Crypto market cycles: how to identify the market phase

On August 28, 2026 bitcoin broke through 80 thousand and briefly tagged 81,240. Up 22% for the week. The Fear & Greed Index reads 68, though seven days ago it stood at 87. The word "bull run" is back in the news feed.

The other half of the picture looks different. This cycle's high was printed in October 2025 at 126 thousand, in early August 2026 bitcoin traded at 64,350, and the drawdown from the top reached 49%. And the rally everyone is discussing started with a liquidation cascade on August 19-20: short positions were wiped out to the tune of $2.77 billion against $264 million in longs.

Same market. Two opposite readings.

The difference comes down to the question of phase. A bounce inside a downtrend and the start of an uptrend look identical on a five-minute chart, and they trade in opposite directions. Crypto market cycles are not there for the sake of elegant theory, they are there so you know which trades are even worth considering today.

What market cycles are

A crypto market cycle is a repeating sequence of asset redistribution between participants with different time horizons and different amounts of capital. Not a calendar, not astrology, and not "every four years on schedule." A process with mechanics you can understand.

It rests on three pillars.

Liquidity is finite. A participant who wants to build a $200 million position will not buy it with a single market order. The order book is empty by the second hundred million, price runs up 5-8%, and the average entry ends up worse than planned. So large capital builds over weeks, inside a range, with limit orders. It unloads the same way. The stretched-out nature of that process is what creates the phases.

Leverage creates forced trades. A liquidation is a forced closing of a position by the exchange when the margin no longer covers the loss. In essence it is a market order placed not by a person but by the exchange engine: it does not look at price and it cannot be cancelled. A cascade of such orders turns a normal correction into a 15% collapse in an hour. On October 10, 2025 the market saw $19 billion in liquidations in 24 hours, the largest event in history. The trigger was the announcement of 100% tariffs on Chinese imports, but the scale came from leverage. The mechanics of collapses like that are covered in the article "Liquidation cascades: why crypto drops so fast".

Attention is reflexive. Rising price generates news, news brings in new money, new money pushes price further. It works the same way in reverse. Hence the asymmetry: rallies stretch out longer than declines, and fear plays out faster than greed.

The four-phase scheme was described by Richard Wyckoff back in the 1930s for the stock market. Crypto added three quirks: round-the-clock trading with no exchange-level circuit breakers on the way down, a halving that cuts issuance on schedule, and a record share of leveraged retail.

And the switch never flips instantly. Between accumulation and markup lies a zone where some participants are already buying while others are still selling, and it can last for months. The exact date of the regime change will be named by analysts after the fact, in the moment nobody sees it.

The 4 phases: accumulation, markup, distribution, markdown

First the overview table, then a breakdown of each phase through order flow.

Accumulation. The phase begins once there is nobody left to sell. Price stops making new lows not because demand showed up, but because supply ran out. The range is wide, the moves are sluggish, and bitcoin's average daily ATR compresses to 1.5-2% against 4-6% at peak activity.

In the order book it looks distinctive. At the lower boundary sits a density level (a cluster of large limit orders at a single price), it holds for tens of minutes, market sells eat into it again and again, and price does not break down. Big volume traded, no movement. Absorption.

The classic indicator marker: price inside a range while OBV trends up. Someone is taking the offer inside the range without driving price higher. Historically, accumulation in bitcoin has run from six to fourteen months.

Markup. The range is broken to the upside, pullbacks get bought, every new low is higher than the last. Volatility expands steadily rather than in jumps.

Order flow changes too. Density levels above stop holding price, they get eaten through in minutes. The funding rate (the periodic payment between holders of long and short positions on perpetual futures) is positive but moderate, around +0.01% per eight-hour period. Breakouts follow through, and there are noticeably fewer fakeouts than in a range. This is exactly where the simplest strategies work, which is what gives beginners the illusion of talent.

Distribution. The most expensive phase for anyone who fails to recognize it. Price sits at the highs, volume is the highest of the cycle, and there is no progress in price.

The signs:

• price inside a range while OBV trends down: a large participant is unloading into everyone who thinks this is accumulation
• the range widens, with wicks in both directions that get retraced quickly
• funding settles into a steady positive above +0.05%, the crowd is paying to hold longs
• BTC dominance declines, money flows into alts, the altseason talk starts
• footprint charts show absorption at the upper levels: volume trades, price does not move

Markdown. Demand is exhausted, but the leverage is still there. The first serious flush takes out stops, stops push price further, and then liquidations kick in. The order book is empty, price jumps across dozens of levels without a single trade, and the spread widens several times over.

Then comes the fade. Volume drops, volatility compresses, interest leaves. And this is exactly where markdown quietly turns into accumulation.

If order flow is still hard to read, watch the free lesson from the "Trading from scratch" course on the Secret Terminal YouTube channel: it shows on a live chart how professionals break the market down through the order book and footprints.

How to identify the market phase

There is no single indicator for this. What works is the intersection of three independent layers: volume and volatility, bitcoin dominance, and sentiment through the Fear & Greed Index. Plus a fourth check through order flow, which separates real interest from imitation. No layer gives you an answer on its own, all three together give you a probability you can actually work with.

Volume and volatility

Start with volatility, because it is measured unambiguously while volume requires interpretation.

The absolute ATR value tells you nothing. ATR(14) = 95 on a BTC chart means the average bar of the last fourteen periods covered 95 dollars. At a price of 67,000 that is 0.14%, at a price of 300 dollars it would be a catastrophe. What you compare is the normalized value.

ATR% = (ATR / current price) × 100

Daily BTC ATR% as a regime indicator:

Now volume. The key question here is not "how much," but "what did price do with that volume."

An example. A five-minute bar closed 12 points above the previous one on 380 BTC of volume. A cumulative indicator will credit all 380 to the buyer. But the footprint (the distribution of volume across prices inside a single candle) may show that market buys were 160 BTC and sells were 220, delta minus 60, and price rose only because the seller pulled his offers above. The buyer was actually weaker.

That distinction is what separates markup from distribution. In markup, price goes up on positive delta. In distribution, it stalls or grinds higher on neutral and negative delta, because someone is handing volume to the market.

Three combinations I check first:

• price rising, OBV flat. The move is being dragged along on thin liquidity, the first serious market sell puts it all back
• price in a range, OBV trending up. Accumulation. On the daily timeframe this signal is rare and deserves attention
• price in a range, OBV trending down. Distribution

A separate note on August 2026. The rally from 64,100 started with a cascade in which 92% of liquidations hit short positions. After the event open interest (the total volume of contracts still open) fell roughly 15%, and funding flipped from negative to positive. Both facts say the same thing: the move was position closing, not money coming in, meaning short squeeze mechanics rather than the start of a markup phase.

Which does not cancel out the second factor. Over that same week spot bitcoin ETFs took in $1.92 billion in net inflows, the best weekly result since October 2025. That is outside money. The signal is mixed, and calling it an unambiguous start of a new phase is not justified.

BTC dominance

Bitcoin dominance shows its share of total crypto market capitalization.

BTC.D = (BTC market cap / total market cap) × 100

As of August 28, 2026 the figure is 59.3% with total capitalization at $2.70 trillion. The current cycle peaked at 65% in June 2025, and the all-time low remains 31.1% in January 2018, at the height of the ICO boom.

A detail almost nobody writes about. Total capitalization includes stablecoins, and there are more than $300 billion of them, which understates the standard BTC.D figure by roughly 6-8 percentage points. Count dominance only among risk assets and you get about 64.4%. The difference matters: 59% and 64% sit in different interpretation zones.

On its own the number is useless. You read it paired with the direction of the market.

The point is that dominance shows risk appetite inside crypto. As long as money is concentrated in bitcoin, participants are cautious. Once the rotation into second- and third-tier alts starts, appetite is maxed out, and that is a late-stage sign.

An important correction to the historical templates. BTC.D has not dropped below 50% for almost three years, since September 2023, and the reason is structural: $56.9 billion of inflows into spot ETFs since January 2024. Institutional money buys bitcoin, not memecoins. The old logic of "dominance drops below 45%, therefore altseason" may simply fail this cycle, because the inflow channel has changed. The mechanics of the metric are covered in "Bitcoin dominance: what it is, where to track it and how to use it".

Fear & Greed Index

The Fear & Greed Index compresses market sentiment into a single number from 0 to 100. It is calculated from six components with different weights: volatility over 30 and 90 days (25%), market momentum and volume (25%), social media signals (15%), investor surveys (15%), bitcoin dominance (10%), and Google Trends search queries (10%).

The scale splits into five zones. Below 24 is extreme fear, 25-46 fear, 47-54 neutral, 55-75 greed, above 76 extreme greed.

Current bitcoin readings as of August 28, 2026: the index is at 68, a third consecutive day in the greed zone. A week ago it was 87, a month ago 50.

You cannot treat this indicator as leading. Half its weight sits in volatility and momentum, meaning derivatives of price. The index does not predict the move, it describes what has already happened in a form that is easy to read.

And here I will say something unpopular that contradicts the famous quote about greed and fear. Historical data shows bitcoin delivered better results after extended periods of extreme greed than after periods of fear. Sustained greed accompanies a strong trend, and a strong trend tends to continue. Shorting on the mere fact of a reading of 85 has cost a lot of people money.

The index works in two cases. First: extremes lasting weeks, not a single day. A reading below 15 that holds for two consecutive weeks statistically coincides with accumulation zones, above 85 for a month with distribution. Second: sharp reversals in the metric itself. A drop from 87 to 68 in seven days while price rises tells you euphoria is cooling faster than price.

When the method fails. In 2018 the index sat in the fear zone for months while price went from 6 thousand to 3.2. Buying on the "everyone is scared" signal cost 45% of the account. In the middle of the scale, from 40 to 65, the reading means nothing at all. The methodology is broken down in the article "Crypto Fear and Greed Index: how it works and how to use it".

Historical BTC cycles

Four completed cycles and one still open. Figures rounded to the significant digits.

The main pattern is obvious right away. Drawdown depth shrinks step by step: 94, 87, 84, 77 percent. The reason is not participant maturity but capital structure: the larger the capitalization, the larger the share of money held by people who do not sell at minus thirty. The second pattern is that the length of the decline has settled around a year.

The current cycle still fits the picture. Ten months have passed from the October 2025 peak to August 2026, the low was set around 64 thousand, the drawdown is 49%. Some analysts expect the bottom in an October-December 2026 window in the 50-55 thousand range. That is a forecast, not a fact.

Caution here is mandatory for a simple reason. Four observations is not statistics. Any pattern built on four points breaks on the fifth with exactly the same probability that it continues.

The connection to the halving

The halving cuts the block reward in half. In April 2024 the reward dropped from 6.25 to 3.125 BTC, and the next halving is expected in April 2028.

Let us count what that means in money. At an average block time of ten minutes, 144 blocks are mined per day: 900 BTC of new issuance per day before the halving, 450 after. At a price of 80 thousand dollars, daily issuance amounts to $36 million. Meanwhile spot ETFs took in $1.92 billion in a single week of August 2026, around $274 million a day, meaning institutional inflows exceed all new issuance by roughly seven and a half times.

The conclusion is unpleasant for fans of the simple model: the direct impact of the halving on the supply-demand balance falls with every cycle, because the quantity being cut keeps getting smaller relative to market turnover.

But the timing relationship still holds so far.

The last three landed in the 17-18 month range. Whether that is coincidence or pattern, nobody has an honest answer. The working version: the halving operates less as an economic event than as a narrative. The date is known in advance, people write about it half a year out, expectations and marketing campaigns get built around it. The inflow of attention and money happens because everyone expects an inflow of attention and money.

The practical meaning does not disappear because of that. If the 17-18 month lag repeats, the next top window shifts toward the end of 2029, and 2026-2027 falls into accumulation ahead of the April 2028 halving. The statistics and mechanics of the event are covered in "Bitcoin halving: what it is, how it affects price and when the next one is".

One more caveat. Every cycle broke some rule of the previous one: in 2021 the top was a double top, with a 55% drop between the two peaks, and anyone trading by the calendar got caught in that drop.

Strategy by phase

The phase does not cancel trading, it changes the parameters. A scalper works in all four regimes, but with different risk settings, different targets, and a different default side.

In accumulation the main enemy is boredom. ATR% is compressed, there is little movement, and the fees have not gone anywhere. You work off the range boundaries, anchoring entries to density levels that hold for a long time. The order lifetime timer becomes your primary filter: a real buyer sits at the level for tens of minutes, while a spoofer blinks in and out in seconds. What you cut is not the risk, it is the number of trades.

In markup the rule is simple: trade in the direction of the move and do not try to catch the top. Pullbacks get bought, the risk-to-reward ratio is better than in any other phase, and you can afford a wider stop because price comes back. What I increase here is not leverage but the number of positions open at once: at the same total risk you get more independent attempts.

In distribution everything changes. The holding horizon shortens: what used to be held for an hour gets closed in fifteen minutes. Long and short become equally valid, structure breaks in both directions. And funding control becomes mandatory. At a rate of 0.05% per eight hours a position pays 0.15% a day: a 10x long on a $1,000 account carries $10,000 of notional, the payment comes to $15 a day, meaning 1.5% of the account. Over a week that adds up to more than ten percent of the account without a single price move.

In markdown the things that work nowhere else start working. Liquidation cascades produce the fastest moves of the year, but you enter them on flow, not on an idea.

An example from practice. I was working through a SOL setup in the 178-181 range. At 178.60 there was a density level of about 23 thousand coins in the book, with a lifetime timer over an hour. By the third approach to price the order had shrunk to 9 thousand and the timer showed eight minutes instead of an hour plus: the old one was pulled and a smaller new one placed. Then the tape took off, market buys came through in 800-1,500 SOL chunks, and the remainder of the density level was eaten in forty seconds.

Entry at 178.55 after a bar closed above the level. Stop under the level at 177.90, a distance of 0.65 dollars. Account $10,000, risk 0.5%, meaning $50. Position size = 50 / 0.65 = 76 SOL. Take profit at 181.20, where the nearest large density level sat above, risk-to-reward 1 to 4. It ran to 180.85 and I closed it manually.

The key point: the reason for the entry was the eaten density level and the acceleration on the tape. The market phase only set the side I was willing to consider a trade on. And what it never cancels is risk per trade. Accounts get blown not by people who read the phase wrong, but by people who put half the account behind a phase they read right.

The market phase in the order book and the tape

Everything described above works on a horizon of weeks and months. A scalper needs the projection onto today, and that is read through order flow. What to watch when the regime changes:

• Tape speed. In accumulation prints come in rarely and small. In markup the flow is steady with one side dominating. In distribution the tape is choppy, large prints go both ways with no result in price
• How density levels behave. The question is not whether the market is rising or falling, but whether the order gets eaten or holds. A density level that has stood for forty minutes is confirmed capital interest
• Delta in the footprint. Volume traded, price did not move, so someone took it. Regular absorption at the upper levels is a distribution sign
• Funding across all exchanges at once. If the rate is -0.9% on one venue and -0.2% everywhere else, the skew is local and there will be no market-wide cascade
• Spread width and book depth. In markdown the order book is empty, and your stop fills with slippage several times larger than you calculated

Secret Terminal has the tools assembled for these five jobs. The density map is built to a depth of 5% on both sides of price and projected straight onto the candlestick chart as colored zones, with an order lifetime timer next to every large order. Footprints with delta show the distribution of volume across prices inside the candle along with the POC. The funding module lists rates from Binance, Bybit, OKX, MEXC and WhiteBIT in a single column with a countdown to settlement, so a cross-exchange comparison takes a second. The built-in screener with a turnover filter shows where there are actually participants today: when the phase changes, the makeup of that list shifts before the picture on the daily chart does. The terminal is free, runs on Windows and macOS, and keys are stored locally.

Common mistakes when working with cycles

Calling the phase off a single metric. A fear index at 20 does not by itself mean the bottom, a compressed ATR does not mean accumulation. You need at least two confirming layers.

Copying the previous cycle's template one to one. Waiting for minus 84% because that is what happened in 2018 means missing the turn at minus 50%.

Confusing a bounce with a reversal. A rally on short covering and a rally on money coming in look identical for the first 24 hours. They differ in open interest: OI dropping 15% while price rises is position closing.

Trading the halving calendar. The date is known to everyone in advance, which means it is already in the price. The 2021 double top showed that even a correct general idea does not save you from a 55% drop in the middle.

Averaging against the move in a markdown phase. The fastest way to turn a manageable loss into a liquidation. Same story with dragging your stop.

To take the order book apart piece by piece and understand where entries come from, the free lesson from the same course helps: limit orders, liquidity and working with levels on a real instrument.

FAQ

• What is a market cycle in crypto in simple terms?

  It is a repeating sequence of four phases: accumulation, markup, distribution and markdown, through which assets get redistributed between participants. Cycles exist because large capital physically cannot enter or exit in a single trade and stretches the process over months.

• How long does a crypto market cycle last?

  Historically around four years, of which the decline takes 12-14 months. The four-year interval comes from the bitcoin halving, but the precision here is nominal: four completed cycles is far too little for statistics.

• What are the crypto market phases and how do they differ?

  Crypto market phases differ not by the direction of price but by the ratio between volume and result. In accumulation volume trades while price sits at the lows. In markup price goes up on positive delta. In distribution volume is the highest of the cycle while there is no progress in price. In markdown the move is driven by forced position closures, not by people deciding to sell.

• How do you know the market is in the accumulation phase?

  Price has stopped making new lows, daily ATR% has compressed to 1.5-2%, and there are density levels at the lower boundary of the range that hold for tens of minutes without breaking. Additional confirmation is a range with a trending-up OBV. Historically this phase in bitcoin has lasted from six to fourteen months.

• What phase is the crypto market in right now?

  As of late August 2026 the signals are mixed. Bitcoin bounced from 64 thousand to 80, but the main impulse came from $2.77 billion of short liquidations rather than money coming in. Against that stands $1.92 billion of net inflow into spot ETFs over the week. The correct wording: this is a test of supply after a ten-month decline, not a confirmed markup phase.

• How does the halving affect the price of bitcoin?

  The direct impact on the supply-demand balance weakens with every cycle. After the April 2024 halving daily issuance is 450 BTC, which at a price of 80 thousand is $36 million a day, while inflows into spot ETFs have reached $274 million a day. The historical timing relationship still holds: the last three tops landed 17-18 months after the halving.

• How does a bounce differ from a trend reversal?

  By the behavior of open interest and funding. If price rises while open interest falls 10-15% and funding flips from negative to positive, that is short covering. A reversal driven by money coming in looks different: open interest rises along with price, and volume stays elevated for several days in a row.

What to take away

A cycle is not a schedule, it is the mechanics of capital redistribution. Phases exist because big money does not fit into a single click, and leverage turns ordinary corrections into cascades.

Calling the phase off one indicator is pointless. You need three layers: volatility and volume give you the regime, dominance shows risk appetite, the sentiment index marks the extremes. Plus order flow, which separates real interest from imitation.

Historical templates are useful as a frame and dangerous as a rule. Drawdowns shrink from cycle to cycle, dominance has not gone below 50% for three years, and the money inflow channel changed after the ETFs launched. And the last point, the most practical one: the phase sets the side and the size of your risk, but it does not give you the entry. The entry is always in the same place it has always been, in the order book and on the tape.

Look at the market where liquidity is visible. Secret Terminal puts the order book with a density map to a depth of 5%, footprints with delta, the tape and funding rates across five exchanges into one window. The order lifetime timer separates confirmed interest from spoofing, the built-in screener shows where the participants are today, and auto-tuning on the C key fits the book to a coin's volatility in a second. A change of market phase shows up in order flow before it shows up on the daily chart, and the terminal stays free.