Do altcoins follow Bitcoin? Their beta on its up days and down days, and whether the wildest pay, tested
Over the year to Oct 14, the typical one of our 46 most traded altcoins moved 1.27 times Bitcoin's daily move (its beta), with a correlation of 0.61: on Bitcoin's up days it rose 1.41 times as far, on its down days it fell 1.45 times as far. In 5 of the 6 full years since 2020 the typical coin fell further with Bitcoin than it rose with it, and it beat Bitcoin over the year in 1 of them; its beta ran from 1.08 (2020) to 1.79 (2025). Part of falling further can be drift: a coin that loses ground to Bitcoin every day falls more on its down days and rises less on its up days for that alone. With the drift taken out, the typical coin's beta was higher on Bitcoin's down days than on its up days in 6 of the 6 years. Each quarter, the half of the 50 that had moved most with Bitcoin over the year before did +0.6% against the other half on average, over 28 quarters: +4.1% in the 14 Bitcoin rose, −3.0% in the 14 it fell. No test is clear of luck (a luck test of 3 with both halves agreeing). Past patterns, not financial advice.
as of · our daily closes, Jan 1, 2020 to Oct 14, 2026
"Altcoins are just leveraged Bitcoin": when Bitcoin moves 1%, the saying goes, the alts move 2 or 3. A coin's beta to Bitcoin measures it - how far the coin moved, on average, for each 1% Bitcoin moved, day by day - and its correlation says how tightly (1: every day in step; 0: no link). Beside them, the two halves of the saying: on Bitcoin's up days, how many times Bitcoin's rise did the coin make, and on its down days, how many times its fall? Here for our 50 most traded altcoins (by dollar volume over the 30 days before; stablecoins, wrapped coins and Bitcoin itself left out), each with 180 or more daily moves beside Bitcoin's over the 365 days.
Over the 365 days to Oct 14, 2026, the typical one of 46 coins had a beta of 1.27 and a correlation of 0.61: on Bitcoin's up days it rose 1.41 times as far, on its down days it fell 1.45 times as far; 63% of them fell further with Bitcoin than they rose with it. Bitcoin moved −28% over the year. With its drift against Bitcoin taken out, the typical coin's beta was 1.29 on the down days and 1.21 on the up days.
The ten highest betas first, the five lowest last. Up days: the coin's average move on Bitcoin's up days over Bitcoin's average there; down days the same on its down days. Every coin's page carries its own line.
§ 02 · Year by year
Year by year
as of · each calendar year's 50 most traded altcoins, from its first close
on Bitcoin's up dayson its down daysthe line: 1×, Bitcoin's ownunder each year: its coins
Year
Coins
Beta
Correlation
Up days
Down days
Down − up beta
Beat Bitcoin
Bitcoin
2026 so far
50
1.18
0.60
1.26×
1.31×
+0.09
38%
−8%
2025
49
1.79
0.67
1.71×
2.11×
+0.31
6%
−7%
2024
50
1.15
0.59
1.16×
1.51×
+0.35
12%
+112%
2023
50
1.12
0.60
1.17×
1.47×
+0.74
18%
+154%
2022
49
1.26
0.72
1.33×
1.41×
+0.19
14%
−65%
2021
50
1.08
0.60
1.08×
0.99×
+0.96
66%
+58%
2020
50
1.08
0.68
0.96×
1.17×
+0.73
16%
+302%
Each year's figures are its typical coin's (the middle one); up days and down days: how far it moved for each 1 of Bitcoin's move on those days, drift and all; down − up beta: its beta on Bitcoin's down days less its beta on the up days, with its drift taken out; beat Bitcoin: the share whose move over the year, from its first close to its last (a coin that stopped trading at its last close), beat Bitcoin's. In 5 of the 6 full years the typical coin fell further with Bitcoin than it rose with it, and it beat Bitcoin over the year in 1 of them.
How far a coin went on the up days and the down days carries its drift: a coin that loses ground to Bitcoin day after day rises less than its beta on the up days and falls more on the down days for that alone, and one that gains ground the other way round. So the test takes the drift out - each coin's beta on Bitcoin's up days and on its down days, fitted together with one intercept, so the ground it loses or gains every day sits there and in neither beta. Down-day beta less up-day beta, a year at a time: +0.55 on average over 6 full years (0.63 in the older half, 0.47 in the newer). A luck test of 2.7, not past the bar of 3.0.
§ 03 · Does it hold?
Does a coin's beta hold?
as of · the coins in two years in a row
If last year's wild coins are this year's, beta is something a trader can plan with. Each two years in a row, the coins in both years' lists, ranked by their beta in each: a rank correlation of 1 keeps every coin in its place, 0 is a fresh draw.
2025 → 2026 0.85 over 26 coins; 77% of the upper half stayed in it
2024 → 2025 0.55 over 22 coins; 64% of the upper half stayed in it
2023 → 2024 0.56 over 24 coins; 83% of the upper half stayed in it
2022 → 2023 0.69 over 29 coins; 79% of the upper half stayed in it
2021 → 2022 0.16 over 23 coins; 55% of the upper half stayed in it
2020 → 2021 0.37 over 24 coins; 67% of the upper half stayed in it
The quarter test below says the same from the other side: the high-beta half of a quarter had a typical beta of 1.32 over the year before it and 1.30 over the quarter it was held; the low-beta half, 1.06 and 1.04.
§ 04 · High against low
High beta against low
as of · each calendar quarter, its 50 most traded altcoins split by beta
Do the coins that swing hardest with Bitcoin pay for it? At each quarter's first close, that day's 50 most traded altcoins with 180 daily moves in the year before, ranked by their beta over that year and split in two; each half held the quarter in equal stakes, the high half against the low half, and each against Bitcoin. A coin that stopped trading counts what it lost.
the high-beta half against the low-beta halfBitcoin's own quarter
Test
Cases
Average
Middle
Ahead
Luck test
High-beta half against the low-beta half
28
+0.6%
−1.1%
50%
0.1
High-beta half against Bitcoin
28
−2.0%
−3.6%
39%
−0.3
Low-beta half against Bitcoin
28
−2.3%
−7.4%
36%
−0.4
High against low, the quarters Bitcoin rose
14
+4.1%
+1.6%
50%
0.6
High against low, the quarters Bitcoin fell
14
−3.0%
−1.1%
50%
−0.6
Down-day beta less up-day beta, drift out, each full year
6
0.55
0.54
100%
2.7
Over 28 quarters the high-beta half did +0.6% a quarter against the low-beta half on average (−1.1% in the middle quarter), ahead in 50% of them. A luck test of 0.1, not past the bar of 3.0. The split is the leverage at work: in the 14 quarters Bitcoin rose, the high half beat the low by +4.1% on average; in the 14 it fell, −3.0%. Those two lean on how Bitcoin's quarter went, which nobody knew at its first close.
§ 05 · How it is tested
How it is tested
as of · methodology
A luck test (t) says how far a result sits from zero, in units of its own noise: about 2 is rarely luck for one test, but many tests at once need more, which is what the first cut is for.
Every number comes from our daily closes, read through the same screen as every lab here (bad prints, frozen closes, pegged and wrapped coins out). A coin's moves are its log moves from one close to the next, paired with Bitcoin's on the days both moved; its beta is the slope of its moves on Bitcoin's and its correlation theirs. A coin needs 180 paired moves in the year to count, so a coin that died in its first months leaves the betas but not the moves against Bitcoin: those count every coin of the year's list, one that stopped trading at its last close.
The design was fixed before any number was read: the 50 most traded at each year's and each quarter's first close, a year of moves for each beta, the split at the middle, a quarter's hold. Each quarter is one case and no two share a day, so each test is a plain one against zero over its cases, read on Student's t since the cases are few, and "fell harder" takes each full year's typical coin as one case. One change after the first read of our data, on Oct 14, 2026: "fell harder" had compared how far coins went on Bitcoin's down days and its up days, drift and all; it now takes the drift out. With 6 tests, a result needs a luck test of 3.0 or more and both halves of its cases, the older and the newer, on its side. Moves are prices only, before trading costs.
Read each half against Bitcoin with care: equal stakes earn the coins' average move, which beats the typical coin's whenever they scatter, and altcoins scatter far more than Bitcoin moves - so a half can lead Bitcoin while most of its coins lag it. The two halves against each other are built alike, and that is the fair test of beta.
§ 06 · Revisions1 dated
Revisions
as of · every change to its headline numbers, dated
This study is re-run every night on the whole history. It was first written on Oct 15, 2026; its numbers last moved on Oct 15, 2026. Each night its headline numbers moved is listed here, newest first.
The high-beta half against the low-beta half, a quarter: +1% · the newest full year's typical beta 1.8