What Do The Midterms Mean For Crypto And AI?

Markets first rallied after the midterms on the assumption that political gridlock would help, then reversed sharply. In the host’s November 5 snapshot, the Dow was up 3%, Bitcoin had climbed above $80,000, and the Nasdaq was down 4% to 6%. Matthew Siegel said investors were repricing AI capital spending. He called the Texas governor’s election result a surprise, given that Texas accounts for 30% of the AI data-center pipeline. Building the full pipeline could require $5 trillion and create nearly 1 million jobs, he estimated; if 90% of data centers are never built, the rough figures fall to $1 trillion and 100,000 jobs. He warned that Nvidia estimates could be vulnerable if there is not enough space to deploy its chips.

Tight credit, greater reliance on equity financing, and investor concern about dilution are pressuring AI projects and stocks. Rob said major AI-related shares had already fallen 10% to 20% in October and thought the previous day’s rally was likely short covering; Siegel had heard of two hedge-fund liquidations but did not know whether Situational Awareness was among them. Rob warned that interconnected revenues and negative unit economics could spread trouble across the technology stack if debt and equity markets tighten. Siegel countered that AI could also cut costs: a 10% efficiency gain on hyperscalers’ roughly $1 trillion in annual operating expenses would save $100 billion. He cited Microsoft offering buyouts to about 25% of its workers.

Speakers said roughly 200 well-funded AI companies could be exposed if GPU supply tightens or prices rise while compute costs fail to fall, particularly if limited partners lose confidence and venture funding becomes harder to secure. Another speaker noted that venture-backed firms are not the main source of demand for CoreWeave and other neoclouds, and expected failures and consolidation to leave compute with the strongest balance sheets and highest bidders. Neoclouds, which own GPUs, were described as about five times more costly than data-center landlords, with contracts shortening from 15 years to three and greater debt and liability risks. Siegel said landlords’ 10-year leases generally have investment-grade counterparties and that valuations imply no terminal value for multibillion-dollar facilities; he saw no evidence of collapsing demand or GPU prices. He also said a proposed Texas framework requiring union jobs, community engagement, transparency, water neutrality, and protection against higher utility bills, without tax subsidies, was relatively reasonable. The political outlook for permits and construction nevertheless remained uncertain.

Siegel said Bitcoin is more closely correlated with software than with the Nasdaq broadly, leaving it exposed to AI-driven disruption of software, lower barriers to entry, and easier vulnerability discovery. Its slower upgrade cycle could add risk. He expects Bitcoin to remain highly correlated with the Nasdaq for six to nine months as both respond to liquidity and technology-sector activity; a break in that relationship, particularly a shift toward gold, would be bullish. His medium-term bull case is that a spending compromise between President Trump and Democrats could weaken the dollar and support Bitcoin and gold. His bear case is tighter credit hurting Bitcoin and other liquidity-sensitive assets, with gold potentially falling less. Jake argued that Democratic control could support Bitcoin through spending and gridlock while resistance to data centers weighs on hyperscaler growth and AI shares. The host said crypto had held up surprisingly well amid AI weakness, election results viewed as unfavorable by crypto supporters, and the Binance hack.

Stablecoins, tokenization, and AI-crypto crossover work were cited as possible sources of demand beyond core crypto trading, although speakers expected continued volatility. Stablecoin velocity has risen about threefold since the GENIUS Act passed, while payments grew roughly 25% to 35% quarter over quarter, with some quarters near 50%. Visa said it was settling $25 billion directly in stablecoins, up from about $7 billion a year earlier and near zero six months before. Faster settlement and lower credit risk can reduce the capital businesses need to hold, including for cross-border and multi-currency services. Yet weaker spot-trading demand has weighed on prices, volumes, and new stablecoin minting; perpetuals volume is below its highs from about a year earlier. The speaker expects Circle to miss Street estimates for USDC growth and margin stability unless conditions change substantially, citing reported margin pressure and competition, despite Circle gaining share in institutional use cases.

Speakers were strongly bullish on tokenization’s long-term potential but disputed where value will accrue. One put 95% conviction on stocks, bonds, currencies, commodities, and capital markets moving onto public blockchains, while acknowledging uncertainty over whether crypto companies or tokens will capture the value. A skeptic pointed to falling stablecoin assets under management, ETFs as an existing token-like alternative, and unanswered questions about duplicated on-chain capital and pricing power, preferring infrastructure to consumer businesses. Tokenized equities have reportedly grown 400 to 500 times from a small base, but Matthew questioned models that require balance sheet and fragment liquidity across duplicative products. He sees possible benefits for overseas investors but no clear winner-take-most dynamic. His preferred example is Figure, which originates home-equity lines and first mortgages for institutions and operates a marketplace for those assets; he cited 50% margins, 100% year-over-year growth, and a 20-times-earnings valuation. He favors its regulator-accepted blockchain record and existing product-market fit. Rob, meanwhile, argued that Polymarket and Kalshi showed how prediction markets can aggregate information rapidly, sometimes signaling that a race is effectively over before news outlets or models do.