Mentioned in This Article
1. Cipher Digital earnings: a profit built on an accounting entry
Cipher Digital, the company that began as a bitcoin miner and is now developing data centers, reported a third-quarter net profit of $31.6 million, equal to 8 cents per share, even though its revenue fell 63% from a year earlier. The results, disclosed in a November 2 filing with the US Securities and Exchange Commission, were lifted by a noncash gain of about $127 million tied to warrants issued to Google.
The headline profit therefore says little about how much cash the business generated. Behind it sit a shrinking mining operation, the first revenue from a new data center line and an operating loss that grew sharply. Taken together, the quarter shows a company in the middle of changing what it sells, and paying for the change.
2. How much did revenue fall, and why?
Revenue for the quarter was $26.7 million, compared with $71.7 million in the same period a year earlier, a drop of 63%. The company's own filing attributes much of the decline to a deliberate shift in how it uses its computing capacity.
Bitcoin mining revenue came to $20.8 million, which is 71% lower than a year before. Cipher has been moving capacity away from producing bitcoin and toward leasing data center space to companies that build and run artificial intelligence systems. Mining is a business in which revenue depends on the price of bitcoin and on how many coins a fleet of machines can earn, while leasing typically rests on multi-year contracts with fixed payments. Moving from the first model to the second means giving up mining income before the lease income has fully arrived, which helps explain why the top line shrank so quickly.
The new business did contribute for the first time. Cipher booked $5.9 million of data center revenue in the quarter, after the first phase of its Black Pearl facility in Texas began operating in August. Because that revenue covered only part of the quarter, it offset just a fraction of what was lost from mining.
3. What is the $127 million warrant gain?
The gap between a falling top line and a positive bottom line is explained by a single item. Cipher recorded a noncash gain of roughly $127.2 million from the change in fair value of warrants that it issued to Google. A warrant is a contract that gives its holder the right to buy a company's shares at a set price. Companies that issue them must periodically restate their value on the balance sheet, and the change flows through the income statement.
The connection to Google comes through a lease arrangement. Google agreed to backstop certain lease payments owed by Fluidstack, an AI infrastructure company that is one of Cipher's tenants. In other words, Google stands behind part of the rent that Fluidstack has committed to pay, and the warrants were issued as part of that arrangement.
The important point for readers is that the gain is an accounting adjustment. It reflects a revaluation of a financial instrument, not money received by Cipher. Without it, the quarter would have been a clear loss, which is the picture that the operating figures paint.
4. Why is the diluted result a loss of 22 cents per share?
The reported earnings of 8 cents per share look different once potential dilution is counted. Cipher's business update shows that, after including the additional shares that could be created by the warrants and other dilutive securities, the quarter becomes a loss of 22 cents per share.
Diluted earnings per share is a standard measure that asks what profit per share would be if every instrument that could convert into common stock actually did so. When a company has a large number of warrants, options or convertible securities outstanding, the diluted figure can differ noticeably from the basic one. Here the two even point in opposite directions, with a profit on the basic count and a loss on the diluted count.
For investors comparing Cipher with other companies in the sector, the diluted number is often the more conservative reference point, since it shows how much each existing share would be spread across a larger pool if the warrants were exercised.
5. What do the operating figures show?
Strip out the warrant revaluation and the quarter looks considerably weaker. Cipher reported an operating loss of $71.9 million, up from $37.6 million in the same quarter a year earlier. The loss nearly doubled while revenue fell by close to two-thirds.
Adjusted EBITDA, a measure that leaves out certain accounting items such as depreciation and non-recurring entries, swung to negative $10 million. A year earlier it was positive $40.8 million. That is a turnaround of roughly $50 million in the metric that many investors use as a rough proxy for underlying operating performance.
The transition is expensive in other ways as well. Cipher stopped mining bitcoin at the Black Pearl site in February so that the facility could be prepared for its new tenant. Taking machines offline removes mining income immediately, while converting a site for AI customers involves construction, power and equipment work that is paid for before leasing revenue ramps up. The combination of lost mining output and upfront investment is what pushes the operating line deep into the red.
6. What is the Barber Lake lease worth?
Management's argument is that the present losses are the price of a much larger business later. The company recently extended the lease at its Barber Lake data center to 20 years, and the contracted revenue from that site now totals about $9.1 billion.
A long lease with a creditworthy counterpart changes the nature of the revenue. Instead of earning whatever the bitcoin market allows in a given month, the company would receive contractual payments over two decades, subject to the tenant meeting its obligations. That is where the Google backstop on Fluidstack's payments becomes relevant, because it addresses the question of whether the tenant can reliably pay.
Contracted revenue is not the same as collected revenue or profit. The figure describes what the leases are expected to bring in over their terms, while the company still has to build, power and operate the sites, and it carries the costs and financing needs that come with that work.
7. Why are bitcoin miners pivoting to AI data centers?
Cipher is not the only company in its industry to reconsider what its facilities are for. Bitcoin miners control something that AI developers need and often struggle to obtain, namely sites with large power supplies, cooling infrastructure and the permits to run them. Converting such a site to host AI computing can be faster than building a new data center from scratch.
At the same time, mining economics are demanding. Revenue rises and falls with the price of bitcoin, and the reward for producing blocks is cut periodically under the network's rules. Leasing capacity under long contracts offers a steadier alternative, which explains why Cipher is willing to accept a lower mining output and a heavier near-term loss in exchange for contracted income.
The trade carries its own risks. The company becomes dependent on a small number of large tenants, and its results are tied to its ability to complete construction on schedule. The third-quarter figures show that, at this stage, the move costs more than it earns.
8. What should investors watch next?
The next test is whether data center revenue, which started at $5.9 million in a partial quarter, grows fast enough to replace the mining income that is being given up. The first full quarter of Black Pearl operations will give a cleaner view of that.
Investors will also watch how the operating loss and adjusted EBITDA develop as the construction phase continues, and whether the warrant valuation continues to swing reported net income in either direction. Because the warrants are revalued every period, future quarters could show a loss from the same item that produced this quarter's gain.
For now, Cipher's third-quarter report offers two readings at once: a reported profit of $31.6 million that comes from an accounting gain, and an operating business that is losing more money while it builds toward a $9.1 billion contracted site. Which of the two proves more representative will depend on how quickly the new leases turn into cash.
9. How should the quarter be read?
Three separate layers sit in the results, and each answers a different question. The first is the bottom line, a net profit of $31.6 million, which answers what the company reported under accounting rules. The second is the operating loss of $71.9 million, which answers how the business itself performed once the revaluation is set aside. The third is the lease backlog, including about $9.1 billion of contracted revenue at Barber Lake, which describes what management expects the transformed business to earn over many years.
Reading any one layer alone can mislead. A reader who sees only the profit would conclude the company is thriving, while a reader who sees only the operating loss and the revenue decline might conclude the strategy is failing. The more accurate view is that the quarter captures the costly middle of a transformation: the old revenue source has been cut back, the new one has only just started, and the long contracts that justify the change have not yet produced cash at scale.
It is also worth remembering that noncash gains work in both directions. A warrant that gains value for the holder generates an accounting gain for the issuer when its fair value is restated, and the reverse can occur in a later period. That is why analysts often look at operating results and cash flow, not just net income, when judging companies with large derivative-like instruments on their books.