- Crypto, Inflation, and AI Infrastructure Expose the Price of Uncertainty
Investors are being asked to reassess three assumptions that have supported technology valuations: that crypto regulation will remain politically favourable, that central banks will soon ease policy, and that artificial intelligence investment will translate quickly into profitable growth.
Those assumptions are now under pressure. The US midterm elections have shifted control of Congress to Democrats; the Federal Reserve has left interest rates elevated as inflation remains above target; and delays to data-centre projects are raising questions about the cost and timing of AI expansion. Separately, a US Treasury order targeting crypto exchange MEXC has put access to banking and settlement infrastructure at the centre of the digital-asset debate.
Utkarsh Ahuja, founder and managing partner of Moon Pursuit Capital, argues that investors must “distinguish between demand for a technology and the returns available to those financing it”.
That distinction cuts across both crypto and AI. Users may adopt new products and companies may keep spending, yet investors can still lose money if financing, operating costs or execution overwhelm future revenue.
Crypto’s immediate test is political and regulatory. Democrats won control of both chambers of Congress in the November 3 midterms, though the new majorities take office in January.
The result changes the legislative environment for an industry that had hoped for a clearer federal framework under a Republican-controlled Congress.
The Digital Asset Market Clarity Act has already passed the House, but it has not become law. On September 15, the Senate failed to invoke cloture on a motion to proceed, with a 49–50 vote.
The election therefore adds another layer of uncertainty to legislation that had stalled before voters went to the polls. The next Congress may revisit market structure, but the form and timing of any compromise remain uncertain.
For investors, the question is not simply whether regulation becomes stricter or more permissive. A durable framework could encourage institutions to participate by clarifying which agencies oversee different assets and what rules apply.
But compliance requirements can also raise costs, especially for smaller firms. The market will have to distinguish rules that create confidence from rules that entrench existing players.
The MEXC case shows that access to financial infrastructure can matter as much as exchange trading volume. On November 6, the US Treasury’s Financial Crimes Enforcement Network issued an order preventing US financial institutions from processing funds transfers involving MEXC.
Treasury alleges that the exchange facilitated at least US$16 billion in transactions connected to sanctions evasion and other illicit activity between May 2021 and May 2026. The order takes effect 30 days after publication in the Federal Register.
Treasury said MEXC processed more than US$1 billion in daily transactions and held approximately US$5 billion in reserves as of September. Those figures illustrate why a restriction on a large exchange can attract attention across digital-asset markets.
But the action concerns a specific platform and Treasury’s allegations; it does not establish that every crypto exchange faces the same risk. Its broader lesson is that apparent market depth does not guarantee that funds can move freely through banking and settlement channels.
Ahuja captures that distinction directly: “For a fund, liquidity means more than executing a trade. It means being able to withdraw the proceeds and redeploy the capital.” Order books may show that buyers and sellers are active, but investors also need reliable custody, banking access and counterparties. If any link in that chain is disrupted, the cost of holding or moving an asset can rise.
The second pressure point is the cost of capital. US consumer prices rose 3.7% over the 12 months to September, while core inflation reached 2.6%. The Federal Reserve held its target range at 3.75% to 4.00% in October. On November 6, San Francisco Fed President Mary Daly said further tightening might be needed if inflation did not come down, while cautioning that the timing and scale would depend on incoming information.
That is a difficult backdrop for businesses whose valuations depend on profits expected years from now. Higher rates increase the discount applied to future earnings and make borrowing more expensive.
The effect is particularly important for AI infrastructure, where companies are committing large sums to data centres, chips and power systems before those assets generate their expected returns.
The physical rollout is already running into constraints. Oracle issued a force majeure notice connected to a major New Mexico data-centre project developed by Blue Owl, as lenders and investors reassessed risks around the pace of AI expansion.
At least 45 projects worth US$68 billion faced community opposition in the second quarter of 2026, following 75 projects worth about US$130 billion in the first quarter, based on project-tracker figures cited in market reporting.
Delays do not automatically mean AI demand is weakening. They can, however, postpone revenue while construction, financing and equipment costs continue to accrue.
A project that opens late may still be valuable, but its economics can deteriorate if the developer must carry debt for longer, pay more for power or renegotiate customer commitments.
Investors therefore need to ask who owns the asset, who bears the delay risk, and whether customers have made firm commitments to use the capacity.
Ahuja’s warning that “demand growth alone does not answer those questions” is apt. The AI investment story is moving from a race to announce capacity towards a more demanding test of utilisation, pricing power and cash generation.
Companies with access to power, committed customers and strong balance sheets may be better placed to manage delays. Developers relying on repeated refinancing face a more difficult calculation.
The common thread across crypto and AI is the durability of access to capital. Crypto platforms need dependable banking, settlement and regulatory arrangements.
AI developers need financing, electricity, permits and customers. Inflation and interest rates influence the price of funding both. Political shifts can alter the rules under which capital is deployed.
That does not erase the long-term case for either technology. It does make broad enthusiasm less useful as an investment thesis. As Ahuja puts it, “Adoption can keep growing while individual investments fall short.”
The next phase of market performance may depend less on whether crypto and AI expand, and more on which businesses can keep operating, raise capital on reasonable terms and convert that growth into sustainable returns.
