Galaxy: Is the "Four-Year Cycle" of Bitcoin still valid? Where is the bottom of this cycle?

Compiled by: Tao Zhu, Jinse Finance; Author: Alex Thorn, Managing Director and Head of Research at Galaxy Digital

This report indicates that the Bitcoin “four-year cycle” may still be valid, but volatility is narrowing. Market and on-chain data suggest that BTC has not yet bottomed out, and this article presents several possible bottoming scenarios.

Introduction

Bitcoin has been around for 17 years, exhibiting long-term cyclical price fluctuations. Approximately every four years, Bitcoin’s price climbs to a high point, then suffers a painful decline to a low, followed by a recovery. This cyclical pattern has historically been closely linked to the four-year halving events, which cut the regular supply in half. Although the impact of halving events is gradually weakening and many predict a “super cycle,” empirical data again shows that the four-year cycle remains effective. This report explores these fluctuations and a recurring pattern throughout Bitcoin’s modern history: each wave is more subdued than the last.

The peak in October 2025 is the calmest in Bitcoin’s history, with subsequent declines also unusually gentle. If the peak is so mild, should we expect the final cycle low to be equally gentle? If so, where might the bottom roughly land?

This report assumes that the current downtrend has not yet bottomed and provides data supporting this hypothesis. Data also indicates that the relatively smooth top in October 2025 could lead to a shallower bottom. Historical experience suggests that the baseline bottom of the current decline could be between $40k and $46k, occurring before Q4 2026. (The baseline scenario is for reference only; actual results may vary significantly.)

Importantly, this analysis relies entirely on market and on-chain data and time-based analysis. The predicted cycle lows do not incorporate or depend on external events (such as regulation, market developments, or geopolitical factors) or their timing or impact.

Bitcoin Four-Year Cycle Overview

Each Bitcoin cycle begins from a previous low, experiences halving, rises to a high, then declines to the next low. The following are four cycles, including the current one:

The bottom of the current cycle has not yet formed; its retracement magnitude and duration are based on data as of June 9, 2026. Note that this report is based on two observed patterns: the peaks and troughs of each cycle are shrinking (from 85% to 84%, then to 77%), and historically, bottoms tend to appear about 12-13 months after each top. The current cycle is only 8 months from its most recent top.

Because the October 2025 top is relatively modest (on an index basis) compared to previous cycle tops, the market’s average price paid for tokens (i.e., the actual price, effectively the “cost basis”) is unusually close to historical highs: about 43.7% of previous all-time highs, whereas in past cycles it was typically around one-third or less. This is a key data point: similar sell-offs at the end of previous bear markets are likely to conclude at higher dollar prices this time. Comparing cycle timing, amplitude, and on-chain indicators suggests that the current decline may bottom within these levels:

The levels above and the analysis in this report suggest that we believe the bottom of this cycle has not yet arrived. Few of our previous cycle bottom indicators have shown rebounds; in terms of time, this decline is still shorter than past declines; and if genuine panic occurs, the cost basis itself could also decrease. Our argument is that the four-year cycle still exists empirically, but its amplitude has narrowed. A calmer top has elevated the bottom but has not eliminated it.

Scope of Analysis

In price cycles, tops and bottoms are nearly impossible to identify precisely in real-time but become obvious in hindsight. Therefore, we evaluate how many conditions that occurred at each previous top and bottom are simultaneously present now. To develop a set of indicators for assessing past tops and bottoms, we examined five major evidence categories: valuation (are prices high or low relative to what holders paid?), profit-taking (are holders selling during price rises or being forced to sell during declines?), miners (are miners profitable or under significant pressure?), trend (how far is the price above or below its long-term average?), and market sentiment (greedy or fearful?).

Applying these five perspectives to the current cycle’s top and bottom yields a clear conclusion: Bitcoin’s volatility is decreasing. Each top’s gains are smaller than the previous, and subsequent declines are also diminishing. If this “compression” of amplitude is real and applies in both directions, it may provide clues about the expected cycle lows during this decline. Based on this, we can estimate where Bitcoin might bottom in this downtrend.

The Two Ends of This Cycle

This analysis requires identifying indicators and establishing baselines to determine the cycle’s top and bottom. We use the same scoring method for each endpoint: comparing current levels to those at previous cycle tops and bottoms.

Tracking the Cycle Top

A top indeed exists but is the calmest on record. At the October high, only two of the eleven classic warning indicators reached the previously lowest top levels, and both just barely. The clearest valuation indicator—the ratio of market value to realized value (MVRV, i.e., price relative to the average price paid by holders)—peaked at only 2.29, compared to 2.93 to 5.91 at previous three cycle tops. The entire “greed” indicator group hit the lowest historical readings for cycle tops, and the Pi cycle top indicator (a timing signal that accurately predicted the previous three tops) was completely absent (a first in Bitcoin history). Yet, the timing was textbook: the top occurred 1062 days after the previous low, roughly aligning with 2017 and 2021 tops. Surprisingly, the real euphoria appeared 18 months earlier, around the launch of the US spot Bitcoin ETF. Even as enthusiasm waned afterward, prices continued to rise. Now it appears to be more institutional buying rather than retail frenzy driving the surge.

The chart below shows our full top indicator suite for the current cycle (using the October 2025 all-time high as a reference).

Of the 11 strength signals: two confirmed, two partially confirmed (at least breaking through 85% of the K-line), and seven inactive. Among the two confirmed signals (RSI and SOPR), both only slightly broke through their weakest 62k-line, reaching peaks again in 2023 and 2024, not at the October 2025 high. Crucially, the Pi cycle top signal was not activated (these two indicators are treated separately because time is a calendar fact, not a measure of top strength). “Previous tops” refer to the range between the 2013, 2017, and 2021 cycle tops; the threshold is the least active among these, the 2021 peak, which is the easiest to break through. “Cycle peak” is the extreme reading of each indicator in the current cycle, with the month noted. Reserve risk and Pi cycle ratio are our internal metrics.

Plotting the Cycle Bottoms

During this decline, only 4 of 13 bottom signals have appeared, three of which are weaker indicators (panic sentiment; trend indicator in the bottom zone; and the first breach of the 200-week moving average). The fourth appeared in early June, marking the first miner signal: the recovery crossover of Hash Ribbons, where the 30-day hash rate average re-crosses above the 60-day average after a decline. Historically, this signal often presages a bottom.

The strongest signals that always appear at genuine bottoms (price below cost basis, holder losses, persistent stop-losses, deep capitulation selling) have not yet occurred. The current decline is -51%, far less than the -77% to -85% lows seen at previous cycle ends, and below the mid-2021 -53% decline.

However, the pace of decline has changed. Using the same point in the cycle (about 8 months after the peak, or 242 days), recent declines have brought the drop slightly below the levels seen in the 2013-2015 cycle (−48%, during a rebound), so it is no longer the shallowest path (most declines then were smaller). The 2017-2018 and 2021-2022 cycles experienced much larger declines (~−68%) at this stage. According to the cycle clock, the bear market bottom window opens only at the end of 2026.

Each line tracks the decline from peak to trough in a cycle, starting at day 0. About 242 days later (dashed line), the current cycle (orange, −51%) is slightly below the 2013-2015 cycle (−48%). This means this cycle is no longer the shallowest (most declines then were the shallowest). The previous two cycles at this stage had declines close to −68%. All these cycles are well above current levels (green zone indicates past bear market bottoms).

The chart below shows our full retracement indicator scorecard for the current decline, using previous bottom indicators.

Of the 13 target indicators, 4 have been reached, 2 are approaching, and 7 are not yet achieved.

To illustrate these bottom indicators, the table lists when they appeared at previous cycle bottoms and compares to current conditions. Comparing these 13 indicators across the first three cycles clearly shows their evolution. At each past bear market low, all 13 indicators eventually reached their bottom zones, with the only difference being timing: some appeared early, others later. Today, only 4 indicators have bottomed, with the only miner-related indicator (Hash Ribbons) being the latest. (Notably, during this decline, the Hash Ribbons reversal seems to have preceded the bottom, unlike previous cycles, possibly due to external factors like Bitcoin miners transitioning to AI, which did not occur in earlier cycles.)

The previous cycle cells show how many days each indicator’s most extreme value was ahead (+) or behind (−) the price low of that cycle (within a 180-day window). The Hash Ribbons line indicates the recovery crossover; the cycle clock marks 12 months after the top. Each indicator reached the three previous bottoms, with signals appearing earlier or later. The current cycle’s low has not yet appeared, so the current column only shows whether each box has been checked since the October 2025 high.

Price Volatility at Both Ends Is Narrowing

Before drawing conclusions, an obvious fact must be emphasized: the volatility of Bitcoin’s price is narrowing at both ends. Tops are declining (MVRV at 5.91, 4.72, 2.93, and 2.29), and bottoms are rising each cycle, with MVRV increasing from 0.56 in 2015 to 0.69 in 2018, then to 0.75 in 2022. In other words, the gap between overvalued and undervalued points in each cycle is shrinking. The sharp declines also confirm this: previous drops of 85%, then 84%, then 77%, while this cycle has so far declined 51%.

The red curve in the chart represents each cycle’s top, the blue curve each bottom, with the price-to-cost ratio (MVRV) near “fair value” (1.0). Data shows that this cycle may not have bottomed yet (hollow diamond indicates deepest reading so far). This describes the cycle pattern but does not guarantee the bottom’s location.

The cooling of tops and rising bottoms describe three completed cycles but are not natural laws. They do not prove the next low will be shallow. But they do allow us to pose a precise question and get an exact answer: if the bottom’s movement resembles past bottoms, how much of the dollar decline is determined by the strength of the top?

A Calmer Top Raises the Bottom

MVRV is today’s price divided by the on-chain cost basis. Conversely, the cost basis is the historical peak divided by the top MVRV. Therefore, the lower the top MVRV, the closer the cost basis is to the peak.

Since the October top was the calmest (MVRV at 2.29), the cost basis ultimately reached 43.7% of the all-time high (compared to 34.2%, 21.2%, and 16.9% at 2021, 2017, and 2013 tops). A calm top does not lower the bottom; instead, it raises the cost basis closer to the peak, thus elevating the bottom (all else equal).

The share of the all-time high that the cost basis represents. Because each cycle’s top was relatively gentle, this share has been rising each cycle, reaching 44% by 2025. The annotations next to each bar show how past typical bottom patterns affected the dollar decline in that cycle.

Assuming the bottom pattern remains unchanged (i.e., each cycle’s bottom occurs at the same MVRV level), the dollar decline in each cycle would shrink, simply because the initial cost basis is higher. This table demonstrates that without any prediction:

Each cell shows the decline from peak to bottom in that cycle, assuming the bottom occurs at the MVRV value in that column, using that cycle’s own cost basis relative to the peak. All rows have the same bottom trend; only the calmness of the top varies. The typical historical bottom (MVRV 0.70) in 2013 implied an 88% decline, but this cycle’s bottom is only 69%. This isolates the impact of a calm top; it’s a simple arithmetic calculation, not an assertion that calm tops cause higher bottoms.

How Far Below the Anchor Will the Bottom Go?

Bottoms are not measured in exact percentages but relative to two key anchors: the cost basis and the 200-week moving average (200w MA). The latter has been Bitcoin’s long-term support line since inception. Using these anchors, the recent bear market lows have fallen well below both: on average about 33% below the cost basis (44% in 2015) and about 14% below the four-year moving average.

Two points are noteworthy. First, the gap relative to the cost basis has been shrinking each cycle (−44%, −31%, −25%), matching the compression trend seen at the top. Second, current prices have not even reached that zone. Even with a 51% decline, Bitcoin’s price remains 14% above the cost basis (never falling below it in this cycle) and only 1.5% below the four-year moving average. Based on past bottoming standards, this cycle has not yet bottomed.

The chart below shows the historical gaps between bear market lows and the cost basis (blue) and four-year moving average (purple). Past lows were well below these levels. Today’s price remains above the cost basis, slightly below the 200-week MA, with the gap narrowing each cycle.

The anchors and the calculations align. Mapping past gaps onto today’s anchors points to the same zone: a decline of 25% to 44% below the cost basis, roughly $30k to $40k, while the gap to the four-year MA is about $41k to $62k. This suggests the real bottom could be below today’s price but well above the previously predicted “75% to 85% decline.”

Where Is the Bottom?

Translating these calculations into price, based on today’s $53k cost basis, yields a range of scenarios rather than a single figure; let’s examine the middle scenario first.

Our base scenario is that the bottom continues to rise each cycle, ultimately reaching fair value (MVRV 0.75–0.86), with prices around $40k to $46k. If the decline is larger, reaching levels seen in 2018/2022 (MVRV 0.56–0.70), prices would be around $30k to $37k. If the decline is smaller, stable buying absorption near cost basis (MVRV 0.95–1.01) would keep prices around $51k to $54k; and touching the four-year average ($62k) would only result in about a 51% decline. (For reference only; actual results may vary significantly.)

The scenarios shown in the chart are based on price. The cost basis and rising four-year MA (which historically tracks bottoms) are well above the previous “75–85% decline” zone (gray, invalidated). The colored bands translate past style bottoms into today’s dollar prices. These levels are based on the premise that bottoms form, not that they are imminent. For reference only; actual results may vary significantly.

The key is how this influences old rules of thumb. A 77%–85% decline (a common measure in past cycles) would push the bottom to between $19,000 and $29,000. But this rule essentially re-uses the calm top assumption, applying a deeper percentage decline relative to the peak to a peak already close to the cost basis, because the top was calm.

The cost basis is the most clearly rising trend in the entire chart and the most obvious signal that the “bottom” could move. Over the past year, the cost basis has risen from about $47k to nearly $56k by the end of 2025 (a 20% increase), as higher bids in this cycle reset the average price. This rise is the fundamental reason why the current bottom is far above the old rules. But afterward, the actual price has fallen about 5% to around $53k, causing some tokens maturing in 2024–2025 to incur losses. By late 2026, the actual price-to-cost basis ratio will be a key volatility factor: a calm decline will stabilize it around $40k, while genuine panic could push it lower, dragging the entire market down.

The crucial point: the bottom may move

The cost basis is reflexive. It appears as a bottom but is built from the last traded price of tokens. During genuine sell-offs, tokens are exchanged at losses, pulling down the average price, so the bottom follows the price decline rather than remaining stable.

This is the most significant limitation under a high-bottom hypothesis. The buffer is thin: current prices are only about 14% above the cost basis (MVRV at 1.14), and in this cycle, it has never fallen below the cost basis. If a sell-off causes the cost basis to drop by 10%, 20%, or 30%, the typical bottom could fall from about $40k to roughly $36k, $32k, or $28k, returning to normal historical ranges.

Maintaining the bottom pattern while allowing the cost basis to decline during sell-offs. Implicitly, the bottom could fall from about $40k back to around $28k, re-entering the normal historical zone (amber area). A calm top has elevated the bottom; genuine panic could reclaim some of the gains.

Stabilizing the bottom while the cost basis declines during sell-offs suggests the bottom could shift downward, with the implied bottom dropping from about $40,000 to near $28,000, re-entering the typical historical range. Calm tops have lifted the bottom; true panic could wipe out some of those gains.

Stabilization of the bottom pattern, with the cost basis falling during sell-offs, implies the bottom could move lower. The implied bottom could fall from roughly $40,000 to about $28,000, re-entering the normal historical range (amber zone). Calm tops have elevated the bottom; genuine panic might erase some of those gains.

The presence of stable spot ETFs and corporate bonds providing steady, price-insensitive buying has been lacking in past cycles, tending to push the bottom higher but also buffering or amplifying declines. Due to their funding channels, digital asset bond firms (DAT) and corporate bonds tend to buy high rather than low, and recent ETF fund flows in 2026 have been net outflows. During deep sell-offs, fund redemptions may force selling rather than absorb it. The 2022 cycle experienced the largest forced sell-off in crypto history, but the decline was only -77%. Therefore, “lower leverage this time” is not necessarily a reason to be optimistic. (These are auxiliary points, not core arguments.)

Higher lower bounds and the risk of erosion during panic are two sides of the same coin: this cycle’s higher cost basis means that during genuine capitulation, the cost basis could fall. That’s why we focus on the range rather than a single value.

Comprehensive Analysis: Data Indicates Retracement Magnitude

Our analysis clearly indicates the scope and duration of the retracement.

A calmer top pattern has elevated the price base to 43.7% of the all-time high, so for any given bottom pattern, the dollar decline is milder than in previous cycles. We believe the rule “Bitcoin has historically fallen 75%–85%, so this cycle’s bottom is near $19,000–$29,000” no longer applies and cannot be used as a basis for price bottoms. Even with a decline comparable to past severe drops, the fall is still much larger than previous ones. Therefore, even our more aggressive forecast suggests a bottom around $40,000 mid-range.

Compared to past cycle indicators and timing data, the bottom may not have arrived yet. Only 4 of 13 bottom indicators are active, and the current decline has lasted about 8 months, whereas historically bottoms tend to form over 12–13 months (and the cost basis itself could also decline).

If deep signals (price below cost basis, holder losses, persistent stop-losses, four-year moving average breach and sustained, and deep bear market declines) begin reversing at prices well above old ranges, it indicates genuine compression on both bullish and bearish sides. Conversely, if a full collapse occurs as expected, the calm top merely delays pain rather than alleviating it. In either case, the calculation of the cost basis shows that the starting point is well above the levels assumed by the old four-year rule.

This is a descriptive study exploring how calm cycle tops influence the calculation of cycle bottoms, not a prediction of price direction or target levels. Our horizontal lines use historical data to compare current retracements relative to the current cost basis (which itself can change).

Appendix A: Chart Library

We provide numerous auxiliary charts grouped by theme. The first set summarizes the cycle; the second set details the complete bottom consolidation list, analyzing one indicator at a time. In each indicator chart, shaded bands represent the ranges reached at the 2015, 2018, and 2022 lows, with orange marks indicating the latest readings.

Cycle in the Charts

Price and its cycle top. Bitcoin’s full price history (log scale), with red indicating the past three cycle tops, orange the October 2025 high.

Price and its cycle bottom. The same history, with reference lows marked: 2015, 2018, and 2022 bear market bottoms (red), plus declines during COVID-19 and mid-2021 (gray).

Cycle clock. Shows how many days each top occurs after the previous low (circle) and halving (square). October 2025 falls within the historical cycle range.

Market euphoria arrived early. The valuation peak of this cycle appeared in early 2024, around the ETF launch; then, as on-chain enthusiasm waned, prices rose another ~70%, reaching the October 2025 high.

Untriggered signals. The Pi cycle peak, marked with an asterisk, identified the 2013, 2017, and 2021 peaks within days. In this cycle, it was never triggered (a first for any cycle peak).

Bottoming Checklist, Indicator-by-Indicator Analysis

MVRV (Market Value to Realized Value ratio). Past lows pushed it well below 1.0; this cycle’s low so far is 1.14.

NUPL (Unrealized Profit/Loss). Past lows pushed it below zero (accumulated losses); currently still positive.

MVRV Z-score. Standardized version of MVRV. Past bottoms showed deep negative values; this cycle remains positive.

Mayer Multiple. Price divided by its 200-day moving average. It has fallen into the bottom zone, the closest to bottom signals among all trend indicators.

Price vs. four-year moving average. The 200-week MA is Bitcoin’s most stable support. Past bottoms touched or fell below this line; now, for the first time in this cycle, the price has dipped below it.

SOPR (Spent Output Profit Ratio). The average profit/loss of coins spent in a given period. After previous bottoms, SOPR was below 1.0 for months (stop-loss); in this cycle, SOPR is only slightly below 1.0.

Realized profit/loss. Daily locked-in profit (+) or loss (−), scaled by market size. The major loss peak indicating a true bottom has not yet appeared.

Puell Multiple. Miner revenue pressure. Past miner bottom ranges were 0.30–0.41; this cycle’s low (~0.44) is close but not reached.

Hash Ribbons. Mining hash power momentum. Below 1.0, miners are capitulating; by 2026, it has persistently fallen below this threshold.

Fear & Greed Index. Our proprietary 0–100 sentiment indicator shows that during this decline, fear was more intense than at previous bottoms. This is the only truly bottomed indicator.

Appendix B: Glossary

Bitcoin Cycle. Bitcoin roughly experiences a four-year cyclical pattern: years of ascent to a high, sharp decline to a low, followed by a long recovery. Each cycle is typically marked by a halving event.

Halving. Approximately every four years, the rate of new Bitcoin creation halves. This is a fixed protocol mechanism, historically serving as a cycle anchor.

All-Time High (ATH). The highest daily closing price Bitcoin has ever reached. The current cycle’s ATH is $124,824 on October 6, 2025.

Retracement. The percentage decline from a peak. A -50% retracement means the price has fallen to half of its all-time high.

Cost Basis, also known as Realized Price. An estimate of the average price paid for held Bitcoin. Technically, it’s the last on-chain transaction value divided by the number of Bitcoins. It’s the most important benchmark in this report; also called network cost basis.

Market Cap. Total USD value of all Bitcoin at current prices (price × circulating supply).

Realized Market Cap. Total value of all Bitcoin at their last transaction prices (not current). Realized cap divided by supply.

MVRV Ratio. Market cap divided by realized cap: equivalently, current price divided by network cost basis. Above 1.0 indicates profit on average; below 1.0 indicates loss. It’s the core metric of this report.

MVRV Z-score. Standardized version of the MVRV gap, enabling comparison of extreme highs and lows across different price periods.

NUPL (Net Unrealized Profit/Loss). The proportion of unrealized profit relative to market cap. Higher positive values suggest greed near tops; negative values (accumulated losses) suggest capitulation near bottoms.

SOPR (Spent Output Profit Ratio). The average profit/loss of coins spent in a given period. Above 1.0 indicates profit-taking; below 1.0 indicates losses (bottom signal).

Mayer Multiple. Price divided by its 200-day moving average. A simple measure of deviation from medium-term trend.

200-Day / 200-Week Moving Averages. The average closing prices over the past 200 days (medium-term trend) or 200 weeks (~4 years, the most durable long-term support).

Puell Multiple. A measure of miner revenue pressure, comparing newly mined Bitcoin’s USD value to its one-year average. Named after analyst David Puell.

Reserve Risk. A measure of long-term holder confidence in price. Shown on a scale; used here only comparatively.

Pi Cycle Top. Triggered when the 111-day moving average crosses above twice the 350-day moving average. It accurately predicted the 2013, 2017, and 2021 tops within days but was never triggered in this cycle.

Hash Ribbons. Comparing 30-day and 60-day hash power averages. When short-term falls below long-term, miners are capitulating; a recovery crossover often precedes bottoms.

Fear & Greed Index. Proprietary 0–100 sentiment indicator based on on-chain, derivatives, and flow data. Low readings indicate extreme fear (near bottom); high readings greed (near top).

RSI (Relative Strength Index). 0–100 momentum oscillator; high readings indicate overbought conditions, often near tops.

Cycle Clock. Days from cycle start low or halving to top or bottom. The last three cycle tops occurred about 1060 days after the previous low; bottoms about 12–13 months after tops.

Reflexivity. A concept popularized by George Soros, meaning that the indicator itself influences the market. Here, the cost basis appears as a bottom but can decline during sell-offs as tokens are traded at losses, making the bottom a moving target rather than a fixed line.