Which capital pays the bill?
The first number I look at in TeraWulf's data-centre transition is the 2.23 billion dollars spent on equipment and facilities during nine months, alongside 178.9 million dollars of operating cash outflow. Financing activities provided a net 1.01 billion dollars over that period. Shareholders and creditors are carrying capital through construction before rental income matures. The timing gap between expenditure and cash generation explains the present company economics more usefully than the headline profit. Progress on new buildings still requires that gap to be financed.[1]
Quarterly net income attributable to shareholders was 709.9 million dollars, alongside a noncash gain of 728.4 million dollars from valuing warrants and derivative liabilities. Such a valuation gain lifts the income statement without producing cash to pay construction bills. The 95.0 million dollar gain on an investment sale also has a different economic basis from recurring rent. I would not capitalise these items as recurring earnings. With an operating loss of 88.2 million dollars, applying headline net income directly to the earning power of the buildings would capitalise a valuation movement ahead of the assets' operations.[1]
It would nevertheless be a mistake to dismiss the leasing progress. High-performance computing lease revenue rose to 63.0 million dollars from 7.2 million dollars in the same quarter a year earlier, while digital asset revenue fell to 13.2 million dollars from 43.4 million dollars. The revenue mix is changing materially. Part of the operating loss could reflect carrying costs for capacity preparing to generate revenue. An alternative is that the new business has persistently heavy operating costs. The aggregate result does not establish the weight of those explanations; dismissing either would be premature.[1]
Cost and timing enter the same calculation
Cash classified without restrictions was 1.60 billion dollars; adding restricted cash brought the total to approximately 2.36 billion dollars. Treating the second total as entirely available for new construction would erase those restrictions. The 1.20 billion dollars raised through equity issuance during nine months also shows the shareholder side of the financing. The question is which capital carries the burden before rent starts, as well as whether funding can be obtained. A large cash balance and heavy construction spending can coexist; the funding mix separately matters to a shareholder's economic interest.[1]
The cost assumption has disclosed boundaries: 336 megawatts are under construction in CB4 and CB5, with management's range of 8–10 million dollars per megawatt of critical IT capacity. Holding capacity constant and moving cost towards the upper end requires more capital for the same size of rental asset. The range is useful for testing the construction scale, but the entire amount cannot be treated as expenditure still to be paid. Adding already incurred investment again to a funding gap would overstate the economics. The capital tied to capacity is a valuation sensitivity alongside the capacity itself.[1]
For me, the decisive assumption is that this capital starts earning rent on management's schedule and within its cost range. Lake Mariner has 102 megawatts operating; CB4 rent commencement is targeted for the fourth quarter and CB5 phased delivery for early 2027. The disclosed progress provides a real basis for that assumption. A delay would leave capital carrying financing costs for longer before generating revenue. I therefore give rising rental revenue weight in assessing TeraWulf, while refusing to substitute the size of a noncash gain for future distributable cash. The shareholder's economic claim depends on invested capital earning revenue on time.[1]