Coinseekly - AI driven crypto insights, details, patternscoinseekly
Data study

What happens after big crypto moves?

After a coin jumps 25% in a week, does it keep going or give it back? After it crashes, does it bounce? We measured the next 7 to 14 days after 19,265 such moves in 33 of the largest coins, January 2019 to October 2026, and compared each with an ordinary stretch.

What we found

  • We tested 15 definitions of a big move. 12 left the next week looking like any other week, 2 were followed by better results than usual and 1 by worse.
  • After rises, 0 of 6 definitions did better than usual and 0 did worse. After falls and pullbacks, 2 of 9 did better and 1 did worse.
  • The most favourable tendency was after a coin lost 25% or more over seven days: the typical next 7 days returned +3.9%, against -0.3% for an ordinary stretch (419 events).
  • The weakest was after a coin closed 20% or more below its highest close of the previous 30 days: -1.5% over the next 14 days, against +0.0% for an ordinary stretch (2,005 events).

“Better” and “worse” need a gap of at least 1.5 points in the typical (median) result and a gap of at least 5 points in how often the price ended higher, with the two pointing the same way. Anything smaller is called no real difference.

After a big one-day rise

Coins in the top 100 that closed a day higher by at least this much, then the next 7 days. Hover or tap a row for the full figures.

Size of moveEventsTypical resultHigher afterOrdinary stretchVerdict
5% or more3,799-0.5%48%-0.1% · 49%No real difference
10% or more1,466-1.1%46%-0.2% · 49%No real difference
15% or more622-1.1%47%-0.2% · 49%No real difference

After a big one-day fall

Coins that closed a day lower by at least this much, then the next 7 days. Hover or tap a row for the full figures.

Size of moveEventsTypical resultHigher afterOrdinary stretchVerdict
5% or more3,537-0.8%47%-0.1% · 49%No real difference
10% or more1,067-0.3%49%-0.3% · 49%No real difference
15% or more319+0.4%53%-0.2% · 49%No real difference

After a big weekly rise

Coins up at least this much over seven days, then the next 7 days. Hover or tap a row for the full figures.

Size of moveEventsTypical resultHigher afterOrdinary stretchVerdict
15% or more1,877-1.1%46%-0.2% · 49%No real difference
25% or more861-0.9%47%-0.2% · 49%No real difference
40% or more360+0.3%51%-0.3% · 49%No real difference

After a big weekly fall

Coins down at least this much over seven days, then the next 7 days. Hover or tap a row for the full figures.

Size of moveEventsTypical resultHigher afterOrdinary stretchVerdict
15% or more1,495+1.0%54%-0.3% · 48%No real difference
25% or more419+3.9%63%-0.3% · 48%Better than usual
40% or more57+8.5%63%-0.3% · 49%Better than usual

After a deep pullback

Coins that closed at least this far below their highest close of the previous 30 days, then the next 14 days. Hover or tap a row for the full figures.

Size of moveEventsTypical resultHigher afterOrdinary stretchVerdict
20% or more2,005-1.5%45%+0.0% · 50%Worse than usual
30% or more993-0.4%49%-0.2% · 49%No real difference
40% or more388-0.1%50%-0.3% · 49%No real difference

How the study works

For each of the largest coins we take every day of its daily closing-price history. A day counts as an event when the coin's move over the look-back period is at least the size shown. We then measure the close-to-close return over the next 7 days (14 for pullbacks). After an event, the same coin cannot start another until that window has ended, so one crash is not counted many times.

The comparison is every single day of the same coins, over the same years, with the same forward window: what an ordinary week looked like. The typical result is the median, because a handful of huge moves drag the average far from what usually happened. Prices are daily closes of each coin's USDT pair on Binance, the same series behind our calculators.

What this study cannot tell you

  • It only includes survivors. The coins are those in the top 100 today with at least a year of history. Coins that collapsed and dropped out are missing, which tends to flatter every result, especially after falls.
  • The events cluster in a few market phases. Big moves bunch up in bull runs and crashes, and the coins move together, so the events are far fewer independent observations than the totals suggest.
  • It is not a trading strategy. No fees, slippage or the difficulty of acting on a close are modelled, and a typical result says nothing about the spread around it.

Disclaimer: This study is educational and does not constitute financial advice. It describes what happened in the past; past performance does not predict future results. Cryptocurrency prices are volatile; never invest more than you can afford to lose.