Fidelity Digital Assets said in its latest Q3 Signals Report that the crypto market has been stuck in consolidation for more than 200 days, and several of its indicators now sit near historical capitulation zones. Even so, the firm said the current adjustment may still be incomplete.
The report’s headline readings were a weighted NUPL of -0.01 and Bitcoin dominance of 68%. Fidelity said those numbers show a market that is slightly below aggregate break-even, with capital still concentrated in BTC while Ethereum and Solana remain in unrealized loss territory.
Fidelity said the drawdown has lasted 203 days so far. That compares with roughly 300-day bottoming cycles in 2018 and 2022. On that basis, the current phase may be only about two-thirds complete. The firm said October 2026 could serve as a time window worth watching, but added that this is not a prediction that the market will bottom then.
Weighted NUPL slipped below zero while Bitcoin remained the market’s main support
Weighted NUPL measures whether a market-cap-weighted basket of digital assets is sitting in aggregate unrealized profit or unrealized loss. Because Bitcoin’s market value is far larger than that of ETH and SOL, Fidelity said the current reading is driven heavily by BTC.
Among the three assets in the report, only Bitcoin still showed unrealized profit. Ethereum and Solana were both in unrealized loss. Put together, the weighted NUPL came in at -0.01, meaning the overall market sat just under the break-even line.
Fidelity said that leaves most remaining unrealized gains concentrated in Bitcoin rather than spread across the market. BTC has acted as a stabilizer inside the basket, offsetting part of the losses posted by ETH and SOL, though not enough to bring the combined reading back into positive territory.
The firm added that if BTC, ETH and SOL were given equal weights, the portfolio would look weaker because the losses in ETH and SOL are deeper. In that sense, the market structure is still somewhat healthier than a setup in which losses are evenly distributed across all assets, even though weighted NUPL has already turned negative.
Fidelity said this underlines Bitcoin’s role as the market’s reference asset. ETH and SOL have continued to weaken relative to BTC, which the report read as a sign that investors still prefer the largest and most liquid asset while staying cautious on digital assets with a history of higher volatility.
By the second quarter of 2026, Bitcoin’s own NUPL had fallen further and the weighted reading for the three-asset basket dropped to -0.01. Fidelity said the market is more likely to remain range-bound or stay under pressure until more assets move back into unrealized profit, rather than quickly returning to broad-based expansion.
Bitcoin dominance rose to 68% as capital stayed selective
After falling through the second half of 2025, Bitcoin dominance has resumed a gradual climb, and Fidelity said the longer-term upward trend has not been broken. Historically, rising BTC dominance has often gone hand in hand with underperformance in the rest of the digital asset market.
The report said that pattern reflects a familiar shift during periods of uncertainty and valuation pressure: capital tends to move toward the most liquid and most established asset. Fidelity pointed to a series of higher lows in BTC dominance and a relatively steady slope higher as evidence that the move is not just a short-lived rotation.
In the second quarter of 2026, BTC dominance edged up from 67% to 68%. Fidelity said there was still little sign of broad capital rotation into the rest of the market. If dominance starts to fall or flatten in the third quarter, that could indicate a recovery in risk appetite and renewed investor attention toward other digital assets, potentially marking an early shift in market structure.
Over the past year, BTC was down about 45%, ETH fell 37%, and SOL lost 53%, according to the report’s rolling-return framework. Year to date, BTC, ETH and SOL were down 33%, 47% and 41%, respectively.
By the end of the second quarter, Bitcoin had dropped below its 200-week moving average. Fidelity said market sentiment had moved into an extremely depressed state, and the broader digital asset market had weakened alongside it.
The firm linked the decline to a mix of factors listed in the report: an unfavorable macro backdrop, capital shifting toward AI-related investments and equities, and fading market momentum. It said several indicators are now near or inside historical capitulation zones.
Spot ETP flows remained negative as well. Fidelity said spot ETPs recorded cumulative net outflows of $5.4 billion year to date in 2026, with $4.9 billion of that coming in the second quarter. June alone saw roughly $4.5 billion in net outflows, making it the worst month since the launch of spot Bitcoin ETPs.
From June 1 to June 4, the market also saw nearly $6 billion in cascading liquidations. Fidelity said forced selling added to downside pressure and disrupted existing positioning.
The report also cited macro and geopolitical headwinds. Inflation remained high, energy prices stayed under pressure, and expectations around monetary policy shifted sharply. At the start of the year, the market was still pricing in the possibility of no rate cuts in 2026; by the end of the second quarter, expectations had shifted toward a possible rate hike by year-end. Short-term volatility remained elevated, and Fidelity said market bottoms typically take time to form.
Bitcoin signals: NUPL at 0.09, momentum negative, Yardstick near undervalued levels
Fidelity said current valuation levels, along with an increasingly visible negative correlation between digital assets and traditional risk assets, could create attractive entry points for longer-term investors, provided that adoption trends at the underlying network level continue to strengthen.
At the end of the second quarter, Bitcoin’s NUPL stood at 0.09, in what the report described as the “hope-fear” zone. That suggests the market still holds a modest amount of unrealized profit, but sentiment has turned cautious. Some holders remain in profit, though Fidelity said there is no broad agreement yet that a durable bottom has already been set.
Bitcoin fell 14% in the quarter, and its NUPL dropped from 0.21 to 0.09, a decline of 0.12. Fidelity said that looked more like a moderate compression in unrealized profits than a full-scale capitulation event. Based on current data, Bitcoin’s market price was about 10% above the aggregate network cost basis, leaving investors with roughly $108 billion in unrealized profit.
Through much of April and May, BTC’s NUPL sat in the “optimism-anxiety” zone as the market gradually warmed to the idea that a bottom might already have formed. The latest slide back into the “hope-fear” zone suggests a move away from sustained optimism and toward caution and uncertainty. Fidelity noted that in past Bitcoin bear markets, NUPL has fallen deeper into capitulation territory, which is why the present reading still calls for restraint.
Historically, when BTC’s NUPL has been within 0.05 of 0.09, the median one-year return that followed was 53%, and the three-year annualized compound growth rate was 69%, based on 73 observations. Fidelity said the correlation between NUPL and future one-year and three-year returns was -0.26 and -0.80, respectively, meaning lower NUPL readings have often been associated with stronger long-run returns. The firm still cautioned that historical relationships can weaken or fail and should be read alongside macro conditions and broader market structure.
On momentum, Fidelity said Bitcoin’s current reading reflects a bearish impulse in this drawdown. Prices failed to keep posting higher highs over the last quarter. The signal turned positive on April 21, 2026, when BTC traded at $78,317 and both short-term volatility and price momentum moved above their medium-term averages. The rebound did not last. On June 1, the signal turned negative again with BTC at $66,540, indicating that momentum had faded and price had failed to hold.
Bitcoin traded between $58,500 and $82,256 during the quarter, with especially sharp swings in April and May. Fidelity said the model produced one false positive in the second quarter: early positive momentum looked as if it might persist, then reversed quickly. The firm said that is an unavoidable cost of trend-following systems, whose purpose is not to call every top and bottom precisely but to participate once an uptrend forms while limiting downside exposure.
Since the signal flipped negative on June 1, Bitcoin has fallen another 10% and remains in a negative momentum regime. Short-term realized volatility was about 34% annualized, below the 40% medium-term volatility reading. For the signal to turn positive again, Fidelity said either short-term volatility would need to rise moderately or medium-term volatility would need to fall further. The indicator is meant to identify phases in which price direction and volatility move together, not to forecast exact tops or bottoms. For now, the reading still points to caution.
On valuation, Bitcoin has fallen more than 50% from its all-time high, while network hash rate is down only about 22% from peak levels. Fidelity said that shows meaningful stress on miners but also notable resilience in the network itself. As a result, Bitcoin’s Yardstick metric has moved close to historical lows, which the firm said suggests BTC may be trading at a meaningful discount relative to the hash power needed to maintain and secure the network.
Fidelity also said this cycle is different in some respects, including lower price volatility and a more mature mining industry. Because price is a direct input into Yardstick, lower volatility makes the relative impact of hash rate more visible in the ratio. At the same time, mining firms have become better at managing energy costs and operating efficiency, giving them more flexibility to adjust machine uptime, relocate equipment, or optimize power contracts as profitability changes.
That means mining capacity can respond more flexibly to price changes, reducing the chances of the kind of extreme divergence between price and network energy input seen in earlier cycles. Price weakness and lower hash rate have pushed Yardstick into what Fidelity calls the “undervalued” zone. Over the last 92 days, the metric spent 76 days, or about 83% of the time, more than one standard deviation below its long-term mean.
Historically, Fidelity said, that zone has often lined up with accumulation phases or relative bottoms. Yardstick stayed there for 298 days in 2018 and for 299 days in 2022 before sentiment gradually recovered. With the current bear market at around 203 days so far, October 2026 is a time period worth monitoring, though not a guaranteed bottom date.
Fidelity also revisited Bitcoin’s store-of-value case by comparing BTC with gold. Measuring Bitcoin in gold rather than in fiat terms can show how it is performing against another traditional safe-haven asset. The report said recent price swings have weakened the store-of-value narrative in the short run, but not enough to invalidate the longer-term investment thesis.
In the second quarter of 2026, Bitcoin fell 14% in U.S. dollar terms, but its price relative to gold was broadly unchanged. After roughly a year of underperformance, investor preferences between BTC and gold may be moving back toward balance. Since Feb. 28, Bitcoin has risen 15% against gold. Fidelity said the relative performance of the two assets has begun to stabilize under ongoing geopolitical uncertainty. The report also noted that gold’s strong performance over the past year has been driven in large part by continued central bank buying, and that BTC-gold dynamics may now be settling into a more stable equilibrium.
On-chain readings still point to an accumulation phase, with some metrics already showing signs of capitulation. For longer-horizon allocators, Fidelity said Bitcoin’s previous stretch of underperformance against gold may now make valuations more compelling.
Mining pressure builds as some infrastructure shifts toward AI
Fidelity said daily average hash rate and the 30-day average remain below the 1,000 EH/s, or 1 ZH/s, milestone reached in September 2025. The report gave two main reasons: growing demand from AI and high-performance computing businesses for power, land and data-center resources, and a bear market that has squeezed Bitcoin mining economics.
Some mining firms may be redirecting power capacity, data-center infrastructure and new capital spending toward AI or high-performance computing. With BTC prices under pressure, Fidelity said AI compute contracts can offer steadier and more predictable revenue, making that kind of infrastructure reallocation economically rational.
The report stressed that Bitcoin ASIC miners are specialized hardware and usually cannot be repurposed directly for AI workloads. In practice, the shift toward AI is more about reusing power, land, facilities and cooling systems than converting existing mining machines into AI servers. Idle machines are also unlikely to stay dormant for long. Fidelity said a more common outcome is equipment sales or relocation to regions with lower electricity costs, rather than a permanent exit from the network.
Hash rate has kept falling since BTC peaked in October 2025, while mining difficulty has stayed elevated instead of dropping in step with price. Falling prices and tighter margins have pushed miners operating near the cost line to power down.
Fidelity said deteriorating mining economics have led to temporary hash rate declines in past bear markets as well. This time, however, competition from AI data centers and power contracts for the same infrastructure base could keep the hash rate downtrend going longer than historical averages would suggest. In the second quarter, daily average hash rate rose 8% quarter over quarter, but the 30-day average fell 6%. Since the start of 2026, Bitcoin has fallen by more than 29% while hash rate is down only about 12%, a sign that miners are under pressure but the network still retains a measure of resilience.
Ethereum: deeper unrealized losses, weak base-layer fundamentals, and lower fee income
Ethereum’s NUPL moved deeper into the capitulation zone in the second quarter, and Fidelity said market sentiment deteriorated further. ETH fell 25% over the quarter, while NUPL dropped from -0.12 to -0.43, a decline of 0.31. Unrealized losses widened sharply. Based on current data, ETH was trading about 30% below the aggregate network cost basis, with total unrealized losses around $87 billion.
On June 6, ETH’s NUPL touched a local low of -0.46 and has not broken below that level since. Fidelity said the metric could still make a lower low, but holding that level for now may be something long-term investors want to watch.
Historically, when ETH’s NUPL has been within 0.05 of -0.43, follow-on returns have often been strong. Since 2018, comparable readings have been followed by a median one-year return of 70% and a three-year annualized compound growth rate of 133%, based on 90 observations in each horizon. The correlation between NUPL and future one-year and three-year returns was -0.13 and -0.81. Fidelity said that is one reason it treats lower NUPL as a positive signal, while also noting that historical patterns may weaken or fail under different market conditions.
Ethereum’s momentum signal turned positive on April 16, 2026, when ETH traded at $2,350 and both volatility and price momentum moved above their medium-term averages. The move did not hold. On May 17, the signal flipped back to negative with ETH at $2,130. During the second quarter, ETH traded between $1,564 and $2,422, with especially sharp swings in April and May.
Fidelity said Ethereum, like Bitcoin, saw one false positive from its momentum model in the quarter. Since the signal turned negative again on May 17, ETH has fallen another 25% and remains in a negative momentum regime. Short-term realized volatility was about 50% annualized, well below the 71% medium-term volatility reading. A positive turn would require either a meaningful rise in short-term volatility or a large drop in medium-term volatility. As things stand, the indicator points to simultaneous weakness in price and volatility, not to a return of upside momentum.
On fundamentals, Fidelity said Ethereum’s base layer cooled off in the second quarter, broadly matching the decline in ETH’s price and volatility. Total transaction count still showed some resilience, down just 5% over the quarter, with daily volume remaining above 2 million transactions throughout.
Transaction fees stayed well below historical averages, though still sensitive to short-term changes in demand. On April 22, Ethereum’s median transaction fee briefly rose to $0.42, then kept falling and ended the quarter at about $0.02. Active addresses and new addresses pulled back after setting record highs in the prior quarter, declining 10% and 31%, respectively. Fidelity said the relationship between on-chain activity and asset price remained fairly strong during the quarter.
The share of addresses linked to real economic activity rose slightly, though Fidelity said a sizable portion of addresses still made no meaningful contribution to Ethereum’s revenue or security. The report said that trend may continue through 2026. The planned Glamsterdam upgrade is expected to focus on expanding base-layer capacity, which could add more block space supply.
Stablecoin transfer value on Ethereum has climbed above historical averages, supported by scaling progress and an improving regulatory environment, according to the report. Over the last 12 months, total stablecoin transfer value kept hitting new highs and exceeded $20 trillion on a cumulative basis.
Growth has started to slow, though. Fidelity said average daily stablecoin transfer value in June was about 9% lower than the average of the prior three months. After a period of rapid expansion over the last year, the firm said the stablecoin market may shift toward a steadier growth rate in the year ahead.
Even with digital asset prices falling broadly, stablecoin transfer volume on Ethereum continued to rise. Fidelity said that points to real usage demand gradually decoupling from market sentiment and asset prices, with stablecoins being used more for payments, settlement and access to on-chain dollars globally rather than only for speculative trading. Average stablecoin transfer costs have stayed below $1 for three straight quarters, which the report said also confirms the practical effect of earlier scaling measures.
Network fee revenue on Ethereum continued to decline over the past year. In the second quarter, rolling 12-month network fees fell 15%, from $344 million to $294 million. Fidelity said scaling at the protocol and infrastructure layers has continued to outpace growth in demand for block space. With developers placing greater emphasis on base-layer scaling again, the report suggested that trend could persist.
The planned Glamsterdam upgrade is expected to add more block-space capacity, which means Ethereum network fees may stay under pressure over the next year. Fidelity added that fee levels are volatile by nature and that a dependable long-run equilibrium is hard to pin down. In the second quarter, daily network fees ranged from $145,000 to $2.75 million, with an average of about $575,000 per day.
One of the main signals investors should watch over the next several years, Fidelity said, is how core developers balance network growth against value capture. As a technology platform still in development, Ethereum has historically prioritized adoption, ecosystem expansion and network utility more than short-term revenue. Unless developers and researchers devote more effort to improving value-capture mechanisms, network fees and protocol income may remain under pressure.
Solana: stablecoin transfers grew while fee revenue approached cyclical lows
Solana’s NUPL stayed in the capitulation zone throughout the second quarter. SOL fell 12% over the quarter, and NUPL slipped from -0.67 to -0.72, a decline of 0.05. Unrealized losses widened further. Based on current data, SOL was trading about 41% below the aggregate network cost basis, with total unrealized losses near $29 billion.
On June 6, SOL’s NUPL rebounded sharply from a local low of -1.08, underscoring the token’s higher volatility in this bear market. Fidelity said the recovery from that low may indicate that a meaningful share of early holders has already sold, with newer buyers starting to step in at lower prices.
Readings around -0.72 plus or minus 0.05 have been rare in Solana’s history, but past performance after those signals has been strong. Since the network’s launch, comparable readings have appeared 21 times, and the median one-year return that followed was 542%. Fidelity said there is not enough long-run history yet to calculate a reliable three-year return series. The current correlation between SOL’s NUPL and future one-year returns is -0.56, which is relatively strong, though the firm cautioned that Solana’s shorter operating history and limited sample size make the relationship less reliable than Bitcoin’s.
Solana’s momentum signal has remained negative since Oct. 28, 2025, with price and volatility broadly trending lower together. That still points to a cautious market backdrop. More recently, however, short-term realized volatility rose above medium-term volatility, at about 63.5% and 61% annualized, respectively. Fidelity said that kind of setup has at times appeared around momentum reversals in the past. If price starts to strengthen as well, it would give firmer support to the case for a local bottom.
Unlike BTC and ETH, SOL did not post a false positive momentum signal during the quarter. It traded mostly between $63 and $97 in the second quarter, with the price near $81 at the start of the period. Under Solana’s own model parameters, even a move to $97 was not enough to turn the signal positive, and the price later made another lower low. When the signal first turned negative on Oct. 28, 2025, SOL was trading near $194. It has since fallen about 60%, and Fidelity said the adjustment still does not look fully complete. For now, the indicator looks more like an attempt to stabilize than a return of bullish momentum.
Even in the bear market, Solana’s fundamentals have shown resilience. Asset prices kept weakening, but demand for on-chain activity did not contract in step. Monthly transaction count continued to rise and was up 1% quarter over quarter in the second quarter, keeping the network on track for a record high.
User growth did not keep pace with activity growth, which means average transactions per user increased. In the second quarter, monthly active addresses and new addresses fell 15% and 7%, respectively, while addresses linked to real economic activity slipped only 4%. Fidelity said economically meaningful on-chain activity remained comparatively stable in this environment, a pattern that differs somewhat from Ethereum.
Another advantage highlighted in the report was stable transaction cost. Solana’s median transaction fee stayed below 0.1 cent throughout the quarter and moved very little, giving users and investors a high degree of cost predictability.
Stablecoin transfer value on Solana has historically been volatile, but the longer-term uptrend has remained intact since growth accelerated more than a year ago. Average daily stablecoin transfer value is now holding above $8.4 billion, up 17% from the prior quarter. Compared with Ethereum, individual stablecoin transfers on Solana are smaller, which Fidelity said reflects differences in user mix and use cases across the two networks.
Over the last 12 months, Solana processed more than $2.6 trillion in stablecoin transfers. While SOL’s price has fallen sharply, stablecoin transaction volume and other on-chain activity have remained steady. Fidelity said, as on Ethereum, a meaningful portion of stablecoin demand on Solana appears sticky and not tightly tied to short-term market moves. If on-chain activity continues to expand, the network’s stablecoin ecosystem could grow with it.
In the second quarter, the total size of the broader stablecoin market shrank by about 1.3%, but stablecoin supply on Solana rose about 3%, or roughly $300 million.
Network fee revenue on Solana remained in a downtrend, though the report said it was beginning to look for an equilibrium level. Rolling 12-month network fees fell 18% in the second quarter to $221 million, with average daily fee income of about $390,000. If second-quarter fee income is annualized, that works out to about $141 million, a figure now much closer to the rolling 12-month number.
At the same time, on-chain activity on Solana is still growing, and several Solana Improvement Documents, or SIMDs, are placing greater emphasis on value capture for SOL holders. Fidelity said those developments reinforce Solana’s positioning as a technology platform capable of generating protocol revenue, with SOL at the center of that economic system.
On Fidelity’s view, Solana’s network fee revenue may already be near a cyclical bottom. If on-chain activity keeps growing and value-capture proposals are rolled out gradually, the report said fee income could begin to recover over the next year.
Appendix: how Fidelity framed momentum, Yardstick and NUPL
The report included a methodological appendix describing several of its core indicators and their limits.
Momentum signal
Fidelity said the momentum signal combines price trend and volatility dynamics to classify the current state of a digital asset. One side of the model compares short-term price change with the longer-term trend. The other checks whether current volatility is expanding or contracting relative to a recent baseline. Those two dimensions are then folded into a single momentum classification used to identify phases in which price and volatility are moving together, diverging, or shifting from one state to another.
Lookback windows and other parameters are selected through an optimization process designed to create relatively clear and stable distinctions across market regimes. Fidelity said the signal is descriptive rather than predictive and should not be treated as a price forecast, investment recommendation or trading call.
Yardstick
Bitcoin Yardstick, also referred to as the hash rate Yardstick, can be thought of as a rough price-to-earnings-style metric for the Bitcoin network. A traditional P/E ratio uses stock price or market capitalization divided by corporate earnings. Yardstick instead uses BTC market capitalization divided by total network hash rate to estimate how richly or cheaply Bitcoin is valued relative to the security resources supporting the network.
A lower ratio suggests BTC is cheaper relative to the hash power needed to secure the chain, in the same general way that a lower P/E is often read as a cheaper stock valuation. Fidelity added that hash rate is not the same thing as earnings, so Yardstick is only a relative valuation framework and should not be equated directly with an equity multiple.
NUPL
Fidelity said the relationship between NUPL and future returns over one-year and three-year horizons is among the stronger on-chain relationships in its research. Even so, Ethereum and Solana have much shorter network histories than Bitcoin, which leaves fewer observations and lowers the reliability of those historical links.
In theory, NUPL can fall below -1.0 when a network’s realized capitalization exceeds twice its market capitalization. In the early years of BTC, ETH and SOL, a meaningful share of token supply moved or was distributed before a liquid public market price had formed.
Fidelity noted that early Bitcoin transfers took place before a mature market price existed, while Ethereum and Solana also went through early token issuance, presales, pre-mines, seed financing and foundation allocations. Those early distributions can affect realized price and may sometimes be recorded at costs above later public market prices.
When market price falls below realized price, aggregate unrealized losses can exceed current market capitalization, pushing NUPL below -1.0. As networks mature and on-chain transaction histories deepen, realized capitalization should reflect actual market transactions more than early distribution events. Fidelity said that is why NUPL generally becomes more informative over time as a network matures.