On 3 March 2020 the Federal Reserve cut interest rates by half a percentage point, an emergency move between its scheduled meetings. A month later bitcoin was 25.43 percent lower. On 15 March, twelve days after the first cut, the Fed cut again, by a full percentage point this time, twice as deep. A month after that one, bitcoin was 31.52 percent higher.
Same committee, same fortnight, two cuts in the same direction, opposite results of almost the same size. Whatever story explains the first cut explains the second backwards. That is the trouble with reading a single Fed meeting: one meeting tells a reader whatever the following month happened to do.
So this article does not read meetings. It counts them: 91 scheduled FOMC decisions from 2015 to 2026, sorted into three kinds. The Fed raised rates (a hike), cut them, or left them alone (a hold). Three questions follow, in order. Does a decision day stand out at all. Does it matter which way the Fed moved. Is a hike different from a cut. The answers are no, yes, and nobody can tell: hikes and cuts never share a year.
The Test That Kills Most Findings
One tool does most of the work here, and it is worth a plain sentence. Bitcoin's fifteen-year price history contains a few enormous booms and busts. Almost anything that drifts slowly enough will line up with them and look like it predicts the price. So every finding on this website gets measured twice: once across the whole stretch, and once inside each calendar year on its own, with the yearly results averaged. The second measurement holds the boom or the bust constant. Most findings do not survive it.
Crude oil's apparent power over bitcoin turned out to be a calendar, not a signal. Turned on this site's own dashboard, the same tool took out one of its indicators.
The Fed makes an unusually clean subject for it. Meetings are scheduled years in advance, the decision lands at 2pm Eastern on the second day, and rates went up, down, or nowhere. Nothing here is estimated. Nothing is a matter of opinion.
How the Sample Was Built
Meeting dates come from the Fed's own calendar and its yearly archive. What was decided at each meeting comes from the Fed's published record of its rate changes, which lists the date each change took effect, one working day after the meeting that ordered it. A meeting with a change against it is a hike or a cut of that size. A meeting without one is a hold. Nothing in this sample was typed from memory or from somebody's summary.
That gives 62 holds, 20 hikes and 9 cuts.
Four events that a looser count would have swept in are left out, each for a stated reason. An October 2019 conference call and an August 2020 written vote changed no rates, so neither is a decision and neither counts as a hold. The two emergency cuts that open this article were real and large, but they were not scheduled meetings, and the whole method rests on the schedule. The March 2020 meeting was cancelled outright. One more, the meeting of 28 January 2015, falls on the only missing day in the price file, and is dropped rather than filled in. That leaves 90 meetings with usable returns.
Returns run from closing price to closing price, 1, 7 and 30 days after each meeting. Three windows, so that nothing rests on the exact minute of a 2pm announcement. Where this article says "the month after", it means that 30-day window.
The Meeting Itself Is a Non-Event
Treat every decision the same and there is nothing there. That is the right first answer, not a disappointing one.
In the month after a meeting, bitcoin returned 5.67 percent on average. In the month after every other day in the same window, it returned 6.40 percent. The difference is 0.73 percentage points in favour of ordinary days, and measuring year by year barely touches it: 0.68. At one week the two figures are 1.00 and 1.03, again in favour of ordinary days. At one day it tips the other way, 0.68 and 0.64 in favour of decision days, which set against bitcoin's daily swings is nothing at all.
The useful part is that both measurements agree. When the whole-period number and the year-by-year number land in the same place, the booms and busts are not doing the work, and the honest reading is that there is nothing there to explain. A market that had learned to treat Fed days as special occasions would not produce this table.
What the Fed Decided Does Matter
Lumping a three-quarter-point rise in with a meeting that changed nothing is a strange thing to do to a number whose entire content is the decision. Split the meetings by what was decided, and at one month the three groups pull apart.
After a hike, bitcoin lost 5.16 percent over the next month on average. After a cut, it lost 3.66 percent. After a hold, it gained 10.59 percent, against 6.40 percent for ordinary days. Measured against that ordinary-day baseline: hikes 11.56 percentage points worse, cuts 10.06 worse, holds 4.19 better.
Then comes the step that has ended most findings on this website. This time it does not. Measured inside each year and averaged, hikes come out 12.29 points worse, cuts 9.33 worse, holds 3.68 better. Every one lands close to where it started, and the hike figure gets slightly worse under the test rather than better. Oil, for comparison, went from a 19.5-point edge to 0.0. Here the test takes nothing away.
The cleanest version is to stop separating hikes from cuts and ask only whether the Fed moved at all. In the month after a meeting where it changed rates, bitcoin lost 4.69 percent. In the month after a meeting where it held, it gained 10.59 percent. A gap of 15.3 percentage points across the whole period, and 15.4 measured year by year.
A gap that size deserves a hostile question before it deserves a headline, and the hostile question is how much of it is one good year.
Enough to matter. The average is built from nine yearly figures, two of which rest on a single meeting. Six of the nine are negative, three positive. Their middle value is 9.65 against an average of 15.36, which is what it looks like when one or two years pull hard on an average. Remove 2017, a year in which bitcoin rose more than tenfold with three hikes inside it, and the gap falls to 8.99. Remove 2018 instead and it climbs to 18.93.
So the honest statement is a range and a direction, not a figure. Depending on how the average is taken, and on which single year is removed, the gap sits between roughly 9 and 19 percentage points, always in the same direction, with six of nine years agreeing on it. The one-day version is the same shape in miniature: 0.74 points, pointing the same way in eight of nine years, the most consistent result in the study. The one-week version is the weakest of the three, 0.02 points, which is to say nothing.
There is no p-value anywhere in this article, on purpose. With 29 moves spread across nine years, and month-long windows that overlap each other heavily, the usual statistical machinery would hand back a figure that sounds far more authoritative than the evidence behind it. These numbers describe what happened. They do not establish why.
The two cuts that open this article make the case for caution better than any caveat could. Twelve days apart, both of them cuts, one followed by a 25.43 percent fall and the other by a 31.52 percent rise. Whatever the average describes, it is not something to lean on at any single meeting.
Hikes and Cuts Never Share a Year
Which leaves the question everyone arrives at next, whether a hike is worse for bitcoin than a cut. This data cannot say. The reason is worth understanding, because more history would not fix it.
Sort the meetings by year and they fall into blocks. From 2015 through 2018, hikes and holds. In 2019, cuts and holds. In 2020 and 2021, nothing but holds. In 2022 and 2023, hikes again. In 2024 and 2025, cuts. In twelve years, not one calendar year contains both a hike and a cut.
That matters because of how the year-by-year test works. It compares like with like by staying inside a single year, where the boom or the bust is roughly constant. When two things never happen in the same year, there is no inside to compare them in. Any hike-versus-cut difference drawn from this data is in truth a comparison of 2022 with 2025, wearing the costume of a comparison between two policy decisions. The whole-period figures do show hikes and cuts 1.5 points apart at one month. That figure should be ignored.
The Fed does not alternate. It turns, and stays turned for years. Its decisions and the state of the economy are close to the same thing.
Which is also why hikes-and-cuts-together against holds was the only split worth reporting. Moves and holds do share years, nine of them, and that overlap is the only reason the comparison can be made at all. The sample is not short of meetings. It is short of years in which the Fed changed its mind.
What to Take Away
Three things, worth keeping apart.
The meeting is not the event. Fed decision days, as a category, are ordinary days for bitcoin, and both ways of measuring agree on that at all three windows.
What the Fed decided does line up with something, within limits. From 2015 to 2026, the month after a meeting where the Fed changed rates was worse for bitcoin than the month after a meeting where it held. The gap is somewhere between roughly 9 and 19 percentage points, and holding the years constant does not weaken it.
Those limits travel with it. Twenty-nine events. A direction that holds and a size that does not. Three of nine years pointing the other way. It is not a forecast, and it is not evidence that the decision caused anything. The Fed changes rates when it is reacting to something, inflation in 2022, a slowdown in 2019, and those stretches are hard on risky assets for reasons that have little to do with the announcement itself. The finding fits "the Fed moves during rough patches" as well as it fits "the announcement moves the price". Telling those two apart needs a record of what the market expected going into each meeting, sourced properly, and that is not in hand.
The comparison everyone wants cannot be built. Hikes and cuts never share a year, so there is no way to weigh one against the other here, at any sample size. More data will not help, because it is the Fed's own behaviour that creates the problem.
Which is why the title of this article is what it is. A reader handed tomorrow's Fed decision in advance still could not say where bitcoin goes next. The meeting day is ordinary. The one gap that survived describes nine years; it does not set a rule for the next month. And the comparison that would sharpen it cannot be built at all. What the Fed does and what bitcoin does are related. That is not the same as one predicting the other.
The meeting of 29 July 2026 falls one day after this sample ends and was still open when it closed, so it is left out rather than guessed at. It is also the next test, and its terms are already set: whether the Fed moves or holds, and what the following month does, can be checked against the numbers above without a single further choice. Every figure here came out of one pass over the daily price history, including the figures that argue against this article's own headline.