The price of Ethereum has declined consistently over the last seven days. Many factors could be responsible for this trend, including the activity (buying and selling of ETH) of large traders, commonly known as whales.

Ethereum price chart for the last 7 days. Source: Coinmarketcap.com

To find out if whales are responsible, we analyzed Ethereum trading data for decentralized exchanges (DEX) from Dune Analytics for the last seven days and answered four important questions.

• Have large traders been buying or selling ETH?
• Has their activity increased or decreased over this period?
• Which whale groups are driving the activities and capital flows?
• Are these activities responsible for the ETH price decline?

We categorized whales into three cohorts, placing them in $100K–$500K, $500K–$1M, or $1M–$5M. The analysis shows that while whales have been active within this period, they are not directly responsible for the price decline.

No Strong Buying Or Selling Bias

The data shows that large DEX traders had a slight buying bias, buying approximately $358 million of WETH and selling approximately $352M during the period, with a $6M net difference. This difference isn’t significant, indicating a nearly balanced buying and selling pattern within the last seven days.

Whale buy vs sell volume for the last seven days. Source: Dune.com

High Volatility But No Clear Trend

Whale activity showed no clear trend over the last seven days, as there is no clear increase or decrease. However, sharp differences in activity are clear, with a huge decline over the weekend indicating significant volatility. This shows that large traders have been active in the market.

Whale activity showing volatility in the last 7 days. Source: Dune.com

Smaller Size Whales Drive Activity But Capital Flow

Large traders with amounts ranging from $100K–$500K accounted for 86% of whale activity by transaction count. This shows that moderately large traders were the most frequent participants over the period.

Whale trading counts. Source: Dune.com

However, the $1M–$5M cohort, with only 8.7% of trades, comes very close in terms of trading volume at 45.13%, compared to 45.64% for the $100K–$500K cohort. This indicates that while moderately sized whales are the most frequent participants, larger whales have a disproportionately greater impact on capital flows despite their significantly fewer transactions.

Whale trading volume in USD. Source: Dune.com

ETH Price Decline Unrelated To Whale Activity

Despite sharp differences in the large traders’ activities by day, ETH price declined consistently, while whale trading volume fluctuated without a corresponding trend over the period in focus. This shows that the price decline has no direct link with large trader activities in the last seven days.

Ethereum price with respect to whale activity. Source: Dune.com

Conclusion

The data reveals that while there is no clear trend in the activity of whales over the last seven days, ETH price continued to fall. The findings suggest that whale activity over this period is not directly responsible for the price decline.

However, the analysis only considers data for the last seven days, which may be too short a period to give an accurate picture of what is really going on. Also, the price decline may be driven by other factors such as general current sentiment among retail traders, who ultimately are the majority of the market.

An interesting direction to also consider, which this analysis doesn’t cover, would be to determine where large traders are sending their ETH to—whether to centralized exchanges (CEX) or to private wallets. The finding could give important insight into the minds of these big players and tell us what to expect from the market next.

Upbit and Bithumb will end trading support for Storj (STORJ), JasmyCoin (JASMY) and ThunderCore (TT) on Sept. 14 after concluding that the issues behind earlier warning designations remained unresolved. 

Upbit’s notice covers STORJ/KRW, STORJ/BTC, JASMY/BTC, JASMY/USDT, TT/KRW and TT/BTC, with trading scheduled to stop at 3:00 p.m. KST. Bithumb independently confirmed the three delistings Friday.

All open orders will be canceled when support ends. Upbit will keep withdrawals available until Oct. 14. Bithumb separately announced the same Sept. 14 trading cutoff and Oct. 14 withdrawal deadline for all three assets, confirming that the action is not limited to one Korean exchange.

Upbit says the warning issues were not resolved

Upbit first placed STORJ under a trading warning on July 28. The exchange said its review considered disclosure of important information, the substance and sustainability of the project and its actual progress. Deposits were suspended when the warning was issued, and Upbit said it would decide during the Aug. 10 to Aug. 14 review window whether to extend, remove or escalate the designation.

The warning came two days after Storj Labs filed for Chapter 11 protection in the U.S. Bankruptcy Court for the Northern District of West Virginia. Storj said case No. 5:26-bk-00512 was intended to address legacy liabilities while keeping its decentralized storage business operating. Upbit did not say the bankruptcy filing by itself caused the delisting.

As crypto.news reported, Storj Labs filed for Chapter 11 while keeping its cloud services active. The company has also floated a possible ownership mechanism for STORJ holders in a reorganized business. That remains a proposal rather than a confirmed recovery for token holders, and any arrangement would depend on the bankruptcy process.

JASMY and TT failed to clear later reviews

Upbit designated JASMY and TT as warning assets on July 31 and suspended deposits. For JASMY, the exchange cited shortcomings in disclosures about matters capable of materially affecting the asset, alongside concerns involving business substance, sustainability and actual progress.

For TT, Upbit also raised questions about the project’s operations, while the original review covered issuance and circulation related concerns. Bithumb’s final notice said the warning reasons for STORJ, JASMY and TT had not been resolved. No new Jasmy response addressing Friday’s delisting was identified in the official materials reviewed.

The decisions come as Korean exchanges continue applying post-listing reviews to supported assets. In related coverage, Upbit moved to remove NKN after its warning review, another case in which a caution designation preceded termination of trading support.

ThunderCore disputes the exchanges’ assessment

ThunderCore has disputed the characterization behind TT’s warning designation. In an Aug. 1 statement, the project said plans involving TT had been publicly communicated and that changes to issuance, network parameters and token economics had been disclosed to Upbit and Bithumb. It maintained that “the network, the issuance mechanism, and the project’s operational condition remain normal.” That is ThunderCore’s position and conflicts with the exchanges’ decision to maintain their concerns.

ThunderCore also updated its circulating supply estimates in June after a governance proposal allowed block issuance to be adjusted from zero to 135 TT by vote. Its official blog lists a new Aug. 14 update concerning Upbit and Bithumb trading support, but both exchanges still proceeded with delisting.

What happens next for STORJ, JASMY and TT

Upbit users can trade the six affected pairs until 3:00 p.m. KST on Sept. 14. Any unfilled buy or sell orders will then be canceled. Withdrawals remain supported until 3:00 p.m. KST on Oct. 14, giving users another month to move assets to compatible wallets or other venues.

Bithumb set the same dates and warned that withdrawal processing may require advance address registration. The exchange also said technical support can become limited after the withdrawal window closes. Delisting does not disable the underlying networks or tokens, but it removes the affected trading markets from two major South Korean venues.

South Korean lawmaker Park Soo-young has called for the government to withdraw its planned 22% tax on virtual asset gains before the levy takes effect on Jan. 1, 2027, arguing that the policy unfairly targets roughly 13 million crypto users while investment taxes on domestic stocks have been scrapped.

According to Digital Asset, the People Power Party lawmaker criticized the tax plan on his YouTube channel, “Park Soo-young’s Economy TV,” on Aug. 13, describing it as a punitive policy that could push more Korean capital toward overseas cryptocurrency markets.

“I hope this punitive tax plan that holds 13 million digital asset users hostage will be withdrawn immediately,” Park said.

The lawmaker compared the treatment of crypto investors with South Korea’s decision to abolish the financial investment income tax, which would have applied to investment income from financial products including stocks.

Park argued that removing the investment tax while retaining a separate levy on virtual assets amounts to telling investors they could face a “tax bomb” if they choose not to invest in the domestic stock market.

South Korea crypto tax faces renewed opposition

Under South Korea’s current Income Tax Act, profits from the transfer or lending of virtual assets will be classified as other income from Jan. 1, 2027. Investors will receive an annual deduction of 2.5 million won, with gains above that amount taxed at 20%.

Once the 2% local income tax is included, the effective rate reaches 22%.

The 2.5 million won deduction is the same basic deduction applied to capital gains from overseas stocks, while South Korea no longer plans to introduce the financial investment income tax that would have covered certain domestic financial investments.

The government has maintained that implementation will proceed next year. In May, Moon Kyung-ho, director of the Ministry of Economy and Finance’s income tax division, publicly confirmed that authorities were preparing to introduce the tax on schedule, as crypto.news previously reported.

The National Tax Service has also been preparing implementation guidance with South Korea’s five major crypto exchanges: Upbit operator Dunamu, Bithumb, Coinone, Korbit and Gopax. The guidance is expected in 2026, while the first full filing period would come in May 2028 for income earned during 2027.

Political opposition has continued despite those preparations. The People Power Party introduced legislation in March seeking to amend the Income Tax Act and abolish the crypto tax before it takes effect. The tax has already been delayed three times following years of disagreement over when and how digital asset gains should be taxed.

Originally introduced in 2020, the regime was scheduled to take effect earlier, but lawmakers repeatedly pushed back its implementation. The latest postponement moved the start date from 2025 to 2027.

Park warns capital could move overseas

Park also challenged the idea that taxing crypto could encourage investors to redirect their money into Korean equities.

“People will not invest in domestic stocks just because of this,” he said, adding that the policy could instead accelerate the movement of Korean wealth overseas.

To support his argument, Park cited data showing that roughly 124 trillion won flowed into overseas digital asset exchanges between January and September last year.

Capital moving from Korean platforms to foreign exchanges and private wallets has already drawn regulatory attention. Financial Services Commission data released in March showed that South Korean exchanges recorded 90 trillion won, or roughly $60 billion, of crypto outflows during the second half of 2025, up 14% from 78.9 trillion won in the first half. The regulator attributed part of the activity to cross-border arbitrage and similar trading, while overseas crypto outflows have remained under scrutiny.

South Korea has also moved to place more cross-border digital asset activity within its foreign-exchange framework. Amendments to the Foreign Exchange Transactions Act created a virtual asset transfer service category and require companies handling qualifying overseas crypto transfers to register with the finance minister. The cross-border transfer rules cover businesses moving virtual assets between South Korea and foreign countries through sales, purchases or exchanges.

Loss carryforwards become another tax dispute

Park separately criticized the treatment of losses under the planned system, arguing that the government intends to collect tax when investors make profits without providing comparable treatment when cryptocurrency prices fall.

“The losses suffered from a crypto crash cannot even be carried forward, yet they are already putting a spoon into the profits,” Park said.

His comments follow other objections focused on how the virtual asset tax compares with the treatment of other investments.

A public petition seeking complete repeal of the levy crossed the 50,000-signature threshold in May, automatically sending the proposal to a National Assembly committee for review. The petition argued that imposing a 22% tax on crypto profits while financial investment income from stocks and bonds remains exempt creates unequal treatment between asset classes. The tax repeal petition also raised concerns about investor protections and the treatment of a market where large price movements can rapidly alter gains and losses.

The disagreement has left the Jan. 1 deadline dependent on whether lawmakers change the Income Tax Act before implementation. The government has said it intends to proceed under the existing law, while no clear opposition to implementation has emerged from the ruling party.

The People Power Party continues to seek either the abolition or another postponement of the tax. Its March bill proposed removing the levy entirely, while Park’s latest comments called for the government to withdraw the planned tax before its scheduled 2027 implementation.

Public Bitcoin mining companies are shrinking their realized hashrate faster than the broader Bitcoin network, according to a BlocksBridge Consulting analysis reported in Miner Weekly. The data suggests some operators are redeploying electricity, facilities, and hardware toward data-center and high-performance computing (HPC) workloads rather than expanding crypto mining.

In the latest Miner Weekly update, BlocksBridge Consulting tracked a cohort of public Bitcoin miners and found realized hashrate declined from 368.3 EH/s in Q4 2025 to 319 EH/s in Q2 2026—down 13.4%. When excluding Bitdeer, the contraction was larger: realized hashrate fell 21.2% from 324.6 EH/s to 255.9 EH/s over the same six-month span. Bitdeer, by contrast, increased its realized hashrate by 44% to 63 EH/s.

Key takeaways

• Public miners’ realized hashrate dropped 13.4% between Q4 2025 and Q2 2026, while the Bitcoin network’s average hashrate fell 10.6%.
• Excluding Bitdeer, the public-miner cohort’s realized hashrate declined 21.2%, indicating a stronger pullback from many large listed operators.
• BlocksBridge’s numbers point to a shift in resource allocation—electricity and compute capacity moving toward data centers and HPC.
• Core Scientific and TeraWulf are increasingly earning from non-mining activities, with colocation and HPC leases outpacing mining revenue in recent quarters.
• The change is tied to the post-China mining boom unwinding and the parallel rise in AI infrastructure demand since 2022.

Public miners retreat faster than the network

The gap between industry-wide network trends and what’s happening at listed miners matters because realized hashrate at major operators often reflects strategic choices: whether to run machines at full tilt, pause less efficient operations, or repurpose capacity. While Bitcoin’s aggregate hashrate fell 10.6% over the period covered by BlocksBridge, the cohort of public miners declined more sharply, dropping 13.4% from 368.3 EH/s to 319 EH/s.

The divergence becomes more pronounced when looking at the composition of the cohort. Without Bitdeer, realized hashrate fell 21.2% from 324.6 EH/s to 255.9 EH/s. That larger contraction suggests that many public operators were not merely adjusting output with the broader cycle—they were reducing mining footprint relative to peers, or limiting the use of their most readily available power for Bitcoin blocks.

Bitdeer’s opposite direction reinforces that the trend isn’t uniform across all listed miners. BlocksBridge reported Bitdeer’s realized hashrate rose 44% to 63 EH/s during the same six-month window, implying that at least one major operator continued to treat mining expansion as economically viable while others pulled back.

Why the shift is happening: mining economics versus AI-driven demand

BlocksBridge framed the pullback as the unwinding of the expansion cycle that followed China’s 2021 Bitcoin mining ban. That regulatory shock triggered one of the steeper declines in network hashrate, followed by a recovery as miners relocated overseas and re-established operations in other jurisdictions.

In North America, the migration and subsequent capital deployment helped public miners expand. Many raised funds and acquired new power sites to increase mining capacity. But with one halving cycle later, the cost-and-demand equation has changed. According to BlocksBridge’s framing, weaker mining profitability combined with surging demand for AI infrastructure since 2022 has encouraged several public miners to redirect sites and power capacity away from pure Bitcoin mining.

That matters for investors because “miner earnings” are no longer tightly linked to one variable—Bitcoin network economics—at least not for the largest operators pursuing diversification. Instead, revenue becomes increasingly tied to how much capacity can be monetized through data-center services, colocation, and HPC leasing, where customer demand is driven by the AI build-out rather than solely by block rewards and transaction fees.

Non-mining revenue is moving to the center of the story

BlocksBridge’s report also points to another signal of this transition: the share of revenue coming from activities other than mining. In the examples highlighted, Core Scientific generated $136.7 million in colocation revenue in Q2, compared with $27.5 million from Bitcoin mining. TeraWulf, in turn, reported $31.9 million in HPC lease revenue versus $12.8 million from mining.

The implication is straightforward: for these operators, facilities and power are being monetized as compute infrastructure for third parties, not only as a way to mine BTC. Cointelegraph previously covered these financial dynamics in separate stories—such as Core Scientific’s revenue mix and TeraWulf’s data-center and HPC expansion—both of which are consistent with the broader BlocksBridge narrative about miners evolving into infrastructure providers.

For context within the cohort, the article notes that Riot Platforms and Bitdeer remained earlier in the transition, with Bitcoin mining still accounting for the majority of revenue in their latest reported quarter. That contrast is important: it helps explain why the overall public-miner hashrate decline isn’t identical to a universal exit from mining. Some companies are deepening the AI/data-center pivot, while others are still heavily dependent on mining cash flows.

What to watch next

The next phase likely hinges on whether the non-mining expansion continues to offset mining profitability pressure—and on how quickly remaining public miners decide to scale down operations in less efficient locations. Investors should watch both realized hashrate trends across cohorts and disclosures about revenue mix, because the sector’s direction appears to be determined as much by electricity monetization strategy as by Bitcoin’s underlying network changes.