Two years after the April 2024 halving cut block rewards from 6.25 BTC to 3.125 BTC, the Bitcoin mining industry is deep into a structural repricing. Hashprice — the core measure of miner revenue per unit of computational power — has fallen to approximately $0.06/TH/s per day, half its pre-halvin...
"Selling bitcoin could become a recurring element of our treasury strategy." — MARA Holdings, Q1 2026 Earnings Report
Two years after the April 2024 halving cut block rewards from 6.25 BTC to 3.125 BTC, the Bitcoin mining industry is deep into a structural repricing. Hashprice — the core measure of miner revenue per unit of computational power — has fallen to approximately $0.06/TH/s per day, half its pre-halving level of $0.12. The three largest publicly traded miners — MARA Holdings, Riot Platforms, and CleanSpark — reported a combined $2.05 billion in net losses for Q1 2026 alone. Network hashrate briefly crossed 1 ZH/s (zetahash per second) in January 2026 before difficulty adjustments and energy disruptions pulled it back to the 800-900 EH/s range.
The sector's response has split into three distinct strategies: treasury liquidation to service debt, aggressive pivots toward AI and high-performance computing (HPC) infrastructure, and pure-play mining consolidation through capacity expansion. Miners sold over 19,000 BTC ($1.4 billion) in Q1 2026 alone. Meanwhile, CoinShares estimates cumulative AI/HPC contracts announced across listed miners exceed $70 billion, with some operators projected to derive up to 70% of revenue from non-mining compute by year-end. The mining industry is no longer a single-variable bet on Bitcoin's price. It is a multi-track infrastructure play where survival depends on energy costs, hardware efficiency, and revenue diversification.
Hashprice — daily revenue per terahash of deployed computational power — is the single most important metric for mining economics. It captures the combined effect of Bitcoin price, network difficulty, block rewards, and transaction fees into one number.
As of late May 2026, hashprice sits at approximately $0.055–$0.065/TH/s per day, or roughly $36.46/PH/s. This represents a decline of approximately 50% from pre-halving levels in early 2024 when hashprice traded near $0.12/TH/s per day.
The math is straightforward. The halving cut the block subsidy from 6.25 BTC to 3.125 BTC per block. Transaction fees contribute an additional 0.2–0.5 BTC per block under normal conditions, constituting approximately 10–15% of total miner revenue (rising to 20–30% during congestion events). But Bitcoin's price has not doubled to compensate for the halved reward. With BTC trading near $67,000 as of early June 2026 — well below the $140,000+ that would be needed to restore pre-halving hashprice levels — miners face a structurally compressed revenue environment.
The average production cost per Bitcoin for public miners has risen to approximately $37,856, according to Spark Research, up from roughly $16,800 before the halving. At current prices, margins exist but remain thin, and any sustained price decline below $50,000 would push a significant portion of the network into unprofitable territory.
Despite compressed economics, miners continued deploying hardware through late 2025 and into early 2026, pushing the network hashrate to unprecedented levels. In mid-to-late January 2026, the 7-day moving average hashrate reached 1.05–1.13 ZH/s, briefly crossing the symbolic 1 zetahash threshold for the first time in Bitcoin's history.
The milestone proved short-lived. A severe winter storm impacting the ERCOT grid in Texas — where a substantial share of US mining capacity is located — forced widespread curtailment. Network hashrate fell approximately 12% from November 2025 peaks, with intraday drops of 30–40% during the most intense days of the weather event.
Mining difficulty has responded with notable volatility through 2026:
| Period | Difficulty | Change | |--------|-----------|--------| | Late 2025 Peak | 148.2T | — | | February 2026 | 144.4T | +15% adjustment | | March 2026 | 133.79T | −7.76% | | April 2026 | 138.97T | +3.9% | | June 13, 2026 (est.) | 125.94T | −9.4% (projected) |
The projected June 13 downward adjustment to 125.94T would represent a notable reprieve for remaining miners, effectively lowering the computational cost of producing each block. Difficulty has increased approximately 5% per month on average through 2026, according to CoinWarz data, but the swings suggest the network is in a period of equilibrium-seeking rather than steady expansion.
The first quarter of 2026 produced uniformly negative financial results across the three largest publicly traded mining companies.
MARA Holdings reported a net loss of $1.3 billion ($3.31/diluted share), compared to a loss of $533.4 million ($1.78/diluted share) in Q1 2025 — a 144% increase in losses year-over-year.
CleanSpark posted a net loss of $378.3 million for its fiscal Q2 ending March 31, 2026, a 173% increase from the $138.8 million loss in the prior-year period. Revenue fell 24.9% year-over-year to $136.4 million, missing analyst expectations of $152.32 million. CleanSpark shares fell over 9% in after-hours trading following the report.
Core Scientific reported digital asset self-mining revenue of $30.1 million in Q1 2026, down from $67.2 million in Q1 2025 — a 55% decline driven by a 45% decrease in Bitcoin mined and an 18% decrease in average Bitcoin price. The company reported a net loss of $347.2 million, which included $266.5 million in non-cash impairment charges.
Combined Q1 2026 net losses across these three companies alone totaled approximately $2.05 billion.
Facing mounting losses and debt obligations, major miners reversed their long-standing "HODL" strategy and began aggressively liquidating Bitcoin reserves.
MARA sold 15,133 BTC between March 4–25, 2026, generating approximately $1.1 billion at an average price of $72,689 per coin. Proceeds funded the repurchase of convertible senior notes: $367.5 million in 2030 notes repurchased for $322.9 million in cash, and $633.4 million in 2031 notes repurchased for $589.9 million. MARA simultaneously cut 15% of its workforce, according to Unchained Crypto.
Riot Platforms sold 3,778 BTC in Q1 2026, generating $289.5 million at an average sale price of $76,626 per coin. Riot sold approximately 2.5 times more Bitcoin than it mined during the quarter. As of quarter-end, the company held 15,680 BTC, with 5,802 BTC pledged as collateral against existing obligations.
Nakamoto Holdings and other smaller public miners contributed additional sales. In total, Riot, MARA, and Nakamoto collectively sold over 19,000 BTC in Q1 2026, according to CryptoPotato reporting. According to CoinDesk, the broader Bitcoin treasury trend among public companies and even sovereign holders (including Bhutan) showed signs of unwinding, with multiple entities exiting positions simultaneously.
Rising energy costs — exacerbated by geopolitical disruption affecting oil prices — further squeezed margins across the sector, accelerating the liquidation timeline.
The most consequential strategic shift in the mining sector is the reallocation of power infrastructure and data center capacity from Bitcoin mining to AI and HPC workloads.
According to S&P Global Market Intelligence, a growing number of public miners have repurposed portions of their infrastructure or announced plans to do so, including Core Scientific, MARA Holdings, Hut 8, Riot Platforms, TeraWulf, and IREN. The economics are compelling: AI data center hosting typically generates higher revenue per kilowatt-hour than Bitcoin mining, with more stable, long-term contract structures compared to the volatile block-reward model.
CoinShares estimates that cumulative AI/HPC contracts announced across listed miners exceed $70 billion. The research firm projects that some operators could derive up to 70% of their revenue from AI-related compute services by end of 2026, up from approximately 30% at the time of publication.
Core Scientific sold $175 million worth of Bitcoin (1,992 BTC) in March 2026 specifically to fund its operational transition toward AI infrastructure. The company's Q1 2026 filing reflects this pivot: while self-mining revenue halved, gross profit actually increased to $30.1 million from $8.2 million a year earlier, suggesting the AI business is already contributing meaningfully to the bottom line.
Stock markets have noticed. All ten of the largest publicly traded mining stocks posted positive year-to-date returns through early 2026, with the most aggressive AI-pivot companies leading: TeraWulf (+85%), Hut 8 (+67%), and Riot Platforms (+46%), according to CryptoBreaking.
Bitcoin mining now consumes an estimated 155–172 TWh annually, according to the Digiconomist Bitcoin Energy Consumption Index and the Cambridge Centre for Alternative Finance (CCAF). This represents approximately 0.5% of global electricity generation, roughly equivalent to the annual consumption of Poland. In the United States specifically, mining consumes approximately 132.6 GWh per day, or about 1.1% of total daily US electricity demand.
Despite a roughly 35% year-over-year increase in hashrate, energy consumption grew by only an estimated 10–15%, reflecting significant hardware efficiency gains. Current-generation ASICs such as the Antminer S23 Hydro achieve approximately 9.5 J/TH, a substantial improvement from prior generations.
Sustainable energy sources now power 56.7% of mining operations, according to industry self-reporting:
| Source | Share of Sustainable Mix | |--------|------------------------| | Hydropower | 42.6% | | Wind | 15.4% | | Nuclear | 9.8% | | Solar | 3.2% |
Marathon Digital reported that over 70% of its energy consumption in Q1 2026 came from carbon-free sources. Carbon-neutral pledges now cover 52% of major mining firms targeting net-zero by 2030.
The profitability survival threshold has narrowed considerably. Miners operating with electricity costs below $0.07/kWh and hardware rated under 15 J/TH remain profitable at current hashprice levels. Those with retail electricity rates above $0.10/kWh face slim or negative margins. The estimated next difficulty adjustment on June 13, 2026 — projected to decrease by approximately 9.4% — would provide marginal relief, but does not alter the structural compression.
The divergence between mining fundamentals (losses, treasury drawdowns) and stock performance (broad year-to-date gains) illustrates a market thesis: investors are pricing in the AI infrastructure optionality, not pure mining economics.
CleanSpark, which has maintained a pure-play mining strategy — adding 585 MW of new mining capacity — saw its stock decline over 9% following earnings. The market's message is clear: pure-play mining, absent revenue diversification, carries a valuation discount.
By contrast, miners with credible AI/HPC pipelines have outperformed Bitcoin itself on a year-to-date basis, reflecting a re-rating of these companies from "crypto miners" to "power infrastructure operators."
The Bitcoin mining industry two years post-halving is no longer recognizable as the single-variable business it was in prior cycles. The April 2024 halving created a structural revenue gap that Bitcoin's price has not closed. Miners have responded not with the mass capitulation that some expected, but with a three-way strategic divergence: treasury liquidation and debt management, infrastructure diversification into AI/HPC, and pure-play capacity expansion for those with cost advantages.
The data suggests the AI pivot is not a temporary hedge but a permanent restructuring of the mining business model. When CoinShares estimates $70 billion in cumulative AI/HPC contracts and operators like Core Scientific report improving gross profits despite halved mining revenue, the sector is effectively undergoing a reclassification from proof-of-work validators to power-and-compute infrastructure providers that happen to also mine Bitcoin.
The next 12 months will determine which strategy prevails. If Bitcoin's price remains range-bound below $75,000, further treasury liquidation and network consolidation are likely. If it rises, the miners who maintained hashrate capacity through the compression will benefit disproportionately. Either way, the era of mining as a pure proxy for Bitcoin exposure appears to be ending.