“Bitcoin falls as Democrats win Congress” is a sentence that writes itself, and it was in many headlines this morning. It is also a causal claim, and a price chart on its own cannot support one. This is Field Guide #75: five checks that turn a plausible-sounding headline into something you can test, applied to the real overnight move. Bitcoin fell 3.05% from $77,981.38 to $75,600.64 on Coinbase between the 21:00 and 03:00 UTC hours, then recovered to $77,801.29 by 07:05 UTC on 4 November 2026 (full report). Did the election do that? The honest answer, after the checks, is “partly, perhaps, and we cannot separate it”.
Why this matters
Bitcoin trades around the clock, so every piece of news lands somewhere in a price series, and some news is always landing. A writer who knows the outcome can pick the window that best fits the story: the low at 03:00 UTC for a “crash” headline, the recovery by 05:00 for a “rebound” headline. Both are accurate and both are selective. The five checks below are designed to stop that selection. They build on our general method for attributing a price move to news, and on Field Guide #74 on “lost $X in market value” headlines, #73 on attribution claims and the prediction-market price guide.
Check 1: Line up the clock
Write down when the information arrived, then when the price moved. If the price started moving before the news, the news is not the whole story. Here the Coinbase hourly candles show a high of $77,981.38 in the 21:00 UTC hour; Bitcoin.com’s own feed puts the start of the slide at about 21:45 UTC. The 22:00 hour held a low of $77,613.75 and the 23:00 hour a low of $77,299.80, at around the time the first US polls close. The standard schedule has most eastern-state polls closing at 7 p.m. Eastern, which is 00:00 UTC. By the 23:00 UTC low, about 29% of the eventual peak-to-trough decline had already happened ($681.58 of $2,380.74). The rest accelerated through the 00:00 to 03:00 hours, as early projections arrived.
Score: partly met. The fall continued after the first projections, which is consistent with a result effect, but it began before them, which means other drivers were already at work.
Check 2: What was already priced in?
A result that the market expected cannot move prices much on arrival; the news was priced when expectations formed. Prediction markets give a rough measure. Before the vote Kalshi and Polymarket gave Democrats more than a 90% chance of the House and more than 70% of the Senate (StockTitan, 3 November); overnight, Kalshi’s Senate price rose to 99% (Bitcoin.com). Nicolas Quatravaux of Paradigm told CoinDesk that Polymarket had priced a Democratic win above 75% “for a while”, so it was not a surprise. By that arithmetic the House result added under 10 percentage points of information and the Senate result about 29 points. The surprise, if there was one, was a Democratic Senate, not a Democratic House.
Score: partly met. The House outcome was largely expected. The Senate shifted from likely to near-certain, which is real information, but not a shock.
Check 3: Is the size unusual?
A 3% move sounds large until you compare it with bitcoin’s own typical range. We pulled Coinbase hourly candles from 6 July to 4 November 2026 (2,894 hours) and measured the largest peak-to-trough decline inside every rolling 10-hour window. The median window had a decline of 1.22%; a 2.97% decline sits at the 5th percentile and 4.10% at the 1st. The overnight window’s 3.05% decline was at least as large as in 4.1% of windows. The UTC-day range so far (3.26%) compares with a median daily range of 2.44% and a 90th percentile of 4.51% across 120 full days.
Score: met, with a caveat. The overnight move was unusual in this window, though one in twenty-five 10-hour windows is that large without an election. On 29 October, the day of the Binance hack, the daily range was 5.39%.
Check 4: What else was moving?
Look at the rest of the board. S&P 500 futures were down 1.1% overnight, Nasdaq 100 futures were down about 2% at the worst point and Dow futures were down 0.7% (Bitcoin.com); CoinDesk later had Nasdaq futures back to a 0.3% loss. So bitcoin’s 3.05% was in the same direction and a bit more than a broad equity sell-off, which is its usual relationship to risk assets and not evidence of a crypto-specific reaction. Meanwhile three other drivers were live: spot bitcoin ETFs had lost about $514 million over the first two trading days of November (SoSoValue data in crypto.news), the $6.3 billion Binance hack was still working through the market (latest update), and the AI-chip rebound and AMD earnings were being digested. None of these is a controlled comparison; the point is that the election was one of at least four candidates.
Score: not met. The election cannot be separated from the concurrent equity move, the ETF flow trend and the Binance aftermath.
Video: a pre-result live stream on bitcoin and the midterms, streamed on 3 November 2026; it is context for what the market was expecting, not evidence of what happened.
Check 5: Did it persist?
Events that really change expected cash flows or risk tend to leave a lasting mark; transient jitters tend to reverse. After the 03:00 UTC low of $75,600.64, bitcoin climbed back above the pre-slide high within two hours, touching $78,066.23 in the 05:00 UTC hour. Dick Lo of TDX Strategies told CoinDesk that the drop reflected uncertainty and that the outcome removed it. A full reversal in two hours is more consistent with a flush of leveraged positions than with a lasting reassessment, but one night cannot prove that. The proper test is the 7-day and 30-day return, which we will add to our study of past midterms on 10 November and 3 December. The 2018 midterm was followed by a 45.5% decline over 30 days, so a quick bounce is not a guarantee.
Score: pending. The initial reversal supports “transient”; the verdict needs days, not hours.
The scorecard
| Check | Question | Finding on 4 Nov 2026 | Status |
|---|---|---|---|
| 1. Clock | Did the price move after the news? | About 29% of the decline preceded the first projections; the rest followed | Partly met |
| 2. Priced in | How much was already expected? | House >90% beforehand; Senate ~70% to ~99% | Partly met |
| 3. Size | Is the move unusual for bitcoin? | 3.05% in 10 hours; larger than 95.9% of 10-hour windows since 6 July | Met |
| 4. Other drivers | What else was moving? | Nasdaq/S&P futures, ETF outflows, Binance, chip earnings | Not met |
| 5. Persistence | Did it last? | Fully recovered within about two hours of the low | Pending |
Reading the table: one check passes, two partly pass, one fails and one is pending. A fair headline would be “Bitcoin dipped about 3% overnight as results came in, then recovered”, not “Bitcoin fell on the Democratic sweep”. Take the difference as a matter of evidence rather than tone.
A reusable worksheet
- Write the timeline first. List the clock times of the news, the first move and the extremes, in UTC, before looking at a chart.
- Quote the prior odds. Use a prediction market or consensus forecast taken before the event, and note the source and date.
- Compare with a rolling baseline. Count how often a window of the same length had a move as large over the past three to four months.
- List at least three alternative drivers. Include equities, flows and any exchange or regulatory news from the same 24 hours.
- Set a review date. Pick the 7-day and 30-day marks now, and publish the result whether or not it supports the first story.
Common traps
- Post hoc ergo propter hoc: “it fell after the vote, so the vote caused it”.
- Picking the window to fit: a 03:00 low for a “crash” and a 05:00 high for a “rebound” are both accurate and both incomplete.
- Quotes as measurement: analysts’ explanations are interpretations; the analysts quoted this morning, such as Dick Lo, Nicolas Quatravaux and Martin Lee, did not run controlled tests.
- Small samples: three or four midterms cannot establish a pattern, as our 12-month study shows.
- Ignoring the 24/7 clock: results arrive when US stocks are closed, so crypto often absorbs the first reaction, and the equity open can look like a second wave.
Video: a pre-election discussion of what a Republican loss could mean for markets, published 2 November 2026.
Sources: Bitcoin.com News, CoinDesk, crypto.news, StockTitan. Related reading: Does bitcoin move on options expiry Friday? and Field Guide #71 on proof-of-reserves reports.
Method: bitcoin prices, highs, lows and closes are Coinbase Exchange candles (UTC) and the Coinbase spot ticker pulled by Bitcoin Mastery at about 07:05 UTC on Wednesday 4 November 2026; the 4 November candle is still open at that time, so the latest price is the 77,801.29 ticker and the last completed daily close is Tuesday 3 November (77,336.75). Long-history bitcoin figures are Yahoo Finance BTC-USD daily closes (UTC). Third-party figures are attributed to their source and date. Hourly statistics use Coinbase Exchange hourly candles from 6 July to 4 November 2026 (2,894 rolling 10-hour windows); the first-polls-close timing is the standard US schedule and was not independently confirmed for 3 November 2026. Prediction-market figures are as reported by the outlets named.
Disclaimer: This article is for informational purposes only and does not constitute investment advice. Bitcoin and other cryptocurrencies are volatile and you can lose some or all of the money you put in. Nothing here is a recommendation to buy, sell or hold bitcoin, any exchange-traded fund, any listed security or any other asset, and the technical levels, probabilities and scenarios discussed are descriptions of published data, not forecasts. Do your own research and consult a licensed financial advisor before making investment decisions.