Summary Maintaining a Buy rating for Cipher Mining, driven by imminent HPC/AI lease revenues from AWS and Fluidstack starting in H2 2026. Cipher Mining projects ~$870 million in annual revenue and 76% blended EBITDA margins, with $8.5 billion in contracted HPC revenue over the contract years. At ~7.9x forward sales and 10.2x EV/EBITDA forward, based on projected sales and earnings, Cipher Mining trades at a discount to data center peers despite potential for a higher growth rate. Execution risk centers on timely site energization; delays beyond August/September 2026 could trigger sharp sell-offs. Cipher Mining ( CIFR ) has been one of the Bitcoin miners executing well these past years. I've been long CIFR for the past few years. My past coverage of the company reflects this stance, which all carried Buy ratings. In between the coverages, there have been good runs and sometimes retracements. In my last coverage, I highlighted the point that Cipher Mining’s rerating story isn't over yet . CIFR is up a lot from 2024 but currently off 2025 highs. In this piece, I'd look at what lies ahead for 2026 as a Bitcoin miner and an HPC/AI play and also touch on the HPC deals secured last year and the implication of those deals into the present thesis. Cipher Mining - From Pipeline to Paychecks in 2026 Cipher Mining is one of the HPC/AI plays expecting an operational inflection point in 2026. The market at the moment has remained muted since the HPC/AI hype late last year for the miners that secured deals. They all rerated in Q4 but have now since mostly retraced those gains. CIFR itself surged to a 52-week high of ~$25 on the news of the $3 billion 10-year Fluidstack lease deal last September, then followed by the $5.5 billion AWS deal, before retracing following the issuance of senior secured notes totaling $1.73 billion in mid-November. I'd like to say that the market seems exhausted in assigning a premium to miners turned HPC plays based on just announced deals, pipeline capacity, and projected cash flow discounted into today's share price. I highlighted these in an article published here on Seeking Alpha last Friday covering Hut 8 Corp. ( HUT ), another miner turned HPC play. While each of these miners has secured what looks like similar multi-billion HPC deals, nuances in the deals themselves, as well as each miner's equity composition (of which I'm currently weighing their specific exposure to Bitcoin volatility through their balance sheets) and projected net operating income [NOI] for the duration of the deal are the specific metrics I'm looking at as I analyze these companies. In my latest HUT article , I downgraded HUT from a Buy to a Hold, despite HUT having seen similar capacity development and deals win as the rest of the HPC pivot companies. The main reason was the longer-dated timeline for their River Bend site energization compared to peers, which was compounded by the company's exposure to Bitcoin as it holds over 13,000 BTC on the balance sheet. These present a double headwind for the company, making 2026 a potentially muted year for HUT. In contrast, as I analyze Cipher Mining going into 2026, I'd be maintaining a Buy for CIFR for the singular reason that the checks for the HPC lease from AWS are projected to hit the books from August this year, and 168 MW Fluidstack hosting capacity at Barber Lake expects to commence rent in September. Impressive top line numbers making the headline are a first place to watch for the next catalysts. I’ve talked about the sentiment of the current market, which has apparently moved beyond maintaining momentum based on just pipeline capacity or capacity under exclusivity. Those were last year's catalysts. 2026 is the execution year, when HPC pivots show actual top-line impact on financials. To project what's coming for Cipher, I'd make some assumptions. Take the Cipher's base mining revenue to be ~$55 million per quarter (I derived that using the average revenue of three quarters already reported in FY25); annualized, this gives a steady mining run rate of ~$220 million. I'm quite positive the company can now produce that amount of quarterly revenue from mining as a steady run rate until the next halving, with current hashrate at 23.5 EH/s and mining efficiency looking good at 16.8 J/Th. Now layer in the HPC contracts. Cipher has roughly $8.5 billion of contracted revenue from the AWS and Fluidstack deals. If this revenue is recognized on a straight-line basis over a 13-year weighted contract life, it implies ~$650 million in annual HPC revenue (calculated as $8.5 billion for the combined deal divided by 13 years). And would put total sales from both mining and HPC to about $870 million. Note that my estimate is still a conservative one as I use a straight-line assumption for the modeling. Cipher's adjusted EBITDA margin improved greatly in 2025. It was just around 12% in Q1 2025 on $49 million revenue and $6.1 million adjusted earnings, but jumped to ~60% in Q3 on a $72.0 million revenue and ~$41 million adjusted EBITDA in Q3. I believe higher adjusted EBITDA is sustainable because it was driven by the upgrade of the mining fleets from older Bitmain T21s to S21 XP Antminers . Economies of scale are likely to stay intact in the meantime, with factors like network difficulty and Bitcoin price not moving too much. For the purpose of my modeling for projected earnings, I'd, however, factor in volatility and go with a conservative ~50% adjusted EBITDA for the mining operations moving forward, which I believe the company can sustain considering the improved 16.8 J/Th average efficiency the S21 XP Antminers have brought to the mining fleet. Valuation is bound to shift from Cipher's current life as a Bitcoin miner when the HPC checks hit in H2. In the deals, management is guiding for an 80–85% NOI margin, which I believe is very well within range, as the AWS deal is a triple N, meaning the clients take care of the biggest three costs (power, cooling, and facility costs). For this modeling, I'll be treating NOI margin as EBITDA margin, which is safe enough considering low overhead and other operational levers in the HPC lease deal which, when accounted for, won't skew our EBITDA from NOI estimate materially. Segment Annual Revenue Projection ~ Adj. EBITDA Margin Annual EBITDA Contribution Bitcoin Mining $220 Million 50.0% (Conservative floor and sustainable with current mining economics) $110.0 Million HPC/AI Hosting $650 Million 85.0% (Guided NOI margin) $552.5 Million Total Pro Forma $870 Million 76.1% (Blended) $662.5 Million At a $6.94 billion market cap at present and $870M projected 2027 sales, CIFR trades at ~7.9x sales, meaning at current market cap, CIFR is trading around 7.9x projected sales. A ~7.9x forward sales multiple for a company with 75%+ blended EBITDA margins and contracted revenue from AWS and Google can be considered modest. Traditional data center providers like Equinix ( EQIX ) and Digital Realty Trust ( DLR ) typically trade in the low mid teens to sales, but they have much slower growth rates as they are well-established firms beyond the rapid growth phase. The case for CIFR is that it is combining high-margin HPC revenue with optionality in Bitcoin upside and potential for rapid growth, which are visible through the pipeline capacity and the AWS and Fluidstack deals themselves, which have clauses that allow expansion down the line. Using the current $6.78 billion EV and our blended EBITDA projection of $662.5 million from both segments, it implies an EV/EBITDA [fwd] of 10.2x (gotten from EV divided by projected EBITDA), and the valuation gets even more attractive. At 10.2x 2027 EBITDA, CIFR is being priced as if the HPC deals are still high-risk speculations. The market is ignoring the fact that the AWS and Fluidstack contracts are legally binding and the sites are already in build-out; the market is underpricing the near certainty of the coming cash flows. As the HPC revenues start to come in, if CIFR catches up to the Sales multiple of data center provider peers (moving from 8x to 10–12x), that would imply a ~25% upside for the rerating from current stock price. If CIFR catches up to the EBITDA multiple (moving from 10.2x to 20–30x), the stock would more than double. Risks The risk I worry about most at this stage is construction and energization slippage risk. As I have noted with Hut 8 and why I downgraded to a Hold because of longer-dated energization timelines, while peers are already locked in for HPC revenue just months away. The market punishes longer-dated timelines. Any hint that the August 2026 for AWS site launch and October for Fluidstack site launch is sliding into 2027 might trigger a sharp CIFR sell-off. Takeaway The market moved with momentum on the AWS deal news in November and is now underappreciating it in January. The gap between current prices around $17 and the top last November around $25 is the execution premium that will likely return as we get closer to that first lease payment in 2H this year. We can do as much rough math as we want, but unless the HPC lease payments don’t start hitting the top line in September/October, CIFR will likely trade as a momentum stock in the latter part of this year, to be driven by the headline top-line numbers from the HPC deals.
