Markets today are running a high-stakes experiment in selective perception. The AI-led surge in capex and stock valuations has become the dominant narrative, casting a long, shimmering shadow over less glamorous but potentially far more consequential developments beneath the surface, and the clock is ticking. Two of the most pressing risks are institutional decay at the top of the US economic system and a brewing storm in the unregulated private credit market. Neither fits neatly into an AI-hype story, which is precisely why they deserve closer attention.

Put simply, we’re in a version of Speed, with our bomb armed and no clear Keanu Reeves to disarm the looming disaster with blank-faced charm. The AI boom is strapped to the underside of the market and the whole system is being driven at a breakneck pace, not necessarily toward a destination, but compulsively away from stillness, which means away from scrutiny, pricing fundamentals, and confronting the unmistakable fragility propping up the broader system. As long as capital keeps flowing and data centers keep humming, the ride continues. But if this machine ever slows down, we’re going to quickly find out what’s duct-taped beneath our bus.

Private Credit: The Hidden Risk Pileup

The First Brands bankruptcy, recently filed in Delaware, is easy to dismiss as a routine corporate failure. That would be a mistake. The auto-parts distributor, heavily backed by private credit, reportedly pledged the same assets to multiple lenders. This basic violation raises red flags not just about the borrower, but about the system funding it.

Private credit has ballooned to over $1.7 trillion, largely outside the regulatory framework that governs traditional banks. What’s emerged is a market built for velocity, not durability. Underwriting standards have eroded under pressure to deploy capital, while financial engineering has replaced credit discipline. Speed is rewarded, scrutiny sidelined. But the deeper issue isn’t just defaults; it’s feedback loops. Lenders roll loans forward, inflate collateral, and use those inflated values to justify the next round. It’s not just risky, but circular: money in, valuations up; rinse, repeat.

We shouldn’t assume this is isolated. A significant share of current credit exposure likely rests on assumptions that won’t survive real stress. Capital flowing into AI may be masking these risks for now, but it won’t prevent the fallout if the private credit structure begins to unwind. The system is being held together by momentum; the real question is what happens when it hits a pothole.

The Circular AI Economy: A Trillion-Dollar Loop

Look closely at today’s AI ecosystem and you start to see how capital is cycling through a tight loop of players, each dependent on the next to justify their own bets. A snapshot of the current structure looks more like a daisy chain than a diversified system:

What emerges is not a traditional supply chain but a circular financial structure where customer, supplier, investor, and partner are often the same entities, rotating hats depending on the press release. Beyond industrial concentration, this is effectively a closed system where valuations are reinforced by strategic demand, not organic profitability.

The Range of AI Outcomes

A Politicized Fed: Crisis Without a Compass

At the same time, another risk is quietly escalating: the politicization of the Federal Reserve. Historically, the Fed has served as the grown-up in the room: not always perfect, but independent enough to stabilize markets in times of stress. The concern today is the degree to which the traditional separation between the presidency and the Fed is being tested.

President Trump has moved quickly to assert more direct influence over the institution. His administration has signaled an intent to nominate officials whose views align closely with its policy preferences, a departure from the long-standing norm of prioritizing independent expertise. (The Supreme Court recently noted in an order that Federal Reserve governorships occupy a uniquely structured role distinct from standard presidential appointments, given the Fed’s quasi-independent design and long, staggered terms. That recognition underscores how unusual the current dynamic is.)

This isn’t hypothetical. The seeds of systemic collapse are often planted long before the crisis. It might not be catastrophic during calm conditions, but crises are defined by uncertainty, and that’s when independent central banking matters most. Imagine a scenario where inflation resurges and the Fed’s instinct isn’t to tighten based on data, but to respond to short-term political pressure; beyond economic risk, it becomes a credibility issue.

Once the Fed’s credibility is lost, it’s not easily rebuilt; and in a confidence-based system, trust is the ballast. If the Fed bends, or is compelled to, the signal to markets won’t be subtle.

Politicians want lower rates and faster growth now; the Fed’s job is to say no when the long term demands it. But if it can’t or won’t, the consequences won’t be measured solely in headlines.

Echoes of 1929

In his latest work, 1929: The Crash and Its Aftermath, Andrew Ross Sorkin unpacks a world that believed itself too advanced, too modern, too financially sophisticated to fail. The markets in the late 1920s had new tools, new capital flows, and new kinds of investor behavior. What they lacked was a true understanding of how circular leverage, or money loaned against inflated asset values used to justify more loans, distorts the perception of safety.

We’re not just repeating history; we’re accelerating it. As Sorkin illustrates in his book, it was never the excess alone that brought the system down; it was the unwillingness to recognize that the excess had become the system.

What made 1929 catastrophic wasn’t the initial shock but the cascade. A collapse in margin lending triggered bank failures. Bank failures triggered credit withdrawals. And with no backstop in place (no effective central bank action, no coordinated response), the system fractured under its own opacity.

The same fragilities exist now, just with different names and technology. Thomas Lamont, the Morgan partner who tried to reassure markets with soothing public statements even as cracks deepened behind the scenes, has his analogues in the various corporate leaders and asset managers today who frame every correction as a buying opportunity and every new pump inflating the bubble as misunderstood innovation. Back then, Lamont told the press the market was fundamentally sound. Today, we hear that “AI is different,” “multiples don’t matter,” and that “this time, infrastructure justifies everything.” The language has certainly changed; the incentives haven’t. Lamont’s reassurances didn’t survive the crash, and neither did his reputation. Optimism has a short half-life when reality shows up.

And the worry is not just valuation. It’s the opacity. Private credit isn’t transparent and AI valuations aren’t stress-tested, the Fed’s autonomy is in question, and the financial system is too interwoven to unwind cleanly when the narrative shifts. What Sorkin reminds us, and what markets often forget, is that crises don’t start with something as straightforward as fraud. They start with belief. Belief in systems too complex to fail, too big to regulate, too modern to collapse. Until, all of a sudden, they’re not.

Two Economies, One Illusion

In her recent New York Times piece, “There Are Two Economies: A.I. and Everything Else” (Oct. 2025), Natasha Sarin outlines the disconnect between a booming AI narrative and weakening economic fundamentals clearly: One economy is visibly surging, fueled by investment in GPUs, cloud compute, and synthetic intelligence; the other, defined by consumer debt, flat productivity, and institutional drift, is quietly eroding.

None of this fits the current mood, which is still defined by tech optimism and a belief that AI will fix, or at least outgrow, whatever macro headaches come along. But just because markets are celebrating innovation doesn’t mean the foundations beneath them are stable. The more performance depends on one sector and one story, the more brittle the whole structure becomes and the more momentum matters, however illusory.

AI technology may be the future, but it won’t stop a credit unwind, it won’t neutralize institutional decay, and it won’t save a market that suddenly decides trust is gone: trust in balance sheets, trust in central banks, or trust in the system itself. And we place that trust too easily in those who have the most to gain from keeping the speedometer at the breaking point.

The deeper question isn’t AI’s price. It’s whether investors understand that the “market” they think they own is really a narrow slice of it, dominated by a handful of AI-linked stocks. The broader market, made up of companies untouched by this narrative and momentum, looks nothing like the benchmark that claims to represent it. The danger isn’t a broad market collapse; it’s whether investors grasp that the index they trust is now a concentrated bet waiting for a few bad days.

Tick, tock.