AI borrowing boom may keep Bitcoin's macro headwind alive after Fed pauses
Bitcoin's next macro test may lie at the long end of the US Treasury curve rather than in the timing of the Federal Reserve's final rate rise, after minutes from the Fed's 15-16 September meeting showed market participants citing heavy private debt issuance for artificial intelligence (AI) infrastructure as one factor pushing yields and term premiums higher.
The minutes said nominal yields rose about 35 basis points across maturities of two to 10 years between Fed meetings. Policymakers raised the federal funds target by 25 basis points to 3.75%-4% in September, and most officials judged that another increase would probably be appropriate before the end of the year. The minutes did not quantify how much of the yield move came from AI-related financing, also citing stronger economic data, expectations of further tightening, geopolitical developments and uncertainty around Treasury buybacks.
The financing demands have drawn the attention of the Bank for International Settlements. The BIS estimates the five largest technology companies will spend more than $1 trillion on AI-related capital expenditure across 2025 and 2026, and industry projections it cites put global AI investment at roughly $500 billion today, potentially rising to between $3 trillion and $4 trillion by 2030. The BIS has said debt is becoming a larger part of the financing mix as firms build data centres, buy chips and secure energy infrastructure. The Fed's trading-desk manager said spreads on debt issued by major cloud providers remained wide because of the amount being borrowed and the long maturities involved.
Real yields are what sharpen the competition with Bitcoin, since they reflect the return available after inflation. Treasury data showed the 10-year par yield at 5.28% on 7 October, with the inflation-adjusted 10-year yield at 2.92%. The Fed said changes in real rates accounted for most of the increase in longer-dated Treasury yields during the intermeeting period, pitting an asset with no contractual cash flow against government securities offering positive inflation-adjusted income. AI-linked equities have so far absorbed the higher financing costs more comfortably: the Fed said companies benefiting directly from infrastructure spending outperformed the broader market, with stronger actual and expected earnings supporting prices even as valuation multiples declined.
Arthur Hayes, co-founder of the defunct BitMEX exchange, argues the data-centre race will ultimately produce excess computing capacity and a downturn. He has pointed to historically overbuilt technology rollouts and expects financial stress to emerge as new capacity comes online, potentially around late 2027 or 2028. His thesis holds that if an AI bust threatens heavily financed infrastructure owners, policymakers would eventually respond with liquidity support, conditions he expects to favour Bitcoin and other crypto assets. That remains speculative: AI demand could grow fast enough to absorb the capacity under construction, and higher productivity and profits could validate the spending before debt burdens become problematic. The BIS nevertheless describes the financing structure as a genuine vulnerability, since investment commitments increasingly exceed internally generated cash.
The immediate signal for Bitcoin investors is therefore less about which meeting ends the tightening cycle than about what happens to long-term real yields afterwards. A sustained decline in the 10-year yield and term premium would weaken the argument that AI capital demand is keeping financial conditions tight, while strong spot demand for Bitcoin amid elevated yields would show investors are willing to accept the higher opportunity cost. The opposite combination would be harder for crypto markets, leaving Treasuries and corporate credit competing aggressively for marginal capital even after monetary tightening peaks. The Fed's next meeting is scheduled for 27-28 October, with officials still focused on inflation and the possibility of another increase by year-end.
