Bitcoin and a 5% US 10-Year: The Record Says Coin Flip
Bitcoin closed Monday at $81,762.56 USD, down 2.29% from the prior close of $83,679.47 USD, and then went lower still. The UTC trading day that begins at 8 p.m. ET Monday took it to $81,197.99 USD, below Monday’s intraday low of $81,681.63 USD, before recovering to a high so far of $84,059.52 USD. The macro candidate for a move like that is the US 10-year Treasury yield, which has spent the last few weeks above 5%.
As of 6:05 a.m. ET on Tuesday, Bitcoin was quoted at $83,503 USD, up 2.13%, and $119,103 CAD, up 2.00%. Ethereum was at $2,511 USD, up 2.74%, and $3,581 CAD, up 2.59%. Live quotes, not closes, from Yahoo Finance.
Three things get in the way of that candidate. The US bond market was shut on Monday for Columbus Day, so the last 10-year print is Friday October 9 at 5.24% and the yield did not trade on the day Bitcoin fell. The level above 5% has no track record to appeal to. And testing the part of the yield that does have a record puts Bitcoin’s next 90 days at 8 up and 8 down, median -1.1%.
The 10-year above 5% has no precedent in Bitcoin’s lifetime
In the daily par yield series from January 2, 2009 to October 9, 2026, 4,446 closes in all, the US 10-year has finished at or above 5.00% on exactly 16 days. All 16 fall between September 15 and October 9, 2026, and the first was September 15 at 5.00%. Bitcoin’s genesis block dates to January 2009, so the series covers essentially the whole of Bitcoin’s existence.
The line was not broken from a distance. The record before this was 4.98% on October 19, 2023, and through September 2026 the yield crept up to it: 4.95% on September 10, 4.96% on September 11, 4.97% on September 14, then 5.00% on September 15. The 10-year started the year at 4.19% on January 2 and closed at 5.24% on October 9, up 105 basis points. Every yield figure here is read from the US Treasury’s daily Treasury par yield curve rates.
The US 10-year had never closed above 5% in Bitcoin’s lifetime. Daily Treasury par yield, January 2009 to October 9, 2026.
Why a government bond yield moves a non-yielding asset at all is the discount-rate mechanism, the same arithmetic that prices equities. Our guide to what moves a stock price works through it.
The level has no sample, so we tested the rate of change
Sixteen days of history, none of them with a completed 90-day outcome, is not a sample anyone can read a 90-day answer from. So we tested what does have one: the rate of change in the yield. We merged the Treasury 10-year daily close with Bitcoin’s daily close, which begins September 17, 2014, giving 3,017 overlapping days. For each day we took the change in the 10-year over the prior 60 business days in basis points, sorted every day into quartiles by it, and measured Bitcoin’s return over the next 30, 60 and 90 calendar days.
At the day level, rising yields look damning
| 60-day yield change | Observations behind the median | BTC next 90 days, median | Hit rate |
|---|---|---|---|
| Falling most (-131 to -22 bp) | 784 | +22.4% | 71% |
| Second quartile (-21 to +4 bp) | 750 | +17.1% | 68% |
| Third quartile (+5 to +27 bp) | 722 | +1.9% | 53% |
| Rising most (+28 to +157 bp) | 700 | -8.0% | 42% |
| All days | 2,956 | +8.1% | 59% |
Yields falling fastest, Bitcoin up a median 22.4% over the next quarter. Yields rising fastest, down 8.0%, with 42% of windows ending higher. A 30-point spread, in the direction the macro story would have it.
Counting each episode once turns it into a coin flip
Those 700 windows are not 700 independent observations. Yields rise in long runs, and every day inside a run generates its own 90-day window that overlaps almost entirely with the day before it, so one bad run enters the sample dozens of times. The top quartile’s 734 days include 87 from the single phase running October 2020 to May 2021, each carrying a version of the same outcome.
Group those days into distinct episodes instead, with a gap of more than 45 days starting a new one, and take one observation per episode on the day the 60-business-day rise peaked. That gives 17 episodes, 16 with a completed 90-day outcome. Their median is -1.1%, the mean -3.2% and the hit rate exactly 50%: eight up, eight down, worst -42.1%, best +48.8%.
Sixteen rising-yield episodes, and no consistent answer. Bitcoin’s 90-day return from the day each episode’s 60-business-day yield rise peaked. Episode median -1.1%; year-matched control, +9.1%.
Bitcoin’s all-days median is the wrong yardstick for that figure, because it is built from the same overlapping windows this section has just taken apart. So we drew a matched control: 16 anchor dates from days outside the top quartile, one per episode peak year, repeated 10,000 times, matching the episodes on sample size and era mix. The control’s median 90-day return is +9.1%, with a 90% band of -2.6% to +27.3%, and only 8% of those draws came in at or below -1.1%. The gap survives onto the right basis, which makes it suggestive rather than settled. Bootstrapping the 16 episode outcomes themselves puts a 90% band of -11.7% to +9.4% around that -1.1%, straddling both zero and the control: sixteen observations cannot settle this.
Removing the double-counting takes -8.0% to -1.1% and a 42% hit rate to a literal coin flip. What does not survive is the idea that a rising 10-year reliably sends Bitcoin down.
The effect was real in one era and modest in this one
On a single global quartile cut, those 700 rising-yield observations return -8.0% with a 42% hit rate. The 508 of them outside 2018-2021 return -0.6% with a 50% hit rate. The 192 inside it return -24.8% with a 20% hit rate, so four years carry the entire headline result.
Recomputing the top quartile inside each era, against that era’s own baseline, says the same thing from the other direction.
| Era | Rising-most median | Rest median | Gap |
|---|---|---|---|
| 2014-2017 | +25.3% (73% hit) | +27.4% (72% hit) | none to speak of |
| 2018-2021 | -23.8% (26% hit) | +14.5% (62% hit) | 38 points |
| 2022-2026 | -5.9% (45% hit) | +3.4% (56% hit) | 9 points |
The yield effect was absent in 2014-2017, enormous in 2018-2021 and modest in the era we are in: a gap of about 9 points, with the rising-yield bucket still rising 45% of the time.
What the day argues, in both directions
The bearish side has more than a story behind it. Only 8% of matched-control draws landed at or below the episode median, and three of the eight losing episodes ran deeper than 30%: -42.1% after the peak on October 5, 2018, -39.1% after March 31, 2021 and -32.9% after October 21, 2021. The current run is also not history yet. The 10-year was up 67 basis points over 60 business days on October 9, which puts today in the top quartile, and episode 17 began on March 26, 2026 with no outcome attached to it.
The other side is one row in the table, and it is the row that sits closest to now. The peak on October 19, 2023 came with the 10-year at 4.98%, the previous record high for the series and the nearest analogue by level we have. Bitcoin rose 48.8% over the following 90 days. One episode is not an effect and should not be treated as one. But it is the single most comparable yield reading in the sample, and ignoring it would be as selective as leaning on it.
A confound sits under both readings. The three deepest losses are the 2018 bear market and the two 2021 cycle tops, and the best covers the run-in to the spot ETF approvals, so cycle position is at least as visible in these 16 outcomes as the yield is. What would change the read is arithmetic: episode 17 acquiring a completed 90-day outcome, or the rise stalling until the 60-business-day change leaves the top quartile.
What a Canadian holder sees
The discount rate that actually prices a Canadian portfolio is not the one in the headline. The Government of Canada 10-year benchmark was 3.93% on October 8, 2026, which is 131 basis points below the US 10-year, and it has risen 46 basis points this year from 3.47% on January 2 against the American 105. Its 2026 high is 3.99%, on September 29. The Bank of Canada’s overnight rate is 2.25% with the next decision on October 28, and we covered what the futures market is pricing for it in our look at the rate path after the September jobs report.
Monday’s close was $116,763.56 CAD, and Tuesday morning’s $119,103 CAD is up 2.00% against 2.13% in USD. TSX-listed companies holding Bitcoin on the balance sheet carry this macro sensitivity with a corporate capital structure stacked on top of it, which is why we rank Canadian crypto stocks on Bitcoin per share rather than on price alone.
Data as of October 13, 2026. US yield figures run to the latest published close, October 9, 2026; the Canadian series runs to October 8.
Disclaimer: The content on bestcanadianstocks.ca is for informational and entertainment purposes only and does not constitute financial advice. Past performance is not indicative of future results. Always consult a qualified financial advisor before making investment decisions. Yield figures are from the US Treasury daily par yield curve rates and, for the Canada 10-year benchmark, the Bank of Canada’s Valet series, which runs one day behind. Bitcoin and Ethereum prices are from Yahoo Finance. The episode study and the matched control are our own arithmetic on those two series.
