Bitcoin Miner Stocks Are Still Priced Like Miners — When They're Not Anymore

MARA, CIFR, and WULF all sold off this week. Morgan Stanley slashed MARA's price target to $5.50 on July 16th, citing deep losses and negative margins. Keefe Bruyette downgraded CORZ to Market Perform and trimmed WULF's target. The headline narrative is simple: Wall Street lost patience, and the miners crashed.

The problem is that the market is scoring these companies on a game they're no longer playing.

The story that carries valuation isn't hashprice. It's the HPC lease pipeline.

Bitcoin is sitting around $64,200 after recovering above $65,000 in recent weeks. Hashprice — the revenue earned per unit of mining hashrate — is near $29/PH/s/day, a post-halving low that has left pure-play mining deeply unprofitable for anyone running older-generation ASICs above $0.05/kWh in power costs. At these levels, the mining business is a cash-flow-negative grind.

That is the story Wall Street is reacting to. The Q2 losses are real. MARA posted a $611.3 million net loss on $174.9 million in revenue. Adjusted EBITDA was negative $360.9 million. CIFR reported $25 million in revenue against a $268 million GAAP net loss, driven partly by a $150.5 million non-cash warrant remeasurement. WULF posted a large GAAP loss as it ramped HPC capacity.

But looking at the mining losses without adjusting for what the capital is actually doing is like judging a construction company by its demolition costs.

Here is the operating reality:

1. HPC revenue is flipping from aspirational to dominant. WULF's Q2 revenue was $46 million, with $31.9 million — approximately 71% — coming from HPC lease revenue. That is a 52% quarter-over-quarter increase. At this run rate, HPC revenue alone annualizes to roughly $128 million. The company controls 60 MW of operational HPC capacity at Lake Mariner under long-term lease to Core42, with CB-3 completion and CB-4/CB-5 delivery targeting 2026. Management is guiding for 250–500 MW of new contracted capacity annually. That is a data center operator's profile, not a miner's.

2. The HPC lease pipeline is contracted, not speculative. CIFR's portfolio totals 5.3 gigawatts across 11 sites, with a 4.4 GW future pipeline. First rental payments at Barber Lake and Black Pearl are expected to begin in the second half of 2026. The company priced an $810 million bond offering at a 6% coupon for the Stingray project — using non-recourse, project-level debt that amortizes during investment-grade counterparty lease terms. These are infrastructure financings, not mining capex.

3. Mining efficiency is actually improving even as the mining headwinds bite. MARA reached 70.3 exahash per second of energized hashrate, up 22% year-over-year. Operating cost per petahash improved 4% to $27.70 from $28.70. Bitcoin production increased 3% to 2,422 BTC. The operational machine is getting cheaper and bigger even as Bitcoin revenue per hash falls. That matters because the hashrate fleet becomes free capacity once the power contracts are re-purposed for HPC tenants.

4. The balance sheet anchor is the Bitcoin treasury, not the mining cash flow. MARA ended Q2 with 35,577 BTC on its balance sheet — worth approximately $2.3 billion at current prices. The company holds $2.5 billion in combined cash and Bitcoin. Post-quarter, it pledged 18,750 BTC as collateral for two new Bitcoin-backed credit facilities providing $600 million in incremental borrowing capacity. At a $5.3 billion market cap, MARA's non-BTC business — including a 4.8 GW power portfolio and the planned $1.5 billion Long Ridge acquisition — is being valued at roughly $3 billion. That is $0.63 billion per gigawatt of power pipeline. For comparison, purpose-built colocation operators trade at multiples far above that.

5. The sector is bifurcating and the market hasn't reclassified yet. The mining industry is splitting into two valuation buckets. Companies transitioning to infrastructure — WULF, CORZ, CIFR, Hut 8 — are being evaluated separately from pure-play miners — MARA, CleanSpark, Riot, Hive. Over $70 billion in AI/HPC contracts have been announced across the listed miner sector. Listed miners could derive up to 70% of revenue from AI/HPC by end-2026, up from roughly 30% earlier this year. But the stocks are still being traded like Bitcoin proxies.

Keefe Bruyette captured the dynamic on July 27th when it revised its miner coverage ahead of Q2 earnings. The firm raised CIFR's price target to $32 from $27, kept an Outperform rating on WULF despite trimming the target to $30 from $33, and explicitly described the pullback as a "pipeline valuation reset" rather than weakening fundamentals. The lone downgrade — CORZ to Market Perform on a cut from $28 to $25 — still came with a composite AInvest aggregate signal that labels the stock a Buy. The analyst consensus is mixed, not uniformly bearish.

What breaks the setup: the HPC transition has to actually deliver rental revenue. Barber Lake and Black Pearl need to hit H2 2026 timelines for CIFR. WULF's Core42 ramp and CB-3/CB-4 energization need to stay on schedule. If HPC leases stall and Bitcoin drops below $50,000 again, the mining operation becomes a genuine drag rather than a bridge. The capital-intensive transition also means debt loads are heavy — WULF carries $5.7 billion, CIFR $1.7 billion — and those obligations don't disappear if tenant onboarding slows.

The stocks may still need to find a bottom before they attract patient capital. But the forward math is already far more attractive than the panic narrative suggests. MARA at $10, trading with roughly $2.3 billion of Bitcoin on its balance sheet, is not a distressed miner anymore. It's a power infrastructure company with a crypto hedge. CIFR at $17 with 5.3 GW of contracted pipeline and H2 rental commencements starting is an HPC platform being punished for its mining headline. WULF at $17, with HPC already generating 71% of its revenue, is priced like a company whose business model hasn't already flipped.

At roughly $3 billion for MARA's non-BTC operations covering a 4.8 GW power pipeline, the stock is trading at less than $0.65 billion per gigawatt. That's the number that matters. The mining losses are the bridge cost. The HPC leases are the destination.