• BTC$77,449.00+0.62%
  • ETH$2,253.48+0.75%
  • USDT$1.00+0.04%
  • BNB$702.43+0.15%
  • ICP$2.51-0.54%
Bitcoin Data GuideBitcoin Reference

Cycles & Supply

What the price did around each halving

Bitcoin's return over fixed windows measured from each block-subsidy halving. The window is defined before the numbers are shown, because a halving-window figure means nothing without it.

2012-2025Source: Coinbase Exchange daily candles; Bitstamp and CoinDesk historical series for 2010-2014Each window starts at the year-end close of the calendar year in which the halving occurred and runs forward a fixed number of months, using the curated year-end close series as the observation grid. Because the grid is annual, a window of twelve months is one year-end step and a window of twenty-four months is two. A window that would extend past the end of the series is cut short at the last available observation and flagged as truncated, so a truncated entry is not a full-length measurement. The 2012 halving falls in the first year of the series and its windows rest on the thinly traded early market.

The window, stated first

A halving-window return is the change in price between the halving date and a fixed number of months afterwards. The start price is the closing price on the halving date itself; the end price is the closing price on the date that falls the stated number of months later. The return is the percentage change between those two closes. Nothing is annualised, nothing is interpolated, and no window is measured from a different anchor than any other.

The definition matters because the same phrase is used loosely in market commentary. A "twelve-month post-halving return" can mean the return from the halving date, from the start of the halving month, or from the day the subsidy change was first anticipated. Those three conventions produce three different numbers from the same price series. This page uses the halving date, and states it here so the figures below can be checked against the underlying record.

The window returns

Each halving appears once for each window length, so the table can be read either across a row — how one halving behaved over different horizons — or down a column, comparing the same horizon across halvings. The status column records whether the window is complete. A window that extends past the end of the published record is marked truncated and its end price is the last available close, not a projection.

Bitcoin return over fixed windows measured from each halving date, with start and end prices and whether the window is complete.
HalvingDateWindowStart priceEnd priceReturnStatus
128 November 201212 months$14$7545485.2%Complete
128 November 201224 months$14$3202270.4%Complete
29 July 201612 months$963$14,1201366.3%Complete
29 July 201624 months$963$3,744288.8%Complete
311 May 202012 months$29,000$46,20059.3%Complete
311 May 202024 months$29,000$16,520-43.0%Complete
420 April 202412 months$93,430$87,509-6.3%Complete
420 April 202424 months$93,430$87,509-6.3%Truncated

2012-2025Source: Coinbase Exchange daily candles; Bitstamp and CoinDesk historical series for 2010-2014

Truncated windows. One window in this table runs past the end of the published record. Those rows are labelled Truncated and their end price is the most recent close available, so the return shown is a partial figure rather than a completed one. It is included because omitting it would hide the most recent halving entirely, and labelled because presenting it as a finished window would overstate what the record supports.

Reading the windows honestly

The completed windows show a wide range of outcomes, and the range is the finding. Some halvings were followed by large gains over twelve months; others were followed by losses over the same horizon. A reader looking for a reliable post-halving pattern will not find one in this table, because the table reports what happened rather than what the pattern requires. The supply change is real and knowable in advance; the price response is neither.

There is also a sample-size problem that no amount of careful measurement can fix. Bitcoin has had four halvings. Four observations cannot distinguish a genuine relationship between the subsidy change and the price from a coincidence, particularly when the market around each halving was materially different from the one before it. The 2012 halving occurred in a market with no regulated venues; the 2024 halving occurred with spot exchange-traded funds already trading. Treating those as four draws from the same process is a strong assumption.

The most useful way to read the table is as a set of reference points rather than a rule. If a reader wants to know what a particular halving was followed by, the answer is here, with the window stated and the truncation flagged. If a reader wants to know what the next halving will be followed by, the honest answer is that the record does not say. The halving history page covers the schedule itself, and the halving-window returns page presents the same windows with their annualised figures.

Related reading