Bitcoin lost 3.2% over the week of US sessions from Monday 2 to Friday 6 November, falling from $78,232.49 at 00:00 UTC on Monday to $75,731.05 at midnight on Friday night, on Coinbase’s public candles. Most of that loss came while Wall Street was open. Our own split of the week shows bitcoin down 2.05% during US cash hours and 1.18% outside them, and a single session, Thursday’s, accounts for more than all of the session loss: the other four sessions together gained about 2.4%. US spot bitcoin ETFs lost $1,281.7 million over the five days, with outflows on four of them, according to Farside Investors.
Ether did not follow the same split. It lost 3.06% in the US sessions and gained 1.48% outside them, ending the week 1.63% lower. The question here is what the US week did to crypto and what could change that next week, answered from Coinbase minute candles, Farside’s fund-by-fund table, Yahoo Finance closes for stocks, yields and the dollar, and Deribit’s options book.
It does not retell Friday. The hour-by-hour story of the last session, including the 15:05 UTC drop to $74,722, is in our Friday analysis of why bitcoin did not join Wall Street’s bounce, and the quiet Friday night and Saturday morning are in this morning’s report on bitcoin’s flat weekend start. At 17:38 UTC on Saturday, the latest Coinbase one-minute candle closed at $75,756.84. Coindaro’s market snapshot at 17:30 UTC put total crypto capitalisation at $2.56 trillion, 3.03% lower over 24 hours, with bitcoin’s dominance at 59.3% and the Fear & Greed index at 25.
Bitcoin this week: two losing US sessions, and one of them did all the damage
The method first, because the result depends on it. We took Coinbase’s one-minute BTC-USD and ETH-USD candles and read the opening price of the minute starting at 00:00, 14:30 and 21:00 UTC each day, plus the next day’s 00:00 UTC. The US cash session runs from 14:30 to 21:00 UTC since American clocks went back on 1 November. The session return is the 21:00 price divided by the 14:30 price. Everything else in the UTC day (the European and Asian hours before the bell and the three hours after the close) is “outside”. Weekly figures compound the five daily factors. Saturday and Sunday are left out.
| Day (UTC) | 00:00 open | 14:30 | 21:00 | Session | Outside | Day | Nasdaq close |
|---|---|---|---|---|---|---|---|
| Mon 2 Nov | $78,232.49 | $77,648.38 | $76,639.75 | -1.30% | -0.70% | -1.99% | -1.07% |
| Tue 3 Nov | $76,674.12 | $77,012.32 | $77,785.96 | +1.00% | -0.14% | +0.86% | +0.83% |
| Wed 4 Nov | $77,336.75 | $78,207.54 | $80,044.02 | +2.35% | +1.61% | +4.00% | +1.44% |
| Thu 5 Nov | $80,428.30 | $79,409.55 | $75,985.05 | -4.31% | -1.29% | -5.54% | -5.52% |
| Fri 6 Nov | $75,969.06 | $75,516.06 | $75,763.84 | +0.33% | -0.64% | -0.31% | +1.09% |
| Week | -2.05% | -1.18% | -3.20% | -3.37% |
Read the table by column. Bitcoin’s US sessions were positive on three days of five, Tuesday, Wednesday and Friday. The two negative ones were larger. Monday’s session took 1.30%, and Thursday’s took 4.31% on its own, from $79,409.55 at the opening bell to $75,985.05 at the close. Without Thursday, the four other sessions add up to a small gain.
The last column is the Nasdaq Composite’s daily change, computed from Yahoo Finance closes. It is not the same window as bitcoin’s session (it runs close to close and includes the overnight gap in stocks), so it is a check of direction, not a like-for-like comparison. On direction the match is complete: on all five days bitcoin’s US session moved the same way as the Nasdaq. On Thursday the two daily moves were almost identical, -5.54% for bitcoin’s UTC day and -5.52% for the Nasdaq. On Friday the Nasdaq gained 1.09% while bitcoin’s session gained only 0.33%, the gap our Friday piece described.
Outside the US hours, bitcoin fell on four days of five. The exception was Wednesday, the day after the midterm vote, when the price rose 1.13% between midnight and the opening bell as results came in, then another 0.48% after the close. Thursday’s “outside” loss of 1.29% happened almost entirely before New York opened; after 21:00 UTC that day bitcoin barely moved, ending the UTC day at $75,969.06.
Ether split the other way
Ether’s figures, from the same candles and the same method, look different. ETH-USD went from $2,272.24 at 00:00 UTC on Monday to $2,235.20 at midnight on Friday, a weekly loss of 1.63%. Its US sessions lost 3.06% compounded and its outside hours gained 1.48%. By day, the sessions were -1.66% on Monday, +0.34% on Tuesday, +4.57% on Wednesday, -5.87% on Thursday and -0.19% on Friday. The outside hours were positive on four days and negative only on Thursday (-1.99%).
The difference with bitcoin sits in the outside hours of three days. On Monday, Tuesday and Friday ether rose outside the session (+0.68%, +0.64% and +0.22%) while bitcoin fell (-0.70%, -0.14% and -0.64%). Small moves, on one week of data: too thin to call a pattern.
Why five observations prove little
One week gives five sessions. A single bad day can flip the sign of the whole column, and Thursday did exactly that. The split also depends on our choice of hours: a reader who takes 13:30 UTC (the hour of US data releases such as Friday’s jobs report) as the start of the “US day” would get other figures. What the table does show, without stretching it, is that the US cash session was where the week’s largest single move happened, on Thursday, the day of the AI-led selloff in US stocks.
Stocks, chips, the dollar and yields: the week that crypto traded against
The weekly changes below are computed by us from daily closes published by Yahoo Finance, comparing the close of Friday 6 November with the close of Friday 30 October. We did not find a Reuters or AP weekly wrap to check them against, so they are our arithmetic, not a quoted figure.
| Market | Close 30 Oct | Close 6 Nov | Week | Thursday 5 Nov |
|---|---|---|---|---|
| Nasdaq Composite | 25,437.00 | 24,579.43 | -3.37% | -5.52% |
| S&P 500 | 7,580.07 | 7,376.86 | -2.68% | -4.36% |
| Dow Jones Industrial Average | 51,770.31 | 51,285.73 | -0.94% | -3.43% |
| PHLX Semiconductor index (SOX) | 10,688.80 | 9,390.25 | -12.15% | -12.25% |
| Nvidia | $204.64 | $186.64 | -8.80% | -9.26% |
| Cboe VIX | 19.42 | 25.64 | +6.22 points | +11.57 points |
| US 10-year yield | 4.93% | 4.86% | -0.07 point | -0.05 point |
| US dollar index (DXY) | 101.88 | 102.30 | +0.41% | +0.04% |
The chip index is the outlier. The SOX lost 12.15% over the week, nearly all of it on Thursday, when it closed at 9,011.14 after 10,268.65 the day before. Friday won back 4.21%. Bitcoin’s weekly loss of 3.2% sits close to the Nasdaq’s 3.37%, which is consistent with the daily direction match in the first table.
Two things did not behave like a classic risk-off week. The 10-year Treasury yield, on Yahoo’s ^TNX series, ended lower than it started (4.86% against 4.93%), after a 4.95% close on Tuesday. And the dollar index rose only 0.41% over the week: it had slipped to 101.55 by Thursday, then gained 0.74% on Friday, the day of the jobs report. So the pressure on bitcoin this week came through equities, and through technology stocks in particular, more than through rates or the dollar.
The macro news of the week is already in our earlier pieces. The midterm vote on Tuesday 3 November produced a Democratic sweep, and what it means for crypto bills is in our analysis of Elizabeth Warren’s likely Senate Banking chair and the CLARITY Act. The October jobs report on Friday showed 165,000 jobs and 4.1% unemployment, according to the Bureau of Labor Statistics. The Bank of England raised its rate to 4% on Thursday. The Fed held at 3.75-4.00% on 28 October, by a 10-2 vote with two members wanting a hike.
Those decisions feed the number that matters most for next week: the odds of a Fed hike on 9 December. After Friday’s jobs data, futures traders raised the probability of a quarter-point increase to about 80% from 75%, according to Reuters as reported by CryptoSlate at 00:40 UTC on Saturday. Readings differ by provider and by hour: StockTitan, citing LSEG data reported by Reuters, put the odds at about 65% early on Friday, before the jobs report. We did not find a CME FedWatch reading for Friday’s close that we could open, so we do not give one.
Bitcoin ETF outflows this week: $1.28 billion, and nearly all of it from IBIT
This morning’s report gave the weekly total; the question here is where it came from. Farside’s daily totals for US spot bitcoin ETFs were -$398.6 million on Monday, -$115.6 million on Tuesday, +$74.6 million on Wednesday, -$430.2 million on Thursday and -$411.9 million on Friday. All figures here are from Farside’s table, read through a browser at about 17:40 UTC on Saturday, in millions of dollars. SoSoValue publishes its own series with its own method, and the two should not be added together or mixed day by day; this section uses Farside only.
The fund-by-fund sums are ours, from Farside’s daily rows.
| Fund (issuer) | Mon | Tue | Wed | Thu | Fri | Week |
|---|---|---|---|---|---|---|
| IBIT (BlackRock) | -405.3 | -95.0 | -201.2 | -105.0 | -403.0 | -1,209.5 |
| FBTC (Fidelity) | +6.7 | +6.8 | +236.7 | -199.9 | +42.8 | +93.1 |
| GBTC (Grayscale) | 0 | 0 | 0 | -49.9 | -11.1 | -61.0 |
| BTC Mini Trust (Grayscale) | 0 | -52.6 | 0 | 0 | 0 | -52.6 |
| EZBC (Franklin) | 0 | 0 | 0 | -32.9 | 0 | -32.9 |
| HODL (VanEck) | 0 | 0 | 0 | -23.6 | 0 | -23.6 |
| ARKB (ARK 21Shares) | 0 | 0 | +19.9 | -18.9 | -20.1 | -19.1 |
| BITB (Bitwise) | 0 | +21.9 | +15.2 | 0 | -20.5 | +16.6 |
| BTCO (Invesco) | 0 | 0 | +4.0 | 0 | 0 | +4.0 |
| MSBT | 0 | +3.3 | 0 | 0 | 0 | +3.3 |
| Total | -398.6 | -115.6 | +74.6 | -430.2 | -411.9 | -1,281.7 |
BlackRock’s IBIT carried 94% of the week’s net outflow, and it was the only fund that lost money on every day. Its outflow run now covers six sessions, from Friday 30 October to Friday 6 November, for $1,443.2 million in total. The only large inflow of the week went to Fidelity’s FBTC, $236.7 million on Wednesday, the day after the vote, and FBTC gave most of it back on Thursday. On Wednesday IBIT still lost $201.2 million, so the one positive day for the group was not a day of broad buying.
At an average price of about $75,500, $1.28 billion is roughly 17,000 BTC. That is our rough conversion, not a figure published by the funds, which report holdings separately.
Three weeks of flows, three directions
Farside’s page starts on 19 October, which gives three full weeks to compare. The trend in bitcoin funds went from net buying to net selling, and the selling grew.
| Week (Farside, $m) | Bitcoin ETFs | Ether ETFs | Solana ETFs |
|---|---|---|---|
| 19-23 October | +301.2 | -220.7 | -17.6 |
| 26-30 October | -202.4 | -431.4 | -71.6 |
| 2-6 November | -1,281.7 | -365.2 | -25.0 |
Ether funds lost money in all three weeks, with BlackRock’s ETHA at -$345.9 million of this week’s -$365.2 million; ETHA had outflows on four of five days, with a $32.8 million inflow on Wednesday. Solana funds had no flows from Monday to Wednesday and a $0.8 million outflow from Grayscale’s GSOL on Thursday, then lost $24.2 million on Friday, split between Bitwise’s BSOL (-$14.4 million), Grayscale’s GSOL (-$5.4 million) and MSOL (-$4.4 million). Farside has no XRP ETF page, so we have no comparable XRP figure.
The week’s outflow also has to be set against the size of the funds. Farside counts $55,766 million of cumulative net inflows into US spot bitcoin ETFs since launch, after this week. The $1,281.7 million that left equals about 2.2% of what the group had gathered in net terms a week earlier, $57,047.7 million. The next independent reading comes from CoinShares’ weekly fund-flow report, usually published on Mondays, which covers exchange-traded products worldwide, Europe included.
Leverage and options: funding near zero, and a big expiry above the price
By Saturday, longs had stopped paying for leverage. On Binance’s BTCUSDT perpetual, the funding rate (what longs pay shorts every eight hours when the contract trades above the index) was 0.0100% at 00:00 UTC on Saturday, the usual baseline, then 0.0013% at 08:00 and 16:00 UTC, according to Binance’s public data read at about 17:40 UTC. Longs now pay almost nothing to hold their positions. Open interest on that one contract stood at about 96,200 BTC at 17:43 UTC.
The forced selling came earlier. Liquidations reached $1.05 billion over 24 hours into Friday, 88.2% of them long positions, according to CoinGlass data cited by Decrypt.
Deribit’s options book, read through its public API at 17:43 UTC on Saturday, holds about 340,100 BTC of bitcoin options open interest across all expiries. Three dates stand out.
| Deribit expiry | Calls (BTC) | Puts (BTC) | Put/call | Max pain (our computation) |
|---|---|---|---|---|
| Fri 13 November | 19,851 | 13,567 | 0.68 | $78,000 |
| Fri 27 November | 39,163 | 32,737 | 0.84 | $80,000 |
| Fri 25 December | 83,723 | 56,725 | 0.68 | $75,000 |
The 27 November monthly expiry is the one to watch: about 71,900 BTC, or roughly $5.4 billion at $75,755, with a higher share of puts than the other two. Max pain, the strike at which option buyers as a group would lose the most at expiry, is our own computation from Deribit’s open interest by strike, not a figure Deribit publishes. For 27 November it comes out at $80,000, well above the price on Saturday. Max pain is a description of where open interest sits, not a forecast; our max pain explainer covers its limits. Two figures we looked for and could not find in a current public source: the CME futures basis and the Coinbase premium. We do not estimate them.
Next week’s calendar: CPI with no yearly rate, then PPI and the Fed minutes
Tuesday 10 November is the date that can move the hike odds. The Bureau of Labor Statistics publishes October’s consumer price index at 8:30 a.m. Eastern time, 13:30 UTC, according to its release schedule. This release comes with a gap. BLS could not collect prices in October 2025 during the federal shutdown, and its shutdown FAQ says it “will not publish 12-month percent changes for October 2026 for indexes where October 2025 data are not available”, which covers most series. The monthly change becomes the headline. Any “year-on-year inflation” figure quoted on Tuesday will be someone’s estimate, not the BLS number.
For the monthly figure, RBC Economics expects +0.3% for both headline and core prices, and the Cleveland Fed’s nowcast stood at +0.24% headline and +0.21% core on 6 November, both as reported by StockTitan. The last print, for September, showed 3.7% over twelve months, up from 3.4%, according to BLS data compiled by Trading Economics; the market reaction on 14 October is in our report on the hot September CPI and the Fed hike bets.
| Date (UTC) | Event | Source |
|---|---|---|
| Mon 9 Nov | CoinShares weekly fund-flow report | CoinShares (usual Monday release) |
| Tue 10 Nov, 09:00 | Fed Governor Christopher Waller, economic outlook, London | Federal Reserve calendar |
| Tue 10 Nov, 13:30 | October CPI, monthly changes only for most series | BLS |
| Fri 13 Nov, 08:00 | Deribit weekly expiry, about 33,400 BTC | Deribit |
| Fri 13 Nov, 13:30 | October producer price index (PPI) | BLS |
| Wed 18 Nov, 19:00 | Minutes of the 27-28 October FOMC meeting | Federal Reserve calendar |
| Fri 27 Nov, 08:00 | Deribit monthly expiry, about 71,900 BTC | Deribit |
| Fri 4 Dec, 13:30 | November jobs report | BLS |
| Tue 8 – Wed 9 Dec | FOMC meeting, with projections | Federal Reserve |
The Federal Reserve’s November calendar lists several more speeches by Vice Chair for Supervision Michelle Bowman and Governor Michael Barr during the week; the minutes on 18 November are the next official text on the two votes for a hike in October.
What would change the picture, and who says so
The price levels below are not ours: each is a condition named by the source cited with it. The last three items are public figures we will check against, not forecasts. None of this is a recommendation to buy or sell.
- On price, Glassnode’s weekly note of 4 November calls the $73,000 to $76,000 band “the level to watch”, where the short-term holder cost basis ($75,700), the 200-day average and a cluster of long liquidations meet. It adds that a sustained break below $73,000 “would point to a longer consolidation”. The note is public.
- Decrypt’s chart analysis of 6 November puts first resistance at the 200-day exponential moving average near $76,497, then the 50-day near $79,756 and about $80,000. Below, it watches $74,780, then the $73,000 to $72,500 area of August’s breakout, then $70,000.
- CryptoQuant analyst Darkfost, quoted by CryptoSlate, places the more important threshold at $75,000, with a failure there opening the way toward $70,000, where many holders bought.
- On flows, the signal is simple to check on Farside each evening: a first inflow day at IBIT after six outflow sessions, or a second week of outflows above $1 billion.
- On macro, the CPI monthly change against the 0.24% to 0.3% range above, and whether the hike odds move away from the 80% area reported after Friday’s jobs data.
- On stocks, whether the Nasdaq gets back to its Wednesday close of 25,736.51 and the SOX to 10,268.65, the levels before Thursday’s rout, given how closely bitcoin’s sessions tracked them this week.
Bitcoin closed the week, at midnight on Friday, right inside Glassnode’s band and just above Darkfost’s $75,000. At 17:38 UTC on Saturday it was still there, at $75,756.84 on Coinbase. The next scheduled figure that can move the hike odds is Tuesday’s CPI at 13:30 UTC.
What we do not know
Several gaps remain after the week. No source we could open names a cause for Friday’s 15:05 UTC drop. Daily SoSoValue figures for Monday to Wednesday were not available to us, so we cannot cross-check those three days. There is no current public reading of the CME basis or the Coinbase premium that we could verify, and no XRP ETF flow series on Farside. The dates of any lame-duck votes in Congress on crypto bills are not set. And our session split covers one week only: it will take several weeks of the same computation before anyone can say whether bitcoin is really weaker during New York hours or whether this week simply had one very bad Thursday.