Altcoin Season: What It Is and How to Measure It

What altcoin season means

Informally, altcoin season is any stretch where holding altcoins beat holding Bitcoin. That informal version is useless for anything except after-the-fact commentary, because it has no threshold, no sample and no time period.

The formal version that most traders actually quote comes from Blockchaincenter's Altcoin Season Index, and it is worth reading precisely rather than approximately.

Three parts of that definition do real work and each of them is a place the metric can mislead.

The threshold is 75%, not a majority. Altcoin season by this definition requires three quarters of the sample to beat Bitcoin, which is a demanding bar. A market where 60% of altcoins beat Bitcoin is not altcoin season under this rule, even though most altcoin holders would have outperformed.

The sample is the top 50, excluding stablecoins and asset-backed tokens. Excluding wrapped Bitcoin and staked Ether derivatives is methodologically correct, since those are claims on assets rather than independent altcoins. It also means this index avoids the double-counting problem that affects Bitcoin dominance, where such derivatives sit in the denominator.

The window is 90 days and it rolls. This is the property that determines almost everything about how the index behaves, and it gets its own section below.

The rolling window problem

A rolling-window indicator reports on a fixed stretch of the past that moves forward one day at a time. That gives it a specific and predictable behaviour which almost nobody accounts for when reading it.

Working it through

Consider a simple timeline of a rotation into altcoins.

Reading Figure 1

The green band is the rotation. The copper bar above it is the stretch during which the 90-day index reads "altcoin season". They describe the same event and they do not line up at either end.

It arrives three quarters of the way in. The 90-day line crosses the threshold on day 45 of a 60-day rotation. Not early, not halfway — three quarters through. Fifteen days of the move are left at the moment the indicator confirms it, and the confirmation is what most people are waiting for.

It stays wrong for longer than it was right. The rotation ends on day 60. The 90-day index keeps reading season until day 106, which is 46 days of a signal describing a market condition that has stopped. It does not reach zero until day 150, a full 90 days after the last day of outperformance, because that is when the final strong day rolls out of the window.

The two lags are the same fact. The window has to fill before the reading turns on and empty before it turns off. There is no setting that removes this; a shorter window shortens both lags together and a longer one lengthens both. It is the definition of a trailing average, not a defect in this particular index.

Which is what the second line is for. The 30-day index crosses up on day 15 and falls back below the threshold on day 76. Between day 76 and day 105 the two disagree: the 30-day says the rotation is over while the 90-day still says season. That disagreement is 30 days of warning, and it is available from two numbers published side by side on the same page. It is not a forecast — the rotation had already ended by then — but it is the earliest the data can tell you, and it beats waiting for the 90-day line to catch up.

Why this happens. The index is not measuring the present. It is measuring the trailing 90 days, and a trailing average stays elevated for exactly as long as the strong period remains inside the window. The lag at the start and the lag at the end are the same phenomenon: the window has to fill up before the reading turns, and it has to empty out before the reading turns back.

What this actually tells you. A reading of "it is altcoin season" is a statement that the last three months contained altcoin outperformance. It is not a statement about today, and it carries no information about tomorrow. A trader who enters altcoins on the day the index crosses the threshold is entering after roughly three quarters of the move that generated the reading — day 45 of the 60 in Figure 1 — and can then hold through 46 further days in which the index still reads season and the rotation has already stopped.

The actionable version. Read the 30-day version against the 90-day version. The 30-day window rolls off three times faster, so when the 30-day reading falls while the 90-day is still elevated, the rotation has already stopped and the 90-day figure is reporting history. That divergence is the earliest signal available from this data, and it costs nothing to check because both are published on the same page. The reverse configuration, a rising 30-day with a still-low 90-day, is the earliest evidence that a rotation has begun.

The sample selection problem

The second structural issue is subtler and affects the index in both directions.

The top 50 is not a fixed list. It is re-evaluated as market capitalisations change, which means membership is partly determined by recent performance. Two consequences follow.

Assets that collapsed leave the sample. A token that fell 90% drops out of the top 50 and is replaced by one that rose. The index measures the coins that are currently in the top 50, so the worst outcomes are progressively removed from the measurement. This is survivorship bias operating continuously rather than at a single cut-off.

Assets that surged enter the sample. A token that rallied enough to enter the top 50 brings its performance with it. Depending on how the calculation handles entrants, recent strong performers can be over-represented relative to a fixed cohort chosen at the start of the period.

The net effect is that the index describes the performance of a set selected partly by that same performance. That does not make it useless, but it does mean the reading is systematically friendlier to altcoins than an equivalent measurement over a fixed cohort would be. If you want to know how your altcoin positions did, the index is not the answer, because your positions did not get quietly replaced when they underperformed.

Why altcoin rotation happens at all

The rotation itself is real, and the mechanism is worth stating properly rather than as folklore.

The third row is the one that has changed most and is least reflected in older commentary. The traditional rotation model assumes capital enters crypto through Bitcoin and can then move onward. To the extent that capital arrives through regulated products holding Bitcoin specifically, that assumption weakens: those flows can raise Bitcoin's price without ever becoming a bid for altcoins. The mechanics are covered in ETF flow basics.

Why altcoin seasons end

They end for reasons that are all versions of the same thing: the marginal buyer stops arriving.

Risk appetite reverses. Any broad risk-off event, whether crypto-specific or macro, hits the highest-variance assets hardest. Altcoins fall faster than Bitcoin, which is the same asymmetry that made them attractive on the way up.

Supply catches up. Rallies attract issuance. New listings, unlocked allocations and new tokens compete for the same pool of speculative capital, so the money required to keep the average altcoin rising grows. The supply side of this is covered in vesting schedules and token unlocks.

Leverage unwinds. Speculative altcoin positioning is heavily leveraged, and a decline that triggers liquidations produces forced selling into thin books, which triggers more. This is the cascade mechanism described in how crypto liquidations happen.

Attention moves. Speculative demand is attention-backed rather than usage-backed, and attention has no floor. When it moves, the demand does not decay gradually.

What the index cannot tell you

• It cannot forecast. It is a trailing 90-day measurement, so it reports on a period that has already ended.
• It cannot time an exit. The same window that delayed the entry signal delays the exit signal by the same amount.
• It cannot describe your portfolio. The top 50 is re-selected continuously, so collapsed assets leave the sample. Your holdings do not.
• It cannot tell you the magnitude of outperformance. It counts how many coins beat Bitcoin, not by how much. A market where 76% of coins beat Bitcoin by 1% reads identically to one where they beat it by 100%.
• It cannot be compared cleanly across cycles. The composition of the top 50 changes completely between cycles, so an equal reading in two different years describes two different sets of assets.
• It says nothing about small caps. The sample is the top 50 by market capitalisation, which excludes the part of the market where the largest percentage moves usually occur.

How to read the index honestly

Conclusion

Altcoin season has a precise published definition, and using it precisely removes most of the confusion around the term. Blockchaincenter's rule is that 75% of the top 50 coins, excluding stablecoins and asset-backed tokens, must have outperformed Bitcoin over the trailing 90 days. That threshold is demanding, the exclusions are methodologically sound, and the exclusion of wrapped and staked derivatives avoids a double-counting problem that Bitcoin dominance does have.

The rolling window is what determines how the index behaves, and it is the part traders consistently ignore. A trailing 90-day measurement only turns positive after the window has filled with outperformance, and it stays positive until that outperformance has rolled out again. The result is symmetric lag: the index confirms a rotation roughly two thirds of the way through it and continues confirming it for up to three months after it has ended. No parameter change fixes this, because the lag is the definition.

The one genuinely useful move with this data is free and takes seconds. The 30-day and 365-day versions are published alongside the 90-day default, and the 30-day window empties three times faster. When the 30-day reading falls while the 90-day remains elevated, the rotation has already stopped and the headline number is describing history. That divergence is the earliest information the index can offer.

Two further limits are worth carrying. The top 50 is reselected continuously, so assets that collapsed are quietly replaced and the index always looks friendlier to altcoins than an unmanaged portfolio did. And the index counts how many coins beat Bitcoin rather than by how much, so a market of marginal outperformance and a market of spectacular outperformance produce the same reading. Used to label what regime the last quarter was, the index is genuinely informative. Used as a trigger, it is being asked to forecast from a window that only contains the past.

Frequently asked questions

Sources and further reading

Primary sources:

• Blockchaincenter — Altcoin Season Index

Related CoinBeaver articles:

• How to read crypto market signals
• Bitcoin dominance and the altcoin cycle
• The Crypto Fear and Greed Index
• ETF flow basics
• Vesting schedules and token unlocks
• How crypto liquidations happen
• What actually moves crypto prices

This article is educational and is not financial advice. The timeline used to illustrate the rolling window, including Figure 1, is a constructed example showing how a trailing measurement responds, not a record of any specific historical period. Figure 1 also models the index as the share of its lookback that contained outperformance, whereas the published index measures the share of top-50 coins that outperformed. Those are different quantities; the lag being drawn comes from the length of the window rather than from what is counted inside it, which is why the shape holds for both. No historical altcoin season durations or index values are quoted here, because durations derived from a 90-day trailing window measure the window as well as the market, and the publisher does not disclose the formula converting the underlying percentage into the displayed score. Check the index page for the current definition before relying on any threshold.