Earnings call transcript: Terawulf Q1 2026 EPS Misses, Stock Volatile By Investing.com

TeraWulf reported Q1 2026 EPS of -$1.01 versus -$0.19 expected and revenue of $34 million versus $36.59 million expected, but HPC leasing revenue surged 117% to $21 million as the company continues its pivot away from Bitcoin mining. GAAP net loss widened to $427.6 million, largely from non-cash warrant fair value adjustments, while adjusted EBITDA improved to -$4.1 million and liquidity remained strong at $3.1 billion in cash and restricted cash. Management reiterated growth in contracted HPC capacity, targeting positive EPS by FY2027 and 480 MW online in the second half of 2027.

Analysis

The key market implication is that WULF is evolving from a hash-rate proxy into a utility-like contracted compute platform, which changes the right comp set and the earnings bridge. That should be constructive for infrastructure capital providers and for customer-facing AI platforms that need scarce power, but it is hostile to pure-play miners that cannot credibly reprice into HPC. The second-order effect is on supply: every successful conversion of mining assets into data-center load tightens the available pool of “ready now” power sites, which should support lease economics for the entire cohort.

The near-term print is messy because the P&L still mixes transition costs, non-cash accounting, and project-level spend that will fade as assets come online over the next 2-3 quarters. That creates a classic gap between economic value and reported EPS, so the stock can remain volatile until investors see a cleaner run-rate on cash EBITDA and utilization. The big tell is that management is effectively signaling a financing moat: cheaper, larger, and more diverse project capital lowers the probability of a forced equity raise and raises the bar for smaller competitors.

The consensus is likely underestimating how much this business becomes a land-and-power arbitrage story rather than a GPU story. If the power bottleneck persists through 2026-2027, WULF’s optionality on utility partnerships and brownfield reuse becomes more valuable than the current GAAP losses imply. The main risk is execution slippage at the newly announced sites: a 3-6 month delay can materially compress project IRR and leave the equity exposed to duration risk while the market waits for contracted revenue to catch up.

Key Decisions for Investors

• Long WULF vs short a basket of legacy miners (e.g., MAR/RIOT/IREN) for a 3-6 month relative-value trade: thesis is multiple re-rating toward infrastructure/HPC economics while miners remain tethered to volatile coin economics; stop if WULF loses site-timing credibility or if BTC rallies sharply enough to reflate miner multiples.
• Buy WULF on pullbacks only, targeting post-earnings de-risking over the next 2-8 weeks; risk/reward improves if the market fades the non-cash loss and focuses on 2H26 revenue conversion, but avoid chasing strength above recent highs given execution-dispersion risk.
• Pair long GOOGL / long NVDA against short WULF only as a hedged way to express AI upside without single-asset execution risk; WULF is the higher-beta beneficiary, but the cleaner trade is to own the demand enablers rather than the capital-intensive landlord.
• For options users: sell cash-secured puts on WULF 1-2 expiries out on any sharp drawdown, because the stock’s implied volatility should stay elevated while catalysts are binary; only do this if willing to own the name through potential permitting/FERC delays.
• Monitor AESI as a secondary beneficiary of power build-out and gas-backup demand; if Maryland/Kentucky progress accelerates, the market may start pricing a broader ecosystem trade in infrastructure-adjacent names before fundamentals show up.