CIFR stock forecast: first HPC rent meets a $72M operating loss

Cipher Digital reported its first high-performance computing lease revenue on November 2, 2026, but its third-quarter filing also showed a $71.9 million operating loss. The company’s results put $5.852 million of data-center rent beside $20.849 million of remaining bitcoin-mining sales. The new business is producing revenue; the question for CIFR stock is how quickly it can become large enough to change the consolidated accounts.

That question became more urgent after U.S. trading opened. Google Finance’s CIFR quote showed $6.70 at 9:53:50 a.m. Eastern Standard Time on November 2, about 8% below the prior $7.29 close. This is a dated intraday observation, not a closing price or evidence of a single cause for the decline. The useful forecast is an operating threshold: keeping the rest of the quarter unchanged, how much additional rent would be needed to offset the reported operating loss?

The first HPC rent is real, but the scale is still small

Cipher’s quarterly income statement separates the two activities. HPC leasing produced $5.852 million in the three months ended September 30, versus zero a year earlier. Bitcoin mining brought in $20.849 million, down from $71.707 million in the prior-year quarter. Total revenue was $26.701 million, down from $71.707 million. The new rent was 21.9% of current-quarter sales; it has not yet replaced the $50.858 million year-over-year contraction in mining revenue.

The direct-cost line is the promising part of the new segment. Reported HPC lease cost of revenue was $0.726 million, leaving $5.126 million after that line, or 87.6% of HPC rent. This is a calculation from the company’s reported amounts, not a disclosed segment operating margin. It excludes company-level compensation, overhead, depreciation, financing and future construction costs. Treating that 87.6% as a permanent mature-project margin would be unjustified; the first delivery is a small, early sample.

The same income statement shows why a high direct margin has not yet resolved the equity question. Compensation and benefits were $38.555 million and general and administrative expense $26.359 million. Depreciation and amortization added $18.821 million. The company reported a $71.915 million GAAP operating loss and negative $10 million of adjusted EBITDA. Those are different measures, but both say the new rent has not yet reached a scale that carries the whole enterprise.

A reproducible threshold for the next phase

Here is the static-cost sensitivity. Divide the $71.915 million quarterly operating loss by the new segment’s calculated direct margin of 87.6%. That produces roughly $82.1 million of additional quarterly HPC leasing revenue needed to erase the Q3 operating loss if mining revenue, all other revenue and costs, depreciation, and the direct margin were frozen at Q3 levels. Add the $5.852 million already reported, and the implied quarterly rent level is about $88 million, or roughly 15 times the first quarter’s HPC rent. The calculation is an analytical hurdle, not company guidance or a prediction of when breakeven will occur.

That hurdle should be read with two cautions. First, additional data halls will bring project expenses and likely more depreciation, so a constant-cost model can make the path look easier than reality. Second, the mining operation is deliberately shrinking; a further fall in mining gross profit would increase the rent needed to compensate. The reverse could also happen: a lower overhead run rate or higher rent yield could improve the bridge. The next two quarterly statements, not the signed headline value of a lease, will show which forces dominate.

A nearer-term checkpoint is adjusted EBITDA. At Q3’s negative $10 million, a simple division by the 87.6% direct margin suggests about $11.4 million of incremental quarterly HPC rent would close that particular gap if every other adjusted-EBITDA component stayed fixed. That is a much smaller threshold than GAAP operating breakeven because depreciation and certain other items sit outside adjusted EBITDA. It does not mean $17.3 million of total quarterly rent would make Cipher profitable or cash-flow positive. Investors should ask management to reconcile each new project contribution to both GAAP and cash outcomes.

Why the reported net profit is a poor shortcut

Readers may see a positive Q3 net income figure and think the transition has already succeeded. The reported reconciliation shows $31.596 million of GAAP net income even though operations lost $71.915 million. A $127.180 million gain from the change in fair value of a warrant liability helped bridge that difference. Such a mark-to-market gain can reverse and does not represent rent collected from a customer. It is part of GAAP earnings, but it is not evidence that the data centers have covered their operating burden.

Interest also matters. Q3 interest expense was $58.994 million, against $32.684 million of interest income. The company issued $850 million of zero-coupon convertible notes due in 2032 and equitized most of a smaller 2030 note balance. The zero coupon limits cash interest on that specific issuance, but the consolidated interest line and project financing still need watching. Nor should a share-price forecast turn a large contracted-rent number into today’s equity value without accounting for debt, construction spend and the timing of deliveries.

The $9.1 billion lease is a schedule, not Q4 sales

Cipher said the expanded Barber Lake lease has approximately $9.1 billion of contracted revenue over a 20-year term, including about $5.2 billion added in the amendment. The amendment was disclosed in a September 24 filing; repeating the total today does not make the contract itself a new November 2 signing. Dividing $9.1 billion by 20 yields $455 million per year as a simple lifetime average, or $113.75 million per quarter. That average is not the contractual quarterly payment schedule, a 2027 revenue forecast or near-term free cash flow. Cipher says deliveries are phased to the tenant’s compute deployment, and the Q3 HPC revenue came from Black Pearl, a different campus.

The actual progression matters more than the headline total. At Black Pearl, rent began when capacity was delivered, and the company says additional halls there and at Barber Lake are scheduled to come online in phases through year-end. Each energized, accepted hall can change recognized rent. A delayed hall can postpone it. The next report needs to show the number of delivered halls or capacity, recognized HPC rent, project operating expense, and whether new debt or equity was needed to finish delivery. Without those pieces, a large multiyear agreement cannot be converted responsibly into a near-term stock price target.

Power rights and financing decide the 2027 branch

Cipher describes a roughly 5.9-gigawatt development pipeline. That is not the same thing as 5.9 gigawatts of energized, leased data centers. In the November 2 update, it reported conditional ERCOT classifications for 1 gigawatt at Colchis and an additional 1.6 gigawatts of studied load with provisional controllable-load designations. It has also begun lateral natural-gas pipeline development that it says could support up to 2.5 gigawatts of on-site generation. These are development steps with regulatory, construction, fuel-supply and financing conditions, not operating megawatts.

The company also identifies 270 megawatts potentially available in 2027 while it advances tenant discussions and design work. Available capacity is not a signed tenant. A bullish 2027 case requires timely power approvals, contracts, construction and rent commencement at attractive project economics. A base case would show steady rent growth but continued consolidated losses while the portfolio is built. A bearish case would see power or tenant delays combine with construction costs or financing pressure. These are conditional paths, not precise share-price forecasts; the disclosed inputs do not justify an honest dollar target today.

This makes dilution and capital structure part of the same forecast. The Q3 balance sheet lists 415.093 million common shares outstanding at September 30, versus 404.963 million at the end of 2025, a 2.5% increase. It also lists $5.447 billion of long-term borrowings and $0.656 billion of cash and equivalents, alongside $1.719 billion of restricted cash. Restricted cash is not interchangeable with freely available cash. Project-level debt may be supported by future rent, but common shareholders still need a clear bridge from completed campuses to cash after interest, maintenance and reinvestment.

What would change the CIFR forecast

The encouraging evidence is tangible: a first $5.852 million rent line, a low direct-cost line at this early stage and more capacity scheduled for delivery. TECHi’s CIFR quote page and forecast page provide dated market and analyst context, which should be kept separate from this filing-based operating test. The counterargument is equally tangible: mining sales fell 70.9% year over year, Q3 operating loss widened to $71.9 million, and reported net income benefited from a warrant revaluation rather than operating profit. The market’s early decline does not settle which side wins over 2027. It signals that investors are demanding proof beyond the size of signed contracts.

For a constructive revision, watch for quarterly HPC rent rising toward the $17.3 million static adjusted-EBITDA hurdle and, over a longer period, for the operating-loss bridge to narrow without an equal jump in debt or shares. Those levels are TECHi sensitivities based on Q3 inputs, not company targets. A flat or falling rent line after promised hall deliveries, higher direct costs, weaker project margins or another delay in firm power access would weaken the thesis. An improving GAAP operating result together with cash generation would matter more than an isolated positive net-income print.

What to watch

The next dated test is Cipher’s fourth-quarter 2026 results, expected in the next earnings cycle; the company has not provided a confirmed release date in this November 2 update. Before then, watch for disclosures on year-end data-hall delivery, ERCOT’s final treatment of conditional load, and new tenant agreements for the 270 megawatts it identifies for 2027. The practical question for CIFR is whether each new megawatt becomes recognized rent and cash quickly enough to outrun the cost of building it.