward | flat: −1% against −2%, both −11% since Sep 30 |
| RAY, the venue leg | with its fees | wrong: −10% on Solana fees −20%; the venue led down |
| UNI | stretch | right: +5% w/w, still −18% since Sep 30, half the stretch left |
| ARB, new | stretch | +10% w/w on no news, −3% since Sep 30: closed if it fills the gap |
| ENA | closed | −8% w/w, −28% since Sep 30: the right close |
Breadth fell 11 points on a flat BTC week, 24 at Sunday’s low: alts sold on their own, which is not beta, and the first week since September the board’s read of a beta-down tape is wrong. The leaders were the names most sold since Sep 30 and the laggards September’s winners, the shape of a positioning unwind rather than a rotation. The sixteen-category chart stays off: CoinGecko throttled the sweep again.
OKX · Kraken · CoinGecko · DefiLlama · 2026-10-20
Confluence
ratios, weighted by where liquidity lives 4TOTAL3/BTC 12% off the 12m low — rotation into alts is tentative at best
TOTAL3 / BTC market cap · 90d · anchor: 12m low
CoinGecko · 2026-10-19
NDX/GOLD flat over 90d — risk appetite hasn't beaten safety yet
90d, indexed to 1.00 · gold = PAXG · rising = risk-on
FRED · Kraken · 2026-10-19
ETH/BTC +0% over 90d — ETH leading the majors
90d
Kraken · 2026-10-19
SOL/ETH +8% over 90d — noise until liquidity reaches alts
90d · weight LOW until liquidity reaches alts
Kraken · 2026-10-19
Flows & positioning
who's bidding, who's crowded 4USDT+USDC −$1.7B in 4 weeks — supply is shrinking, no impulse to score
weekly net issuance, $B · 12w to 2026-10-19
DefiLlama · 2026-10-19
October: green 9 of the last 11 years, median +14%
BTC monthly return, October 2015–2025 · rolls to next month in its last week
Coinbase · 2026-10-19
Breadth 45% — under half of alts above trend, not alt season
% of top-84 alts above their 50D MA · 90d · anchor: 50%
CoinGecko · 2026-10-19
Funding flat across the board — BTC highest at 0.010%, nobody is crowded
8h funding, OKX USDT perps · anchor: >0.03% = crowded
OKX · 2026-10-19
Theses
claims with receipts 1Stablecoin supply leads TOTAL3 by 4–8 weeks
CLAIM
Net mints front-run alt liquidity; supply is the tell, price the echo.
FALSIFIER
Two mint impulses >$4B with no TOTAL3 response inside 8 weeks.
MONITOR
▲ live · weekly issuance vs TOTAL3, weekly, hermes-ledger
STATUS
2 confirm · 0 misses · 0 pending — Aug impulse answered inside 2 weeks; lag shorter than claimed
Arb & pegs
spreads net of costs 1Kimchi premium +0.9%, inside its 90d range — no spread clears costs
spreads net of fees, funding and withdrawal time
| PAIR | SPREAD | NET | ANCHOR |
|---|---|---|---|
| kimchi · BTC upbit/okx | +0.9% | -0.0% | < 90d p85 (+1.1%) |
| coinbase premium · BTC vs okx | -0.12% | — | 44d US offered streak |
| USDe peg | 0.9998 | — | quiet |
| USDT peg | 0.9993 | — | quiet |
Net assumes a 0.9pt round trip on the KRW leg; the Coinbase premium is quoted gross.
Upbit · OKX · Coinbase · Kraken · FRED · 2026-10-19