MARA, CIFR, and WULF Aren't the Same Story Anymore

The headlines bundle Marathon Digital, Cipher Digital, and TeraWulf into one declining basket: bitcoin miner stocks dropping as price targets get cut. Morgan Stanley slashed MARA's target to $5.50 in mid-July, trimmed CIFR's to $47 from $48.50, and Keefe, Bruyette & Woods cut WULF's to $30 from $33 last month. Bitcoin itself is hovering around $65,000 — roughly 44% below the $126,000 peak late last year. On the surface, it looks like one story playing out across three tickers.

But the operating evidence splits these companies into two very different stories. Two of them are building data-center businesses that already have leases and partners. One is still running a cash-incinerating mining operation that lost $611 million in a single quarter last week. The market hasn't finished making that distinction.

The old story — pure mining — is a losing proposition

Bitcoin mining economics right now are structurally unprofitable for anyone running older equipment. The hashprice (the revenue a miner earns per unit of computing power per day) sits around $29 per petahash. At the current Bitcoin price, the estimated cost to mine one Bitcoin is roughly $84,000. That means each block a miner finds costs more to produce than the coin is worth. Mining difficulty recently fell 10% in June as operators shut down hardware that could no longer justify its electricity bill.

MARA is the company still anchored to this reality. Q2 2026 results, released August 6, showed revenue of $175 million — missing the $208 million consensus by 16%. The adjusted EBITDA swing was brutal: from $1.2 billion in the year-earlier quarter to negative $361 million. The net loss hit $611 million. Roughly $343 million of that came from mark-to-market losses on digital assets as Bitcoin declined, but the operating picture didn't help either.

Over the trailing twelve months, MARA burned through $1.34 billion in free cash flow. The balance sheet showed $421 million in unrestricted cash at quarter end against $2.27 billion in total debt. Management is selling Bitcoin to stay afloat — total holdings fell 29% year-over-year to 35,577 BTC. On August 6 the stock closed at $10.65; today it's trading around $10.15. Morgan Stanley's $5.50 price target is the lowest of twelve analysts covering the name.

This is not a company where the market has become bored and the numbers are improving underneath. This is a company where the numbers got worse and the market is finally reflecting it.

The pivot story — not finished, but real

CIFR and WULF look different on paper. Both are transitioning away from mining toward contracted data-center infrastructure, and the evidence isn't just slide-deck language.

WULF has four secured lease agreements with hyperscalers and AI tenants. KBW cut its target to $30 last month, but that still implies roughly 82% upside from today's $16.96. The consensus across fifteen analysts sits at $34.29. Needham, Citigroup, Chardan, and B. Riley all have buy ratings. The KBW cut came on the same day the broader neocloud sector sold off — APLD was down 33% over the prior month, IREN down 25%. The market is rotating away from pipeline promises toward executed contracts. That's a discipline shift, not a thesis break.

CIFR is even more stressed on the tape — down 21% over five days to $17.56 today. But look at the underlying structure. The company is the only miner-turned-infrastructure operator partnered with Amazon Web Services. KBW flagged unallocated ERCOT "Batch Zero" capacity (a tier of new power capacity in Texas reserved for large industrial loads) as a potential major value unlock. Morgan Stanley named CIFR a preferred ERCOT large-load beneficiary. The company holds $832 million in cash against $6.9 billion in total debt — a loaded balance sheet, no question, but the debt is financing a buildout, not a mining operation.

Q2 2026 revenue for CIFR missed by roughly $7 million, but that quarter's miss happened as the company was actively winding down mining and ramping infrastructure. Revenue was still up 20.3% year-over-year. The AI data-center pivot is a different business model with different timelines than bitcoin mining.

Why the three-name bundle persists

The reason all three stocks trade in the same fear loop is simple: they all started as bitcoin miners. Investors see the name, see the hash rate, and anchor to the old risk profile. The difficulty curve, the halving cycle, the $65,000 price — those are the numbers that still dominate the narrative.

MARA's management is trying to build a pivot too. The Q2 slides describe a "Digital Infrastructure Triad" targeting 4.8 gigawatts of total capacity, with a 2-gigawatt site planned in Matagorda County, Texas, and a pending acquisition of Long Ridge Energy adding another gigawatt. The timeline stretches into 2027 and 2028, and the execution risk is material. But the Q2 result was a $611 million loss and the free cash flow burn is already $1.34 billion over twelve months. A company can't pivot on aspiration alone when the operating numbers are this far in the hole.

CIFR and WULF don't have the same luxury of scale, but their lease books and partnerships represent contracted cash flows that don't depend on Bitcoin staying above the $84,000 cost line. That's the structural difference the market is still catching up to.

What changes the setup

For CIFR and WULF, the next earnings seasons carry the proof. WULF reported Q2 on August 5th; CIFR's Q2 results will come later this month. What matters isn't whether mining revenue holds — it doesn't, and that's not the business anymore. It's whether infrastructure revenue grows, lease margins print in line with guidance, and capex converts into contracted throughput. ERCOT Batch Zero auctions for CIFR and AI tenant ramp timelines for WULF are the two concrete checkpoints that would make the market stop treating these as mining proxies.

For MARA, the $5.50 Morgan Stanley target sits below half the current price. That's not a misreading — it's a reflection of what the cash-flow path says when you strip out the infrastructure slides. The company has to prove that Q3's projected $0.43 EPS hold up, that the debt service is manageable on a shrinking cash pile, and that the Texas buildout can attract tenants before the debt matures.

The tripwire for CIFR and WULF is the same: a quarter where infrastructure revenue shrinks or lease margins collapse below what the debt service requires. If that happens, the pivot story is broken and the valuations will compress to mining multiples fast.

The tripwire for MARA is simpler: another quarter of deep negative free cash flow without a material non-mining revenue line. At that point, there's nothing left to rerate toward.

The market is still pricing the old story across all three names. But the cash-flow evidence already says they aren't the same company. The question over the next 12 months isn't whether Bitcoin stays above $65,000 — it's which of these businesses can generate free cash flow without it.