While TeraWulf relies on credit support for lease commitments, IREN has adopted a different strategy. The company combines Microsoft’s customer prepayment with loans that use graphics processing units (GPUs) and contracted payments as collateral.

The distinction raises a pivotal question. How much of a multibillion-dollar AI contract can ultimately translate into shareholder returns?

Construction expenses, hardware purchases, maintenance, insurance and employee costs can reduce the income available to investors. Furthermore, borrowing introduces interest payments and repayment obligations that developers must meet.

July reporting identified leveraged loans and high-yield bonds as potential components of TeraWulf’s Kentucky financing package. At that stage, the proposal did not establish that the company had borrowed the entire amount.

Loan availability also differs from actual borrowing. For instance, an operator with a hypothetical USD 1 billion credit facility might initially draw only USD 200 million. Interest expenses and fees on unused funds could follow different calculation methods.

As construction progresses, further borrowing could increase financing expenses. Commercial agreements also determine whether developers must purchase GPUs, servers and networking equipment or primarily provide facilities for customers.

Read More: Bitcoin Miner Jiang Zhuoer Exits DeFi Tokens After BTC Slides Toward USD 62,000