The Big Short Partners Reunite: Rates, AI, Gold and Two Stock Picks | The Real Eisman Playbook Ep 75

Summary

Steve Eisman, Vincent Daniel, and Porter Collins spend much of the episode on the tension between large deficits, elevated inflation, and the policy desire for lower rates. Collins argues the fiscal and monetary toolkit is stretched, while Daniel says the administration needs to push rates lower over the next six to nine months and would benefit from lower oil via an end to war. Eisman is less alarmed about immediate fiscal collapse but says Scott Bessent may have made a mistake by drawing a line in the sand that now has to be defended. They also debate gold as a reserve asset with no counterparty liability, AI capex concentration and return-on-capital risk, OpenAI’s weak economics relative to its growth narrative, and the operational structure of modern hedge fund pods at firms like Citadel and Millennium.

Notes

• Collins says the Treasury/Fed are using extraordinary measures because they are out of room on deficits, inflation, and yield control.
• He frames interest expense plus entitlements as about 100, or roughly 10% of tax receipts, leaving little fiscal flexibility.
• Collins says he has unusually high conviction in precious metals because the debt and rate problem looks unsolved.
• He ties gold demand to central banks preferring assets rather than liabilities on balance sheets.
• Gold is described as a debt-free asset that simply sits there.
• Collins says gold is effectively a rock whose dollar price rises as the dollar falls over time.
• Daniel says government expenses are higher than receipts and the debt-service burden is enormous at roughly 40 trillion times the interest rate.
• His core policy objective is simply to get rates lower.
• Daniel says he would ask Trump to end the war to push oil from about 86 to 60, bringing inflation expectations and rates down.
• He says the administration has a six- to nine-month window to engineer lower rates and is using market signaling to pull that forward.
• Collins says the war was a major strategic error from an interest-rate and debt perspective.
• Eisman says he is less focused on the deficit than the others, though he thinks Bessent now has to defend a public line in the sand.
• Eisman explains QE as Treasury-bond buying meant to lower rates and stimulate the economy, but says it mostly helped stocks.
• Collins says the entire fiscal side is wide open because deficits are around $2 trillion, or 6% to 7% of the budget.
• Daniel says DOGE lasted only a month and cut nothing material.
• Eisman says the government will only change course when it gets close to the edge, and that real budget pressure sits in entitlement and defense spending.
• Eisman points to Nvidia customer concentration, saying the top five direct customers accounted for 70% of accounts receivable at quarter end.
• Daniel says a friend saw AI query volume up 30% to 40% while costs fell about 60% because enterprises centralized routing.
• Eisman clarifies that firms are sending only a small share of critical prompts to Anthropic and OpenAI, routing the rest to cheaper open-weight models.
• Daniel says the real question is whether AI capex will produce adequate return on invested capital, and he expects boom-bust dynamics.
• Eisman cites Torsten Slok saying GDP may grow 2% this year and at least half of that may be AI capex.
• Collins contrasts OpenAI and Anthropic with hyperscalers that used to generate huge cash and buy back stock.
• He argues AI companies will eventually be forced public, and that new stock supply is usually bad for markets.
• Eisman says 70% of hyperscaler AI revenue comes from Anthropic and OpenAI, which implies a large share of cloud revenue.
• He says Anthropic’s June-quarter revenue was about 11.5 billion and OpenAI’s was 6.5 billion, with OpenAI growing more slowly.
• Eisman says OpenAI is bleeding cash, losing people, and facing a worsening narrative as a nonprofitable company.
• Collins jokes that an eventual bankruptcy buyer of OpenAI would get a great deal, like past distressed acquisitions.
• Daniel explains the pod model at firms like Millennium and Citadel as centrally allocated capital run by PM pods.
• Eisman says Citadel-style allocators can hand a PM a defined capital slice, such as 250 million.
• He describes a financials pod as a long-short book focused within financials.