
MARA, Riot, Cango and Bitdeer have reduced holdings while executives emphasize power strategy and HPC-style revenue.
Several major miners have already sold meaningful Bitcoin and reduced leverage with the next halving expected in April 2028. Executives are framing the run-up as a tougher, more power-constrained cycle where diversified revenue and long-term energy procurement matter as much as hashrate.
Miner treasuries are getting lighter well before the usual pre-halving window. MARA Holdings sold more than 15,000 Bitcoin in March to reduce leverage. Riot Platforms sold over 3,700 BTC in the first quarter. Cango sold 2,000 BTC to pay down Bitcoin-backed debt.
Bitdeer took the cleanest line: it said its Bitcoin holdings had fallen to zero as of Feb. 20.
The common thread is balance-sheet resilience. Instead of leaning on a price-led margin expansion to carry the business, multiple operators are monetizing reserves and reducing leverage early. That behavior reads less like a tactical treasury choice and more like a sector-wide acknowledgment that the next issuance cut will arrive with less room for error.
The subsidy math is straightforward and brutal. In April 2024, the block subsidy fell from 6.25 BTC to 3.125 BTC, with Bitcoin trading around $63,000 at the time, according to CoinGecko.
The next halving is expected in April 2028, cutting the subsidy from 3.125 BTC to 1.5625 BTC per block. In BTC terms, that is another 50% reduction in the fixed issuance component of miner revenue, separate from transaction fees.
Executives are already framing the implication as “higher input costs for half the new coins.” If costs are sticky and network competition stays high, miners cannot assume the same operating leverage they enjoyed when price strength did more of the work.
Industry messaging around the 2028 cycle is converging on tighter competition and more selective capital. The setup is being described as tougher than 2024 due to record hashrate, higher energy prices, and tighter power markets, with a CoinWarz chart referenced for the hashrate claim but without a specific level or date.
Juliet Ye, head of communications at Cango, said the 2028 halving arrives in “an environment that looks almost nothing like 2024,” citing a widening efficiency gap “forcing real decisions around fleet upgrades.” She also warned, “There is less room in the middle now,” adding, “Operators with scale and diversification will be fine. Those without will find the next halving very difficult.”
Mark Zalan, CEO of GoMining, put the shift in capital terms, saying “capital discipline now matters more than hashrate maximalism,” with new deployments needing to clear tougher return thresholds.
Alejandro de la Torre, co-founder and CEO of Stratum V2 pool DMND, argued the cycle mechanics remain familiar: “There is actually very little fundamental difference between this mining cycle and previous ones,” and “The same dynamics repeat.” He expects hotspots to peak and realign as “no region keeps dominance for long,” potentially opening room for via new energy partnerships.
Regulation is also being positioned as part of the financing narrative into 2028, with references to more specific US custody and banking access rules, the EU’s MiCA framework, and new ETFs, , and settlement rails out of Hong Kong.
The strategic pivot being described is away from pure block-reward dependence and toward power and data-center style monetization. Zalan called mining a “thinner business than it used to be,” and pointed to incremental revenue from curtailment, grid services, and heat reuse.
Ye described a multi-use facility model: “The facilities that will matter in five years are the ones that can do more than one thing,” using mining to fill capacity while positioning sites to toggle between AI workloads and hashpower. That framing aligns with the broader emphasis on long-term power procurement and diversified revenue, not just fleet growth.
The regulatory angle is being used to support that buildout. Zalan argued “capital moves faster when those rules are clear and usable,” tying clearer regimes to how miners finance expansion and how institutions position ahead of the next issuance cut.
I read the recent treasury drawdowns as a liquidity and risk-management tell, not a one-off. When MARA sells more than 15,000 BTC to reduce leverage, Riot sells over 3,700 BTC in a quarter, Cango sells 2,000 BTC to pay down BTC-backed debt, and Bitdeer runs holdings to zero, the sector is signaling it wants less dependence on a pre-halving price run-up to stay solvent.
The threshold that matters is whether this becomes a persistent supply overhang from public miners or a short, front-loaded reset. The real test is whether miners can lock in long-term power procurement and convert the “power-and-HPC” narrative into disclosed curtailment/grid-service agreements and AI/HPC workload contracts, while network competition keeps rising into 2027 and the April 2028 subsidy cut to 1.5625 BTC approaches. If those non-mining revenue lines show up in filings and guidance, the setup starts to look structural rather than narrative-driven, and that is what would make this shift matter for BTC supply dynamics and equity dispersion across miners.