Bitdeer Technologies Group (NASDAQ: BTDR) reported record third-quarter revenue of $320.7 million, up 89% from $169.7 million a year earlier and above the $251.11 million analysts had expected. However, the company’s per-share loss was wider than anticipated, while higher operating costs and a sharp increase in its net loss weighed on the results. The quarter thus presents a split picture: a top line running well ahead of forecasts, with profitability moving the other way.

BTDR closed Monday at $6.61, up 9.9% ahead of the earnings report. Trading was more volatile on Tuesday: premarket reports showed moves ranging from a 1.5% gain to a 2.7% decline, and the stock later fell about 7%. Bitcoin was down about 2% at the same time. The pullback came despite the revenue beat, keeping the focus on the earnings side of the report.

The company reported a loss of $0.46 per share, compared with the $0.37 loss analysts expected. An earnings summary published by Wall St Engine cited a $0.34 estimate and reported adjusted EBITDA of $69.6 million, up 173% year over year.

BITDEER $BTDR Q3’26 EARNINGS HIGHLIGHTS Revenue: $320.7M (Est. $256M) ; +89% YoY EPS: -$0.46 (Est. -$0.34) Adj. EBITDA: $69.6M; +173% YoY Announcements: AI Cloud Backlog: ~$2.9B; ~332 IT MW secured Tydal: $4.7B 16-year AI/HPC lease; 121 IT MW; ~$8.0B… pic.twitter.com/x7rDcZqmcT — Wall St Engine (@wallstengine) November 10, 2026

The company’s official results are available in its third-quarter 2026 earnings announcement.

Bitcoin Production and Holdings

Bitdeer mined 3,855 BTC during the quarter, compared with 1,109 BTC a year earlier. Its proprietary hashrate increased to 86.4 EH/s from 35 EH/s.

Despite the higher production, the company held only 47 BTC as of Sept. 30, down from 2,029 BTC a year earlier. Bitdeer received $221.5 million from selling digital assets during the quarter, primarily mined bitcoin. Selling output shortly after mining rather than accumulating it leaves the balance sheet with minimal direct exposure to bitcoin’s price.

At the end of September, the company held $43.2 million in digital assets and had $343.8 million in digital-asset receivables from a related party, for a combined $387 million. Cash and cash equivalents totaled $213.7 million.

Higher Costs Reduce Profitability

Cost of revenue rose to $309.9 million from $130.0 million, mainly because of higher electricity costs and depreciation — a 138% increase that outpaced the 89% revenue growth. As a result, gross profit declined to $10.8 million from $39.7 million, compressing gross margin from roughly 23% to about 3%.

Net loss widened to $125.5 million from $30.1 million. The quarterly loss included a non-cash charge of $54.6 million related to a digital-asset loan, an item that reduces reported earnings without a cash outflow and represented roughly 44% of the quarterly loss.

Adjusted EBITDA nevertheless increased 173%, reaching $69.6 million compared with $25.5 million a year earlier. Self-mining revenue rose to $217.5 million from $130.9 million, while co-mining contributed $59.9 million.

AI Cloud revenue increased to $20.4 million from $1.8 million. The AI Cloud backlog reached approximately $2.9 billion, with about 332 IT megawatts of capacity secured. That backlog is more than 140 times the quarter’s AI Cloud revenue, meaning most of the contracted work has yet to be recognized as revenue.

Chief Financial Officer Michael G. Potter said Bitdeer is “converting AI demand into long-duration contracts.” He said customer prepayments and project-level financing are intended to fund AI capacity as it is contracted. The two lines of business carry different revenue profiles: self-mining is tied to bitcoin mining, while the AI/HPC contracts run as long as 16 years, the term of the Tydal lease.

Construction at the Tydal site in Norway remains on schedule for energisation during the fourth quarter of 2026. The earnings summary also referenced a $4.7 billion, 16-year AI/HPC lease involving 121 IT megawatts at Tydal, alongside approximately $8.0 billion in related figures. Energisation is the nearest operational checkpoint; beyond it, the markers to follow are the pace at which the AI Cloud backlog converts into recognized revenue and whether electricity and depreciation costs keep growing faster than sales.

The original report was published by Blockonomi.