There is a familiar line that surfaces in institutional commentary whenever the world feels tense: the market has already priced in the risk.

It is a reassuring phrase, and a flexible one. Over the years it has been applied to trade wars, pandemics, debt-ceiling standoffs, and banking wobbles. Lately it has been applied to a prolonged closure of the Strait of Hormuz and the spike in oil, freight, and inflation that would follow. The claim, roughly, is that markets are braced for the bad case, so that if tensions ease even modestly, capital can rotate swiftly into the parts of the market everyone has been avoiding.

We want to sit with that phrase for a while, because we think it has quietly become something more than an observation. It has become a permission slip.

An anecdote that stayed with us

Some months ago, on the same day a prominent banking house circulated a note arguing that markets were "priced for persistent risk," Dell Technologies—a large, unglamorous, genuinely real company that sells servers, infrastructure, and services—rose on the order of 30 to 40 percent in a single session on the strength of AI demand. Not a meme stock, not a SPAC, not a token. The corporate equivalent of Walmart, moving like a lottery ticket.

The two things sat oddly together. On one screen, a careful argument that investors were bracing for a dangerous world. On another, one of the most established hardware businesses in America trading as though the future were not only bright but urgent. We kept the image because it captured a tension we have been circling for a while: the gap between how institutions narrate risk and how the market actually prices it.

Two ways to hear the word "risk"

It helps to separate two meanings that ordinary language runs together.

There is the risk that gets narrated—the risk that fills the notes, the scenarios, the strategy pieces. Today that is Hormuz, energy, inflation, and the question of what the Federal Reserve does next. And then there is the risk that gets priced—the compensation the market actually demands to hold assets, expressed in valuations, credit spreads, and the real stance of policy.

The first is a story we tell. The second is a number we pay. The interesting question is never whether risk is discussed, but whether the discussion and the price are pointing in the same direction.

When a note says risk is "in the price," it is making a claim about the second thing while usually describing only the first. And the two, we would argue, are presently pointing in opposite directions.

The tape tells a different story

Step back from the commentary and look at what the market is doing rather than what it is saying. Equities sit at or near record highs. The economy is not limping; it is running warm. And this is not merely a handful of AI names—the strength is broad enough that an investor who owned nothing but the index would be sitting comfortably. Against that backdrop, investors talk openly about rate cuts, central banks have declined to tighten into an inflation reading still meaningfully above target, and the reassuring explanation for hot prints is that they are "just energy," temporary, safe to look through.

This is not the posture of a system braced for a decade of conflict and inflation. It is the posture of a system that has quietly priced the resolution of those risks—of Iran, of energy, of inflation—and layered that optimism on top of a real, but narrow, AI boom. Seen that way, Dell's leap is not an oddity but an emblem: when an industrial-technology company trades like a speculation, you are watching optimism and crowding, not caution, express themselves in the very core of the market.

The same joke, told twice in one week

We were reminded of all this recently when two notes from separate, well-known houses crossed our desk within two days of each other. Both concerned the Federal Reserve. One argued, in effect, that the Fed will not hike, and that market pricing was too worried. The other argued that whether the Fed hikes hardly matters, because the move is small and already absorbed. Read side by side, the two reach the identical destination by opposite roads. If rates hold, own the rally. If rates rise, the increase is digestible, so own the rally anyway. Heads or tails, the recommendation does not change.

That is the quiet comedy of the moment, and it is more instructive than any single day's price move. When the conclusion is fixed and only the reasoning is adjusted to fit, the analysis is no longer testing the world—it is decorating a position already taken. The phrase "it's already priced in" does a great deal of work in both notes, and almost no one stops to ask the obvious follow-ups: priced into which assets, over what horizon, and with which other risks implicitly set to zero.

The elephants the phrase steps around

Grant the most generous version of the argument. Suppose the AI boom and a prolonged energy shock are genuinely opposing forces that roughly cancel at the index level. Even then, the world does not become safe, because a long list of slower, heavier risks simply keeps standing in the room—and almost none of them is meaningfully in any price.

There is a further trap worth naming before the litany, because it cuts against the whole logic of underwriting a fixed future. The market has a habit of building cathedrals to a problem just as the problem quietly dissolves. Recall the sequence: a frontier lab stuns everyone, capital floods toward the assumed bottleneck, and then—DeepSeek, and then DeepSeek again—someone demonstrates the same capability at a fraction of the assumed cost, and the bottleneck everyone was funding turns out to have moved. The same thing is now visible at the physical layer. Even as the consensus insists that power and cooling are hard constraints that justify almost any price for the incumbents positioned around them, teams are already working on photonic interconnects and laser-based cooling to lift inference speed and collapse the energy cost per token.

This is the first contingency, and it runs in the opposite direction from the popular fear: the danger is not only that the solution never arrives, but that it arrives early and cheaply, from an unexpected direction, and strands the very capital that was deployed on the assumption that today's constraint would persist. A portfolio built to be paid for scarcity can be quietly repriced the moment scarcity is engineered away. The market underwrites the problem as permanent; problems rarely are.

And here is what should genuinely frighten anyone paying attention: the risks that remain are not tail scenarios or clever contrarian talking points. They are large, visible, and compounding in plain sight, and the market has decided, more or less in unison, to look straight through every one of them. That is the part that should not be normal.

The quiet arithmetic of inflation and rates

Start with inflation, and stop pretending the argument is about next month's print. Over the past decade the dollar has lost roughly a third of its purchasing power—a dollar today buys about seventy-two cents of what it bought ten years ago. Read that slowly, because the comforting language of "cooling" and "disinflation" obscures what it actually means: prices are not coming back down, they are merely rising a little less quickly from a level that already robbed a third of the currency's value. This is not a footnote. It is a slow, silent expropriation, and the market treats it as background noise.

And it compounds a second error, one we hold with some conviction: the persistent belief that rates are about to fall. We have been told to extend duration for four years running, and the trade that actually worked was the opposite—staying short, staying patient, and declining to bet that cuts were imminent. If rates stay higher than the consensus expects—and the consensus has been wrong in that direction for years now—then every long-duration asset priced off a benign discount rate is mismarked, and the erosion above continues to fall hardest on savers, retirees, and anyone whose income did not sprint to keep pace. "Higher for longer" is not a slogan to us. It is the base case the market keeps refusing to underwrite.

The debt nobody wants to add up

The sovereign balance sheet is worse, not better, and almost no one wants to say it plainly. Federal debt now sits at levels the country has not carried outside of total war. The deficit is structural, not cyclical—it does not close when the economy is good, which means it will not close at all without a rupture. Interest expense has grown so large that we increasingly borrow simply to pay the interest on what we already borrowed, the definition of a trap rather than a cycle. And a meaningful share of that debt sits in the hands of a strategic rival, which means some portion of our cost of capital is a decision made in Beijing, not Washington.

Beneath the sovereign line sits the household ledger, and it is straining in ways the headline numbers flatter. Student debt has swollen past a trillion and a half dollars, deferring the milestones—homes, marriages, children—that used to mark early adulthood. Credit card balances have pushed to record highs at interest rates that would once have been called usurious, which is what borrowing to cover the cost of living looks like when wages lag prices. Auto loans have stretched to absurd lengths, with delinquencies among younger and lower-income borrowers already climbing. Layered atop it all is a housing market that has categorically priced an entire generation out of ownership, severing the primary ladder by which ordinary families built wealth for eighty years. None of this is priced. It is assumed away by models that still treat perpetual growth and benign real rates as a birthright—models that quietly assume the consumer can absorb whatever "higher for longer" actually costs.

The private-credit edifice

Then there is a private-credit market that has never seen a cycle. A corner of finance that was a niche a decade ago—a few hundred billion in assets—has ballooned roughly tenfold to around $3 trillion today, on a path some project toward $5 trillion by the end of the decade. It is sold as equity-like returns with bond-like risk, which is a contradiction in terms, and the underwriting has visibly degraded as the asset-gathering accelerated: a large share of payment-in-kind loans are marked above 95 cents on the dollar, with regulators openly questioning whether those marks are real. These loans have largely never defaulted at scale, because the market has never lived through a true downturn. We are pricing a default experience that has not happened yet—and when it does, it will arrive in the least liquid, most generously self-marked corner of the system.

The secondaries mirror

Sitting right beside private credit, and rhyming with it, is the explosion in secondaries—the machinery built to manufacture liquidity for private assets that were never designed to be liquid. The uncomfortable truth is that the exit routes these assets depend on—the IPO window, the M&A cycle—are themselves cyclical and, at present, largely shut. When the only way out is a secondary sale, and secondary volume becomes a growing share of all exits, that is not a sign of a healthy, self-clearing market. It is a sign of a mismatch between what investors were promised and what the asset can actually deliver—the same ballooning, same-story dynamic as private credit, wearing a slightly better suit.

The tell is not in the marks

It is in the behavior of the people who own this paper. Across these private markets, limited partners are now tugging at their managers' sleeves demanding liquidity—distributions, exits, their money back—on assets they explicitly agreed to lock up for a decade or more. When you cut a check into a ten- or twenty-year vehicle, that is the deal: you do not expect to see anything until the far end, and you cannot stomp your feet in year three because the liquidity you never bargained for hasn't appeared. Yet that is precisely what is happening, at scale, and an entire industry—secondaries, continuation vehicles, semi-liquid "evergreen" wrappers sold through the bank channel—has sprung up to manufacture an exit these assets were never built to provide.

This raises a genuinely awkward question about the demand side. Every one of these investors signed a subscription document attesting that they were accredited, or qualified—that they understood the illiquidity, the lockup, the risk of total loss. The stampede for early liquidity is the clearest evidence that many of them did not. They wanted the yield and the private-market cachet without underwriting the one feature that defines the asset class: you cannot get out. If you needed the money to be there, the answer was always available and always cheaper—you could have bought a stock. Instead, a great deal of capital signed up as long-term, patient money and is now behaving like short-term, impatient money, and the machinery being built to bridge that gap is itself the mismatch, dressed as a solution.

We are not saying these asset classes, private credit or secondaries, have no reason to exist; a well-sourced, deeply diligenced secondary bought at a real discount to a real mark can be a fine transaction - a real project, with real cash flows, and real underwriting may be a great vehicle of risk-adjusted-asset backed credit. But too much of what is marketed today is someone else's dirty shirt—a fund of a fund of a fund, layered and re-layered, sold on the promise of liquidity rather than the merit of the underlying. When a vehicle pitched as balanced turns out to be ninety-plus percent secondaries, the label and the contents have parted ways. That gap—between what a thing is called and what it is—is the recurring signature of a market that has grown too big, too fast, for its own plumbing.

The rest of the room

And these are not the only occupants. Start with the capex wave itself: hundreds of billions poured into data centers, chips, and power on the faith that a return will eventually justify it—a return that is, as yet, simply not there.

Then the geopolitical fractures, which do not resolve because one strait reopens. A deepening axis between Russia and China. China and Taiwan. Russia and Ukraine. Iran and the Gulf. The steady hardening of the BRICS bloc into something that looks less like a slogan and more like an alternative settlement system. The Middle East, perpetually one miscalculation from a wider war. None of these is a single event to be priced and cleared; they are slow-grinding structural fractures, and the market treats each as a headline to fade rather than a fault line to respect.

And then, closest to home, the domestic fractures. The wealth gap, most glaringly in your face every night. The social division that has curdled into something the country has not carried since the last time people spoke, however loosely, of civil war. These are not market risks in the tidy sense—no spread to widen, no curve to steepen—but they are the ground the market stands on, and the ground is not steady.

Any one of these, in a saner moment, would dominate the conversation. Instead they have been filed, collectively, under "priced in"—waved past with the same three words used to dismiss a spike in Gulf freight rates. To call a system braced for persistent risk while it steps around all of this at once is not analysis. It is a decision to not look. The elephants did not leave the room. We simply agreed to stop seeing them.

The contradiction hiding inside the AI trade

There is a sharper version of the same point buried inside the very trade everyone loves. The AI build-out is not a clean software story; it is a power and infrastructure story—steel, copper, transformers, transmission, and above all electricity, at a scale the country has not attempted to add in a generation. Every time the market pays up for a server maker on AI guidance, it is quietly underwriting that abundant, reliable power will be there at tolerable cost, that grids and permitting will clear, and that the inflation thrown off by all this spending will stay manageable.

Now hold that beside the insistence that the market is braced for a persistent energy shock. Both cannot be true. Either energy is scarce and dangerous, in which case the most energy-hungry trade of the cycle should be de-rating rather than melting up—or energy is abundant and benign, in which case the bracing is theater. The market is trying to hold both beliefs at once: pricing the AI complex as if power were free, while narrating the macro as if power were the central threat. It wants the upside without underwriting the downside that is its precondition.

Why this echoes what we wrote before

Earlier this year we wrote about the quiet risk of homogeneous deal flow—what happens when nearly every pitch becomes an AI story regardless of the underlying business, and capital begins chasing narrative rather than scarcity. The pattern we are describing here is the public-market cousin of that private-market observation. The comforting myth is that herding is a retail phenomenon, the stuff of cab-driver stock tips. The less flattering truth is that herding is human, and professionals simply dress it in better language. The echo chamber has moved upstream, into memos and committees and portfolio reviews. By the time a theme saturates the public, it has long since saturated the institutions—which is exactly why two respected houses can, in the same week, arrive at the same conclusion by contradictory means and call it independent analysis.

None of this means AI is a mirage or that a crash is imminent. AI is a real shift, and exponential progress in technology is real. But exponential progress in technology does not repeal cyclicality in capital. Adoption can accelerate while valuations overshoot, while capital is misallocated inside the AI complex itself, and while everything else stays underfunded. Believing "this time is different" because the technology is real is precisely how investors confuse inevitability with investability—and narrative with risk premium.

What we take from it

We are participants, not spectators, and we have no interest in calling a top. Our takeaway is quieter and more durable. When someone tells us the risk is priced, we treat it as a hypothesis to be tested rather than a fact to be accepted, and we ask which assets, over what horizon, and which other risks have been set to zero to make the sentence true. Narrated risk tells you what the crowd is worried about. Priced risk tells you what it is actually willing to pay for. When the two drift apart—when the risk lives mostly in the story and the resolution mostly in the price—it is usually a good moment to think less about how much more there is to make, and more about how many assumptions one is underwriting without being paid for them.

Not because we know which song is the last one, but because we do not need to dance to all of them.