Summary I reiterate my 'sell' rating on TeraWulf due to deteriorating mining economics and extended valuation versus peers. Rising power costs and declining Bitcoin production have significantly reduced gross margins, despite higher revenue from increased BTC prices. TeraWulf's cash burn and large capital expenditures suggest a need to raise capital within a year if current trends persist. While the HPC data center pivot is promising, I see better opportunities elsewhere in both the Bitcoin mining and HPC sectors. In conjunction with a Q1-25 earnings report that I felt was 'laughably bad' for TeraWulf Inc. ( WULF ), I downgraded the stock from 'hold' to 'sell' and argued that long-term investors should generally stay away. From where I sit, Bitcoin USD ( BTC-USD ) mining stocks can be decent ways to express shorter-term price moves in BTC provided coin price growth can outpace global hash rate capacity. We have yet to meaningfully see that scenario develop during this halving cycle with hash rate generally rising with price: Data by YCharts During the 2020/21 cycle, price moved faster than miners could plug in. This led to an explosion in mining profitability that took most mining stocks to mania levels that have yet to be revisited even 4 or 5 years later. While TeraWulf's Q2 wasn't nearly as bad as Q1, I think there are still important signals in the report that are worth examining. Q2-25 Earnings During the quarter ended June 2025, TeraWulf mined 485 BTC at an average value of $98,219 per coin. This was good for $47.6 million in revenue during the quarter - an increase of roughly 34% year over year and 38% sequentially. GAAP Metrics Three MonthsEnded Q2 2025 Three MonthsEnded Q2 2024 YoY Change Revenue $47,636 $35,574 33.9% Cost of revenue (exclusive of depreciation) $22,094 $13,918 58.7% Gross Margin 53.6% 60.9% -11.9% Source: TeraWulf, $ in 000s COGS grew by 58.7% year over year which resulted in a 12% reduction in gross margins from 60.9% last year to 53.6% in Q2 this year. However, it should be noted that Depreciation & Amortization also outpaced revenue growth at 46.4% year over year. Despite the 34% increase in top-line revenue from Bitcoin mining, the number of coins mined by TeraWulf went down by 30% year over year: Non-GAAP Metrics Three MonthsEnded Q2 2025 Three MonthsEnded Q2 2024 YoY Change Bitcoin Self-Mined 485 699 -30.6% Value per Bitcoin Self-Mined $98,219 $65,984 48.9% Power Cost per Bitcoin Self-Mined $45,555 $22,954 98.5% TeraWulf operational hashrate (EH/s) 12.2 8 52.5% Source: TeraWulf, $ in 000s The table above shows power costs being a significant driver of TeraWulf's year-over-year reduction in gross margins. The power cost to mine Bitcoin was nearly double what it was in Q2-24 - and again, this was for less Bitcoin. Power cost outpaced both the value of the Bitcoin mined and the operational hashrate of the company. Essentially, the economics of mining BTC deteriorated substantially for TeraWulf on a year-over-year basis. That said, the company actually pulled off a sequential improvement judging from the bottom line. TeraWulf's non-HODL approach to mining does two things; it theoretically limits the upside of the equity since such a large portion of the company's assets generally depreciate over time and it also gives us a more realistic glimpse into the economics of mining since quarterly net income doesn't have the tailwind/headwind of FASB accounting adjustments. Q2-25 Income Statement (TeraWulf) One of the biggest expenses during the first quarter has come back down to earth in the most recent quarter. SG&A of just $14 million in Q2 was more in line with Q2-24 than Q1-25. TeraWulf took a $3.8 million loss on miners that it sold in the quarter. All told, the operating loss before interest expense came in at $15.6 million in the quarter; again, an improvement from Q1 but a steeper operational loss than what was observed in Q2-24. Factoring in interest expense on the company's debt, TeraWulf produced a net loss of $18.4 million in the quarter. Balance Sheet & Valuation At the end of June, TeraWulf had just $90 million remaining in cash. During the quarter, the company spent $213.6 million on purchases and deposits for plant and equipment. Given TeraWulf's burn rate in the latest quarter, back-of-the-envelope math implies the company has a little over a year of cash remaining before it will have to raise capital again at the same Bitcoin mining economics we have today. Balance Sheet, Assets (TeraWulf) Of the $869.4 million in total assets, $604.8 million is plant and equipment; of which, $236.6 million is Bitcoin miners and $278 million is 'construction in progress.' The $55.5 million in Goodwill is from TeraWulf's purchase of Beowulf Electricity & Data. This was an interesting acquisition because Beowulf had previously been providing services to TeraWulf. The company was considered a related party because it was owned by TeraWulf CEO Paul Prager. Management hopes this acquisition will improve the optics of investing in WULF for long-only investors now that an important related party has been rolled into the company. Balance Sheet, Liabilities (TeraWulf) On the liabilities side, TeraWulf has $695 million in total liabilities; $151.3 million of which are current. The biggest line item here is the $488.7 million in convertible notes that the company sold in October. Those notes are due in 2030 and pay 2.75% with a conversion price of $8.48 per share. As I see it, TeraWulf isn't in immediate financial trouble. But it isn't difficult to envision scenarios where the price of WULF faces pressure. Data by YCharts Valuation is already quite extended relative to both the info tech sector median and to its close peer IREN Limited ( IREN ). At a price-to-book ratio of 12.4, shares of WULF are getting a 4x premium over shares of IREN. WULF also trades at a 2x premium through forward sales. Closing Takeaways Longer term, TeraWulf is continuing to build its HPC data center offering with Lake Mariner projected to reach a capacity of 750 MW when completed. Revenue from TeraWulf's HPC efforts will begin to be recognized in the company's next earnings report. In my view, this will be an important quarter for TeraWulf to prove its pivot to AI will be beneficial to shareholders. The economics of mining Bitcoin have not improved this year and the mergers/acquisitions that were expected following the halving over a year ago haven't meaningfully manifested. More importantly, TeraWulf's power cost increases are problematic, in my view, and this has been exacerbated by the sale of the Nautilus facility last quarter. I'm going to reiterate my 'sell' rating on WULF stock. I continue to see better opportunities in both the Bitcoin mining space and the HPC space, should one choose to invest in those themes.
