Dear Reader,
Mitt Romney reportedly turned roughly $450,000 in retirement savings into as much as $100 million.
Peter Thiel reportedly transformed less than $2,000 inside a Roth IRA into approximately $5 billion, both in less than 2 decades.
Those are extraordinary outcomes – not typical ones – and I would never pretend otherwise.
But they reveal something important:
Wealthy investors often study the account structure and access mechanism before the investment becomes obvious to everyone else.
That lesson matters right now.
Because while ordinary investors chase the same crowded AI names, an enormous stream of capital is moving into the infrastructure beneath the entire industry.
It doesn’t need to invent the next chatbot.
It doesn’t need to replace Nvidia.
Its opportunity comes from the money pouring into the physical and financial foundation AI requires.
And President Trump reportedly recognized the setup early – investing as much as $25 million of his own money.
The danger is assuming you can wait until your employer adds something similar to its retirement menu… or until another AI headline finally makes the opportunity obvious.
By then, you may be trying to evaluate it alongside a much larger crowd.
I’d rather give you the evidence now, while you still have time to consider the complete case calmly.
Good investing,
Alexander Green
Chief Investment Strategist, The Oxford Club
P.S. The research behind this presentation connects the opportunity to a projected $216 trillion AI event. Watch the briefing to see how that figure fits into the investment case.
A $1.3 Trillion Credit Market Just Got Its First Fed Spotlight. Gold Investors Should Pay Attention.
The Federal Reserve does not supervise private credit. It cannot compel disclosures. And until this week, it had no systematic way to measure what is happening inside a market that has grown to roughly $1.3 trillion, comparable in size to the entire high-yield bond market. That changed on November 9, when the Federal Reserve Banks of Dallas and New York opened a pilot survey exploring lending trends in the U.S. private credit direct-lending market, open to eligible firms from that day through November 18.
The timing is not coincidental. Given the growth of private credit and limited visibility into lending activity, the survey is designed to provide insight into the availability of credit, the evolution of lending standards, and the implications for the broader U.S. economy. Aggregate findings are expected to be published in Q1 2027. That is a long lag for a market already showing signs of distress.
Consider what has happened in the past week alone. Bloomberg reported that Dutch bank ING put a roughly $500 million portfolio of data-center loans up for sale in a rare auction, just days after a Democratic sweep of U.S. midterm elections raised doubts about the future of many of the projects. The auction is the first of its kind at that scale to exclusively contain data-center debt, and it arrived amid growing market uncertainty about the level of debt in the sector. Among the loans on the block: a $25 million slice backing the Stargate project data-center site in New Mexico, the location where Oracle sent the developer a force majeure notice.
ING’s willingness to test auction prices on AI infrastructure debt is a signal worth reading carefully. Banks do not run competitive auctions on performing, liquid assets. They do it when they want a market price they cannot otherwise discover and when holding the position feels like the larger risk. Regulators have struggled to assess the potential dangers of private credit to banks due to a dearth of data and the inability to force the unregulated industry to disclose information. The Fed survey is the first formal attempt to close that gap.
The stress is not confined to data centers. The Financial Times reported that listed vehicles managed by KKR and Blue Owl, as well as one run by Apollo Global known as MidCap Financial, were among the funds in which repayments outstripped new lending in the quarter, with executives at KKR pointing to limited dealmaking and its push to exit certain loans. The firm’s listed fund, FS KKR Capital Corp., reported that 7.1% of its loan book was on non-accrual as of June 30, 2026, before falling to 6.4% as of September 30, 2026. Several major lenders have already marked down loans to struggling borrowers, with the Blackstone Private Credit Fund valuing its position in software company Medallia at less than 50 cents on the dollar at the end of June 2026, down from just over 60 cents three months earlier. Ares, meanwhile, carries one of the highest credit concentrations in the sector.
For gold investors, the significance lies not in any single auction or survey result. It lies in the pattern. A market the size of the high-yield bond universe is shifting in ways that remain largely opaque, and the central bank itself only now has a questionnaire to begin understanding it. AI-fuelled turbulence has revived demand for many traditional safe-haven assets as growing concerns about the stability of the AI boom have knocked the S&P 500 and Nasdaq from recent highs. Gold slid below $4,000 an ounce on Friday, October 30, its lowest level in more than a year, despite a broadly risk-off backdrop of falling stock prices, widening credit spreads, and lower bond yields. The yellow metal has recovered ground since, with spot gold at $4,051.50 on November 10, but the disconnect from credit stress is itself informative.
When a trillion-dollar credit market begins to show non-accruals at levels not seen since 2017, when banks auction AI infrastructure debt at open prices, and when the Federal Reserve admits for the first time that it cannot see inside the market clearly enough to monitor it, the conditions that historically precede a flight to hard assets are assembling. The trigger may not be this week. But the structure is being built.