Bitcoin Whales Taking Profits? Understanding SOPR and Realized Profit Metrics in One Article

On October 8, single-day realized profit on-chain reached approximately $1.03 billion, the second-highest value so far in 2026 and only slightly below the annual peak of $1.04 billion. According to statistics from Santiment cited by DailyCoin, this concentrated realization occurred after Bitcoin spiked to around $87,000, before the price quickly fell back to near $81,000, a weekly decline of about 5%. During the same period, roughly $520 million in long positions were liquidated, and U.S. spot Bitcoin ETFs saw net outflows exceeding $702 million. According to Fortune's market record on October 13, BTC traded at $83,408.20, up about 7.95% from a month earlier and down about 27.64% year-over-year, still well below its all-time high of $126,198.07 set on October 6, 2025.

When price stagnates while realization volume expands, this is precisely the moment when the question of "are holders selling?" needs to be answered quantitatively. On-chain data can answer it because every spent bitcoin carries two prices at once: the price when it last moved, and the price when it moved this time. Dividing the two gives SOPR; subtracting the two and multiplying by the amount gives realized profit or realized loss. The former measures the nature of realization behavior, while the latter measures its scale. Together they describe facts that have already occurred, not unrealized gains or losses on paper.

Key Takeaways

The dividing line for SOPR is 1. According to Glassnode's metric documentation, SOPR equals the USD value of outputs spent during the period at the time of spending divided by their USD value at the time of creation. A reading above 1 means coins moved during the period were on average sold at a profit, while below 1 means they were on average sold at a loss.

155 days divides holders into two categories. The same documentation explains that STH-SOPR only counts spent outputs with a lifespan shorter than 155 days, while LTH-SOPR only counts those longer than 155 days. The two reveal completely different behaviors.

Realized profit and loss is the dollar-denominated version of SOPR. Glassnode's definition of net realized profit/loss provides the conversion relationship: net realized profit/loss equals realized profit minus realized loss, and also equals 1 minus the inverse of SOPR, multiplied by the USD transfer volume for the period.

SOPR looks at behavior; MVRV looks at valuation. The definition of MVRV is market cap divided by realized cap, measuring the market-wide level of unrealized gains relative to the cost basis. SOPR measures whether the portion of coins that have actually moved were sold at a profit or loss.

2026 provided samples from both extremes. In early June, long-term holders realized approximately $2.4 billion in losses within 48 hours, and LTH-SOPR fell below 1. After mid-August, SOPR climbed back above 1 and stayed there for three weeks, the longest profit-taking stretch of the year.

Ratios and amounts must be read together. SOPR tells you how many times over sellers on average profited, while realized profit tells you how large those realizations were in dollar terms. Looking at only one can easily mistake small-scale panic for a large-scale exit.

Why Profit-Taking Is Worth Quantifying Now

What $1 billion in one-day realization means

The $1.03 billion single-day realized profit recorded by Santiment does not in itself constitute a bearish conclusion, and the firm also explicitly stated that this dataset does not guarantee a major reversal. Its real value lies in confirming one thing: the October push from around $81,000 toward $87,000 encountered actual seller supply, and that supply came from holders sitting on unrealized gains rather than from passive panic selling. The subsequent price decline shows that this realization was not absorbed by buying pressure of a similar scale.

This is consistent with the direction of capital flows during the same period. That week, spot Bitcoin ETFs saw net outflows exceeding $702 million, meaning incremental allocation capital was contracting while on-chain realization was expanding. For an ongoing method of tracking ETF flows, see Bitcoin ETF Flow Tracker.

Price is still about 30% below its all-time high

Over a longer time frame, the current price remains about 34% below the all-time high of $126,198.07 and is down about 27.64% from a year ago, but up about 7.95% from a month ago. What makes this position unusual is that holders with different cost bases are in completely opposite situations: coins bought near the highs in Q4 2025 are still underwater, while coins bought at lower levels in mid-2026 already have sizable unrealized gains. Price alone cannot distinguish which of these two groups is selling. The role of SOPR and realized profit/loss is precisely to separate "who is moving coins, and whether they are doing so at a profit or loss" out of aggregate data. For a framework for judging the current cycle position, see A Systematic Review of Bitcoin Market Cycles and Historical Bull Market Drawdown Distributions.

What Exactly SOPR Measures

Calculation method and the dividing line at 1

Every unspent output of Bitcoin records the block time when it was created, and therefore also corresponds to a USD price at that time. When that output is spent, it corresponds to a new price. Glassnode's metric documentation sums the USD value at the time of spending for all outputs spent during the period as the numerator, and sums the USD value at the time of creation for those outputs as the denominator. Dividing the two gives SOPR.

The interpretation is therefore very direct. Above 1, coins moved during the period were on average sold while in profit; below 1, they were on average sold while at a loss, a situation usually accompanied by panic or capitulation-style selling; equal to 1 means coins were sold at the average cost basis. The metric was proposed by Renato Shirakashi, and in its launch note on June 24, 2019, Glassnode also offered an empirical observation: in bull markets, SOPR falling below 1 tends to be repaired quickly because holders expect prices to continue rising and are unwilling to exit at a loss; in bear markets, SOPR rising above 1 is likewise repaired quickly because sellers fear further declines and view breaking even as an acceptable outcome. The line at 1 is therefore not only an accounting boundary between profit and loss, but often also a behavioral support or resistance.

Why aSOPR excludes outputs younger than one hour

Raw SOPR includes all spent outputs in its statistics, including a large number of very short-lived change outputs, internal consolidations, and restructuring transactions, none of which represent real buy or sell decisions. Glassnode's documentation explains that aSOPR (adjusted SOPR) excludes outputs with a lifespan shorter than one hour to more clearly reflect actual buying and selling behavior. In daily analysis, aSOPR is usually cleaner than raw SOPR, especially on days when exchanges conduct large-scale wallet reorganization.

Breaking Holders Apart: STH-SOPR and LTH-SOPR

The dividing line at 155 days

Glassnode's documentation defines STH-SOPR as including only spent outputs with a lifespan shorter than 155 days, and LTH-SOPR as including only those longer than 155 days. CryptoQuant's metric explanation describes the short-term holder definition more specifically, with a statistical range of outputs older than one hour but younger than 155 days, also using 1 as the threshold.

The number 155 days is not arbitrary; it corresponds to the historical time window in which holding behavior changes significantly. Holders beyond that period are statistically more inclined toward long-term holding rather than short-term trading. Reading the two curves separately carries very different implications: short-term holder realization often reflects trading sentiment and leverage conditions, while long-term holder realization is closer to structural supply release.

The two curves told two stories in 2026

This year's market happened to provide samples from both ends. In early June, according to an analysis by Coinspeaker republished via Yahoo Finance, long-term holders realized approximately $2.4 billion in losses over the 48 hours ending June 5, and LTH-SOPR fell below 1, meaning coins held for more than 155 days were sold at a loss. The analysis compared this pattern to lows in January 2015, December 2018, and November 2022. The price subsequently touched its yearly low of $58,035 on June 25, the lowest level since September 2024. Long-term holders capitulating has historically tended to appear near cycle bottoms rather than tops.

The other end was the repair after summer. According to a report by COINOTAG citing CryptoQuant data, as of early September, SOPR read 1.002 and had remained above 1 for three consecutive weeks since August 19, the longest profit-taking stretch of 2026. Entering October, according to a chart posted on X by analyst Darkfost and relayed by ChainCatcher, the weekly average of long-term holder SOPR returned to 1.25 and the monthly average to 1.16, both above the annual average of 1.10, corresponding to an average realization gain of about 10% for long-term holders over the past year. This dataset comes from an individual analyst's publication and has not yet been confirmed by an official data platform standard, but its direction is consistent with the aforementioned expansion in realized profit.

It should be emphasized that SOPR remaining above 1 does not by itself equal a bearish signal. Profit-taking is a normal component of an uptrend. The combination truly worth watching is when realization scale keeps expanding while price no longer makes new highs.

Realized Profit and Realized Loss: The Dollar Scale of the Same Event

Conversion between net realized profit/loss and SOPR

Glassnode's definition of this metric set is quite clear. Realized profit is the total USD gain generated by coins moved on-chain during the period relative to their previous movement price, realized loss is the corresponding total loss, and net realized profit/loss equals the former minus the latter. The documentation also provides its algebraic relationship with SOPR: net realized profit/loss equals 1 minus the inverse of SOPR, multiplied by the USD transfer volume for the period.

This conversion reveals the true division of labor between the two metrics. SOPR is a dimensionless ratio that only answers how much was gained on average; net realized profit/loss multiplies that ratio by transaction scale to answer how many dollars those realizations were worth. With SOPR equal to 1.05, the actual market pressure is completely different on a low-volume weekend versus a high-volume up day.

Why ratios and amounts must be read together

According to Glassnode's explanation, positive net realized profit/loss means profits are being realized. If price is still rising at that time, it indicates incremental capital is absorbing sell pressure; if peaks keep rising while demand weakens, it may correspond to a local top. A negative value means losses are being realized, and extremely deep negative values are often associated with capitulation-style selling and historically tend to appear before reversals. A value near zero means coins are moving near their cost basis, which in a bull market pullback may indicate confidence repair, while in a bear market it may indicate trapped supply exiting en masse near breakeven.

Putting October 8 back into this framework: SOPR above 1 indicates sellers were overall in profit, $1.03 billion in realized profit indicates the scale of this realization ranked second for the year, and the price decline from $87,000 to $81,000 indicates buying pressure did not fully absorb it. The three datasets point to the same conclusion, and any one of them alone would be insufficient to support it.

On-chain data describes what has already happened, while markets keep changing. Open the BTC spot market and verify on-chain signals against the order book

The Division of Labor Between SOPR and MVRV

MVRV is defined as market cap divided by realized cap, where realized cap represents the sum of the cost basis of the entire circulating supply. According to Glassnode's documentation, when the ratio is above 1 and continues rising, it means the market's unrealized gains relative to the cost basis are expanding, and holders' incentive to realize profits strengthens accordingly. Historically, above 3.5 corresponds to large-scale distribution in the later stages of a bull market; below 1 means most of the supply is below its cost basis, historically corresponding to capitulation and accumulation near bear market bottoms.

The two metrics answer completely different questions. MVRV answers "if everyone sold now, how much would they gain on paper," measuring unrealized gains that have not yet occurred, making it a valuation thermometer for potential sell pressure. SOPR answers "for the portion already sold, was it actually a profit or a loss," measuring behavior that has already occurred. High unrealized gains do not mean holders will sell, which is MVRV's blind spot; SOPR's blind spot is that it only sees coins that have moved, not the vast majority of coins sitting still.

In practice, a more effective approach is to link the two into a chain: MVRV tells you how much pressure exists, while SOPR and realized profit/loss tell you whether that pressure is being released and how fast. For specific interpretations of MVRV and realized price, see Assessing Bitcoin Valuation with MVRV and Realized Price.

How to Use It, and What It Cannot See

Three common misreadings

The first is using SOPR directly as a buy or sell signal. The metric oscillates frequently around 1, and single-day crossings are extremely common in any market. Only deviations lasting several weeks, or patterns that diverge from price action, have analytical value.

The second is ignoring transfer volume. SOPR is a ratio, and when on-chain activity is light, a small number of transactions can push the reading very high or very low. Extreme values at such times are often unrepresentative and must be checked against the absolute amount of realized profit/loss.

The third is equating realized profit with sell pressure. When coins move from one address to another, it is counted as a spend on-chain, but this is not necessarily a sale to the market; it could also be a transfer to custody, a position swap, or internal reorganization. True sell pressure needs to be judged alongside exchange inflows, funding rates, and open interest. For related methods, see <