KBW raises TeraWulf to outperform, citing AI pivot as a key growth driver

Investment bank KBW upgraded TeraWulf (WULF) to “outperform” from “market perform” and raised its price target to $24 from $9.50, citing the company’s pivot from bitcoin mining to AI and high-performance computing (HPC) leasing as a major growth driver.

KBW said the market is underestimating the earnings upside from TeraWulf’s strategic shift. Analyst Stephen Glagola wrote in a Wednesday note, “We think investors underappreciate the magnitude of the BTC mining to HPC leasing mix shift in 2026–2027 and robust growth catalysts on 646 MW net of visible HPC leasing pipeline through 2027.”

Shares were modestly higher in early trading Wednesday, at $11.18.

The move reflects a broader trend among bitcoin miners, which are increasingly repurposing existing data centers to host AI and HPC hardware, boosting profitability as traditional mining margins decline.

Glagola estimates TeraWulf’s current leases could drive a 505% EBITDA compound annual growth rate (CAGR) from 2025 to 2027, supporting multiple expansion from the stock’s current 13.8x EV/EBITDA valuation. His bullish thesis focuses on the company’s 646-megawatt HPC leasing pipeline through 2027 and the rapid reduction of mining’s contribution to revenue.

KBW expects HPC leasing to generate roughly two-thirds or more of TeraWulf’s revenue in 2026 and the vast majority of contribution profit, with bitcoin mining becoming largely immaterial by 2027.

The report also highlighted that execution risks are lower than the market assumes, citing secured financing for major build-outs, a proven track record of delivery, and supportive debt markets. KBW noted that recent share weakness largely reflects sector-wide selling in bitcoin miners rather than company-specific issues.

The bank added that these discounts should narrow as HPC lease revenues scale in 2026, driving cap-rate compression and valuation upside, with additional optionality from potential new HPC deal announcements over the next year