How to Read the ETH/BTC Ratio

Most crypto prices are quoted in dollars. This one isn't. ETH/BTC is the price of one ether in bitcoin, and it's the pair people check when they want to know which of the two biggest coins is winning.

It's also the chart that comes with a promise. When ETH/BTC starts climbing, the story goes, money is leaving bitcoin for riskier coins, and "alt season" is next. So this week we measured it: Binance's daily ETHBTC close since 14 July 2017, Kraken's weekly close back to August 2015, and 17 other big coins priced in bitcoin, to see what the ratio said about them.

What the number measures

Take ether's price in dollars and divide it by bitcoin's. On 23 September 2026 ether closed at $2,684.71 and bitcoin at $84,397.60, so ETH/BTC was $2,684.71 ÷ $84,397.60 = 0.0318. One ether was worth 0.0318 bitcoin. Flip it round and one bitcoin was worth about 31.4 ether.

That's all it is. It doesn't care whether crypto went up or down. If both coins rise 50%, the ratio doesn't move. If both fall 50%, it doesn't move either. It only moves when one coin does better than the other.

You can trade it directly on most big exchanges, as the ETHBTC pair. Buying it means swapping bitcoin for ether. It's a direct bet that ether beats bitcoin, with no dollars involved.

Finding 1: it's a scoreboard, and ether has mostly lost

The ratio started near 0.00167 in October 2015, when ether was a year-old experiment. By June 2017 it closed a week at 0.1401. That's the peak of the whole record, when one ether bought a seventh of a bitcoin. Then it fell for two years, to 0.0164 in September 2019.

The second rally ran through 2020 and 2021 and topped at 0.0879 on 8 December 2021. It never got back there. By 21 April 2025 it was 0.0181, the lowest since January 2020. A summer bounce took it to 0.0421 in August 2025. Today it's 0.0318, 76% off the 2025 low and still 64% below the 2021 high.

The table is the scoreboard. ETH/BTC rose in 2016, 2017, 2020 and 2021, and fell in every other year. In 2021 ether gained 399% in dollars against bitcoin's 60%, and the ratio more than tripled. Since then it's fallen five years running, counting this one so far. The worst of them was 2024: bitcoin +121%, ether +46%. Ether still made money that year. It just made much less than bitcoin, and the ratio fell 34%.

That's the first thing the number teaches. A rising ether price doesn't mean ether is winning. In 2023 ether rose 91% and still lost a quarter of its value against bitcoin.

Finding 2: when bitcoin falls, ether usually falls more

The popular version is that ether is "high beta": it goes up more than bitcoin in a rally and down more in a sell-off. Half of that is true. In the 51 months bitcoin fell, ether did better in only 16. The ratio's median move in those months was −5.8%. When bitcoin fell more than 10%, ether held up better in just 7 of 23.

The other half isn't. In the 57 months bitcoin rose, ether did better in 27, which is a coin flip. Even in months bitcoin rose more than 10%, ether won 18 of 38. So the pattern isn't "more up, more down". It's "about the same up, more down". That's why the ratio has spent most of the decade falling.

It isn't a fluke of one period. From 2018 to 2021, ether beat bitcoin in 13 of 23 up months and 10 of 25 down months. From 2022 on, it was 13 of 31 and 6 of 25. The gap between the two kinds of month is there both times, and it's wider in the second.

Day to day, the two coins move almost together. Ether's daily moves have a correlation of 0.80 with bitcoin's, and for every 1% bitcoin moved, ether moved about 1.03% on average. That's closer to one-for-one than ether's reputation suggests. It's been above 1.2 in only a few years, and 2025 was the highest at 1.47.

Finding 3: it describes the rest of the market

Here's where ETH/BTC earns its reputation. We priced 17 other big coins in bitcoin (XRP, Litecoin, Cardano, BNB, Chainlink, Dogecoin, Solana and 10 more) and took the middle coin's move in each 30-day window since 2018. When ETH/BTC fell over a window, the middle coin fell 9.0% against bitcoin, and only 24% of the basket beat bitcoin. When ETH/BTC rose, the middle coin was roughly flat (−0.6%) and 47% of them beat bitcoin.

The two move together. The correlation between ETH/BTC's move and the basket's is 0.59 over 30 days and 0.82 over 90. So when you look at ETH/BTC, you're looking at a decent summary of how every other coin is doing against bitcoin. It's the biggest of them, and it's the one with a liquid pair everyone watches.

One more thing the basket shows: most coins lose to bitcoin most of the time. Over any 90 days, the median share of the basket beating bitcoin was 29%. Part of that is the basket, which is old coins that have faded (see the limits at the end). But it's the base rate you're betting against whenever you swap bitcoin for something smaller.

Finding 4: alt season needs it, but it doesn't make one

"Alt season" has a common definition: 75% of the big coins beat bitcoin over the last 90 days. By that measure on our basket, 298 of 3,129 days since 2018 were alt season, about one day in 11. The longest run was 97 days, from 29 March 2021 to 3 July 2021.

On 276 of those 298 days, ETH/BTC was up over the same 90 days. So the claim is right as far as it goes: alt season almost always comes with ether beating bitcoin. The 22 exceptions all fall in one stretch, from November 2024 to January 2025, when the basket rallied while ETH/BTC fell.

Now run it the other way. ETH/BTC was up over the previous 90 days on 1,274 days. Only 276 of them, 22%, were alt season. Even when it was up more than 25%, it was alt season on 206 of 429 days, fewer than half. A rising ETH/BTC is nearly always there when alt season happens. Most of the time it happens without one.

This month is a live example. ETH/BTC is up 21% over the last 90 days, and 64% of the basket's remaining 11 coins have beaten bitcoin. That's more than usual, and it's not alt season. The last one ran from 28 August 2025 to 9 October 2025, and it started with ETH/BTC up 65% over 90 days.

Finding 5: it doesn't forecast

Everything so far is about the same window. The question traders actually ask is whether a rise today tells you anything about next month. It doesn't. After a month ETH/BTC rose, it rose again in 23 of 46 cases, exactly half. The basket's middle coin fell 3.6% against bitcoin the next month. After ETH/BTC fell, the basket fell 5.5%. That's a small difference either way. The correlation between this month's ETH/BTC and next month's basket is −0.17. Over 90-day windows it's 0.00.

Trend rules don't rescue it. The usual one says to own ether when ETH/BTC is above its 200-day average and bitcoin when it's below. Since 30 January 2018, the next 90 days after a close above the average had a median of −5.4%. After a close below, −3.4%. Above the line was the worse place to buy.

Switching between the two coins on that rule 94 times would have left you with 88% of the bitcoin you started with, before any fees. That's 12% less than holding bitcoin and doing nothing. It did beat holding ether, which lost 70% against bitcoin over the same stretch. That's not much of a win.

How to read it yourself

• Read it as a score, not a price. The ratio only moves when one coin beats the other. Ether can rise in dollars and still lose ground, and it's done exactly that in most years.
• Check what bitcoin is doing first. A falling ETH/BTC in a falling market is the normal case: in 35 of 51 down months ether did worse. A falling ETH/BTC while bitcoin rises is the one that tells you something about ether.
• Use it to describe the market, not to time it. When ETH/BTC is rising, most coins are usually doing better against bitcoin than normal. That's useful to know about today. It's not a reason to expect more of the same next month.
• Don't wait for it to call alt season. A rising ratio was there during almost every alt season, and it was there on 998 days that weren't one.
• Watch the base rate. Most coins lost to bitcoin over most 90-day stretches in this sample. Swapping bitcoin for anything smaller is a bet against that, not a free option.
• Treat trend lines on it with care. The 200-day average didn't separate good 90-day stretches from bad ones, and trading on it lost to doing nothing.

Method, and what this cannot tell you

ETH/BTC is ether's price divided by bitcoin's. The daily series is Binance's ETHBTC spot pair, closing price for each complete UTC day from its listing on 14 July 2017; bitcoin and ether in dollars are Binance's BTCUSDT and ETHUSDT from 17 August 2017, so the year and month tables start there. Before Binance, the history is Kraken's ETH/XBT weekly close back to August 2015. A month is the last close of one calendar month to the last close of the next. The basket is 17 coins that were among the biggest on Binance in 2017-2019, each priced in bitcoin on its own BTC pair and counted from listing to delisting; a window's basket move is the median coin's, and "alt season" is 75% or more of the basket beating bitcoin over the trailing 90 days, the definition the most-quoted alt-season index uses. The 200-day test compares ETH/BTC's close with the average of the 200 closes before it, and looks 90 days ahead.

• One venue for the daily data. Binance's ETHBTC is one of the busiest markets for the pair, and its closes agree with Kraken's to a median of 0.05% a week, but a price on another exchange can differ on the day of a big move.
• The basket is old coins. It's the 17 biggest coins listed against bitcoin in 2017-2019, so it leaves out most of what launched later and is heavy with coins that have faded. That's why its typical coin lost to bitcoin most of the time. An index built from today's biggest coins has the opposite bias: it only includes coins that have already done well.
• The basket shrinks. Binance delisted six of its bitcoin pairs between 2024 and 2026, so it's 11 coins by September 2026. A coin's history stops at its last day rather than counting the delisting as a loss.
• No fees and no taxes in the switching test. Ninety-four switches between ether and bitcoin would have cost something at any exchange. The rule lost to holding bitcoin before those costs.
• Nine years is a few cycles. The sample holds two big ETH/BTC rallies (2017 and 2020-21) and one smaller one (2025). A finding that rests on how those went is a finding about three episodes.