The Great Disconnect: Why the Stock Market Is Ignoring Everything That Should Matter
Equity markets sit at all-time highs while institutional credit desks buy record amounts of default protection — a contradiction that history suggests resolves violently, and soon.
The S&P 500 has gained $6 trillion in market cap in just over a month, even as oil sits at $125, core inflation accelerates past 3%, private credit redemptions are being gated, and the equity risk premium has compressed to generational lows. This deep dive maps the mechanical forces actually driving prices, what credit markets see that equities don't, the historical record of how these divergences resolve, and the specific trigger calendar — starting next week — that could break the equilibrium.
This publication is for informational and educational purposes only and does not constitute financial, investment, or trading advice. The analysis, opinions, and commentary presented here should not be interpreted as a recommendation to buy, sell, or hold any security. Always conduct your own research and consult a qualified financial advisor before making investment decisions. Past performance does not guarantee future results.
Here is a fact that should bother you more than it does.
On May 1, 2026, the S&P 500 closed at an all-time high. On the same day, the put/call ratio on HYG — the iShares high-yield corporate bond ETF, the single most liquid instrument for expressing a view on credit risk — also closed at an all-time high. Not near it. At it.
The same institutional investors who are buying stocks at record prices are simultaneously buying record amounts of insurance against corporate defaults. They are, in the same portfolio, expressing the view that everything is fine and the view that everything is about to break. One of these positions will be wrong. The question is which one, and what happens to everyone else when the answer arrives.
The instinct is to call this cognitive dissonance, but that’s too generous. Cognitive dissonance implies the person holding two contradictory beliefs is unaware of the contradiction. These are not unsophisticated actors. The desks buying equities and the desks buying credit protection often sit on the same trading floor, sometimes report to the same CIO, occasionally share the same Bloomberg terminal at lunch. They know. They just disagree about timing.
And timing, in markets, is not a detail. It’s the whole game.
So let’s lay out what the equity market is choosing to ignore, or at least to defer worrying about. Oil is at $125 a barrel, driven by the effective blockade of the Strait of Hormuz and the cascading sanctions regime that has turned the Persian Gulf into an insurance nightmare. Core PCE — the Federal Reserve’s preferred inflation gauge — printed at 3.2%, more than a full percentage point above target and accelerating. Q1 GDP came in at 2.0%, down from 3.1% two quarters ago. Consumer sentiment sits at 53.3, a level previously associated with recessions that have already started. There is an active shooting war in a region that controls 20% of global oil transit. The new Fed chair hasn’t held his first meeting yet and has already signaled that forward guidance — the single most important innovation in central banking communication over the past two decades — is over.
Against this backdrop, the S&P 500 has risen 14.2% in 23 trading days. The index has gained roughly $6 trillion in market capitalization in slightly over a month.
And here is the number that frames everything that follows: the equity risk premium — the additional return that stocks offer over risk-free Treasury bonds — is, on a trailing-earnings-yield basis, negative 92 basis points. A 10-year Treasury currently yields more than the S&P 500 earnings yield. Investors are paying for the privilege of taking equity risk.
A note of precision is required here, because the ERP is one of those metrics that changes shape depending on how you hold it. Use forward earnings estimates instead of trailing, and the premium is thin but positive. Use TIPS-adjusted real yields and the picture shifts again. On certain constructions, the ERP has flirted with zero multiple times in the past five years without the predicted catastrophe arriving. The bears who declare “this has only happened twice before — 2000 and 2007” are overstating the measurement’s precision. But the bears who say “the compensation for owning equities over Treasuries is at or near its lowest level in a generation” are not overstating anything. Every construction is compressed. The directional signal is consistent across all of them. And what is not ambiguous is what happens if the 10-year yield rises further: at 5%, the ERP drops deeper into negative territory on every construction, into ranges that have no historical precedent of resolving benignly.
The question this piece attempts to answer is not whether the market is “right” or “wrong.” Markets are not sentient beings with opinions; they are the emergent output of millions of individual decisions, many of them made by algorithms that do not have opinions at all. The question is more specific: what is actually driving prices, who is on each side of the trade, what does the historical record say about how this resolves, and what are the specific events in the next sixty days that could break the equilibrium?
The answers are more interesting than the question.
The Surface Explanation
There is a version of this story that is simple, intuitive, and reassuring. It goes like this.
The market sold off sharply in late March 2025 on fears that the tariff regime would actually stick. Liberation Day produced a 12% drawdown in roughly a week. And then the president blinked. Tariffs were paused, modified, exempted, renegotiated, or simply not enforced, depending on the country, the product, and the day of the week. The pattern was familiar enough that traders gave it a name: TACO. Trump Always Chickens Out.
TACO is not an investment thesis in the traditional sense. It does not appear in any SEC filing or investment policy statement. But it is, functionally, the dominant behavioral model driving positioning. It says: the worst-case scenario never materializes. Every threat is a negotiating tactic. Every crisis is a buying opportunity. And the evidence since January 2025 supports it. Every selloff has reversed. Every dip has been bought. The conditioning is nearly perfect.
The second pillar is AI — not AI as a concept but AI as a line item on an income statement. Microsoft’s Azure AI services are generating $37 billion in annualized revenue. Google Cloud crossed $12.3 billion in a quarter, up 63%. Amazon Web Services reported $29.3 billion. These are audited numbers from 10-Q filings, not projections from a pitch deck.
The third pillar is the labor market: initial jobless claims at 189,000, a 57-year low. Wage growth moderating from 3.9% to 3.5% — the Goldilocks zone for corporate margins. Consumer spending grew 2.4% in real terms despite sentiment at 53.3.
The surface explanation says: the market is correctly weighting earnings and employment over headline noise about oil and inflation. It is forward-looking, as it always has been, and what it sees forward is AI-driven earnings growth, a resilient consumer, and a policy regime that always moderates at the brink.
This explanation is not stupid. Parts of it are correct. The market IS forward-looking. AI earnings ARE real. The labor market IS strong. If you had bet, at any point in the past sixteen months, that the selloff would reverse and the market would make new highs, you would have been right every single time.
But here is the thing about being right every single time: it is the precondition for being catastrophically wrong once. The people who bought every dip in January 2008 had been right for seventeen consecutive months before they were ruined. The question is not whether the surface explanation contains truth. It does. The question is whether it contains enough truth to justify the most compressed equity risk premium in a generation, record positioning, and the complete dismissal of the most aggressive credit-protection buying in the history of financial markets.
If the surface explanation were sufficient, institutional credit desks would not be doing what they are doing. And what they are doing deserves its own section.
The Earnings Case — Why the Bulls Have a Point
Before we get to the prosecution, let’s give the defense its full hearing. Because the bull case is not just plausible. It is, by certain metrics, the strongest earnings case in a generation.
The Magnificent Five — Apple, Microsoft, Google, Amazon, and Meta — reported a combined $541.8 billion in quarterly revenue for the most recent period. That is up 19.4% year over year. Not estimated. Not projected. Filed with the SEC, audited by Big Four accounting firms, subject to Sarbanes-Oxley certification by the CEO and CFO personally. These are real numbers.
Apple posted $111.2 billion, a record. The services business now accounts for over a quarter of total revenue and is growing faster than hardware. Microsoft came in at $82.9 billion, up 18.3%, with the AI business alone annualizing at roughly $37 billion and growing at 123% year over year. That is a business that did not meaningfully exist three years ago and is now larger than many S&P 500 companies in its entirety.
Google Cloud’s 63% growth is worth lingering on because of what it represents structurally. For a decade, the knock on Google was that it was a one-trick pony — an advertising company masquerading as a technology company. At a $49 billion run rate and accelerating, the cloud business changes that math. Amazon reported $181.5 billion in total revenue. AWS margins expanded again. Even Meta, which spent $36 billion on AI infrastructure in the past year, is converting that investment into 33% revenue growth by making its ad-targeting algorithms measurably better at predicting what you want to buy.
The collective AI capital expenditure budget across these five companies now runs at approximately $725 billion annualized — by a wide margin, the largest corporate capital investment program in history. And here is the critical difference between this cycle and the telecom bubble of 1999: the telecom companies were funding their capex with debt. The hyperscalers are funding theirs with cash flow. Apple has $162 billion in cash. Microsoft generates $23 billion in free cash flow per quarter. Amazon’s operating cash flow runs at $114 billion annualized. These companies could fund their AI buildout for years without accessing a single credit market.
This matters because the AI investment cycle is not vulnerable to the mechanism that killed the telecom bubble — the moment when lenders stop lending. You cannot have a credit crunch in a sector that doesn’t use credit. The bears who compare AI capex to fiber-optic overbuilding are drawing the wrong analogy. A better analogy is the early buildout of cloud computing itself, circa 2010-2014, when AWS was spending seemingly absurd amounts on data centers and skeptics questioned whether demand would ever materialize. It materialized. The skeptics were wrong for a decade.
The labor market reinforces the earnings case through a different channel. Initial claims at 189,000 are the lowest since 1968, when the labor force was less than half its current size. The wage moderation from 3.9% to 3.5% threads the needle between consumer health and corporate margins. Workers are getting thin but positive real wage gains while companies face less cost pressure. This is the economic sweet spot that rarely persists but, when it does, produces exactly the kind of earnings growth we are observing.
Consumer spending tells the same story from a different angle. Real personal consumption grew 2.4% in Q1. Retail sales are up 4.2% year over year. Auto sales are running at 16.7 million units annualized. Credit card delinquency rates are actually improving — 2.94%, down from 3.06%. Debt service ratios are stable at 11.3%. The consumer SAYS they feel terrible. The consumer’s credit card statement says otherwise.
The bulls look at all of this and say: the S&P 500 is not at all-time highs because the market is irrational. It is at all-time highs because the five largest companies in the index are generating the fastest revenue growth in their histories, consumers are spending, workers are employed, and the tax code is about to get more favorable. The macro headwinds are real but the earnings are realer.
This is a defensible position. It is also incomplete. Because it treats the stock market as a fundamentals-pricing machine, and the stock market in 2026 is something considerably stranger than that.
The Plumbing — What’s Actually Moving Prices
If you want to understand why the S&P 500 is at all-time highs, you can study the earnings reports, or you can study the plumbing. The earnings reports will tell you a story about AI and margins and consumer resilience. The plumbing will tell you the truth.
Start with the autopilot bid. Every two weeks, roughly 80 million American workers receive a paycheck. A portion of that paycheck — typically 6% to 10%, sometimes more — is automatically deducted and sent to a 401(k) plan. That money flows into target-date funds, which allocate mechanically between stocks and bonds based on the participant’s expected retirement date. The stock allocation, for anyone under 50, is typically 80% or more. And that allocation does not change based on valuations, earnings, geopolitics, oil prices, inflation readings, consumer sentiment, credit spreads, or the equity risk premium. It does not change at all. It buys.
The aggregate scale of this flow is approximately $50 to $60 billion per month. Every month. Regardless. It is the financial equivalent of the tide: relentless, predictable, and indifferent to the weather.
To appreciate the significance of this, consider what happened during the most recent week of mutual fund outflow data. ICI reported that actively managed mutual funds experienced net redemptions of $19.5 billion. This was reported as evidence of “investor nervousness.” What was not reported, because it is not dramatic enough for a headline, is that passive ETF and index fund inflows during the same period exceeded $25 billion. The active money was selling. The autopilot money didn’t notice. The market went up.
This is arguably the single most important structural change in equity markets over the past twenty years. When roughly half of all equity assets are held in vehicles that buy regardless of price, the traditional mechanisms by which markets correct — human beings deciding that stocks are too expensive and selling — are fundamentally impaired. The marginal price-setter in the equity market is increasingly not a human being making a judgment about value. It is an algorithm following an allocation formula written in 2003.
Now layer on the buybacks. S&P 500 companies have authorized over $1 trillion in share repurchases on an annualized basis. Apple alone has a $110 billion authorization outstanding. These programs operate on roughly similar mechanical logic to the passive flows: they execute steadily, on schedule, with relatively little sensitivity to price. There is one critical exception: the blackout period. Companies cannot buy back shares during the weeks surrounding their earnings announcements. The most recent blackout period ended, across much of the S&P 500, in the last week of March 2026.
The S&P 500 bottomed on March 30, 2026. The rally began on March 31. The temporal alignment between the end of the buyback blackout and the start of the rally is striking, and the mechanism is straightforward: the largest single buyer in the equity market was prohibited from buying for several weeks, and the buying resumed precisely when the prohibition lifted. Other things happened too — the ceasefire, earnings beats — but the blackout calendar provides a mechanical explanation for the timing that narratives about sentiment do not.
The third mechanical force requires a brief detour into the options market. When investors buy put options for protection, the dealers who sell those puts must hedge by selling stock as the market falls and buying stock as the market rises. This creates a self-reinforcing loop. In late March, the VIX was above 30 and dealers were short gamma. Then vol began to fall. As it fell — from 30 to 25 to 20 to 15 to the current 12.7 — dealers were forced to buy stock to rebalance their hedges. Every tick lower in vol required more buying. The buying pushed prices higher, which pushed vol lower, which required more buying. This is the vanna/gamma feedback loop, and it is as close to a perpetual motion machine as financial markets produce.
The explosion of zero-days-to-expiration options has amplified this considerably. 0DTE options now account for roughly 45% of all S&P 500 options volume. Because they have almost no time value, their gamma is extreme: small price moves produce large hedging flows. The feedback loop that would have taken weeks to play out a decade ago now plays out in hours.
Fourth: the CTA flip. Commodity Trading Advisors are systematic trend-following funds managing an estimated $350 to $400 billion. They buy things that are going up and sell things that are going down. They do not care why. After the March selloff, most CTAs were short or neutral equities. As the rally began, they hit their trigger levels and flipped to long. By mid-April, most models were at maximum long equity exposure. The estimated one-time flow from the short-to-long flip was $50 to $70 billion. Like the other mechanical forces, CTAs are entirely indifferent to fundamentals. They are buying because the price is going up. The price is going up partly because they are buying.
Fifth: the TACO trade as behavioral economics. When retail investors bought $3 billion net during a 5% down day in April 2025, they were not making a calculated assessment of discounted future earnings. They were responding to a stimulus with a conditioned response that had been reinforced by consistently positive outcomes over the preceding sixteen months. The conditioning works like any conditioning: the pattern repeats, the response strengthens, the organism stops evaluating and starts reacting. Retail now accounts for approximately 20% of daily equity volume. This is broad-based buying across index ETFs and mega-cap tech — rational in the sense that it has been consistently profitable, fragile in the sense that all conditioning eventually encounters a stimulus that does not produce the expected reward.
Finally, risk parity and volatility-targeting strategies mechanically increase equity exposure when volatility is low. As the VIX has compressed to 12.7, these strategies have been forced to add equity exposure, probably $30 to $50 billion in aggregate.
Add it all up. Passive autopilot flows: $50-60 billion per month. Buyback resumption: enormous. Vanna/gamma dealer buying: self-reinforcing as long as vol falls. CTA flip: $50-70 billion one-time. Retail dip-buying: $3 billion+ on down days alone. Risk parity rebalancing: tens of billions. My rough estimate is that fundamental re-rating based on AI earnings accounts for 35% to 40% of the rally from the March lows. The remaining 60% to 65% is mechanical amplification: flows that are indifferent to the reasons and responsive only to the direction.
This matters for one reason above all others. Every single one of these mechanical forces works in reverse — with a caveat. The passive autopilot bid is partially a one-way ratchet: 401(k) contributions don’t stop in a downturn, which creates a genuine floor that did not exist in 1973, 2000, or 2007. But target-date funds do rebalance away from equities as prices fall, and ETF holders can and do redeem. The floor is real but permeable. It prevents flash-crash capitulation. It does not prevent a grinding 20% decline over months. And every other mechanical force — CTAs, dealer hedging, risk parity, vol-targeting, retail conditioning — amplifies the downside with exactly the same indifference it brought to the upside. The dealer who was forced to buy as vol fell will be forced to sell as vol rises. The CTA that went max long will go max short. The retail investor whose conditioning says “buy the dip” will, after the first dip that doesn’t bounce, discover what psychologists call extinction — the violent unlearning of a response when the reward disappears.
The market that gained 14.2% in 23 trading days is capable of losing 14.2% in 23 trading days. Not because the fundamentals deteriorated that quickly — fundamentals rarely do — but because the mechanical forces that drove prices higher are the same forces that will drive them lower, and they do not know or care about the difference.
What Credit Markets See That Equity Markets Don’t
There is an old adage in finance that equity markets are for optimists and credit markets are for realists. Like most old adages in finance, it is both oversimplified and directionally correct. Equity investors get the upside and can tell themselves stories about growth and transformation. Credit investors get a fixed coupon and the return of their principal, and their entire analytical framework is organized around one question: can this borrower pay me back? When the two markets diverge sharply — when one is partying and the other is heading for the exits — you should pay attention to the one that is heading for the exits.
The credit market is heading for the exits.
Credit default swap trading volumes in the first quarter of 2026 hit what industry sources report as an all-time high. Traders Magazine, citing industry data, put the figure at $4.5 trillion — up 69% from a year ago. The precise number depends on whether one measures traded notional, gross notional outstanding, or net exposure, and the primary data sources (DTCC, ISDA, BIS) have not yet published their quarterly figures. But the direction is not in dispute. Someone — many someones, at many institutions — is spending enormous sums to protect against corporate defaults.
It gets more specific than that. In April, S&P Dow Jones Indices launched the CDX Financials index, a new CDS index covering 25 North American financial entities — banks, insurers, REITs, and, for the first time, business development companies linked to the private credit market. Apollo Debt Solutions, Ares Capital, and Blackstone Private Credit Fund collectively make up 12% of the equally weighted index, according to Reuters. The inclusion of BDCs in a standardized CDS index is itself a data point: you do not add a new asset class to a credit derivative product without institutional demand for protection against it. Banks including BofA, Barclays, Deutsche Bank, and Goldman Sachs are market-making.
The private credit universe gives them reason. The redemption crisis — and “crisis” is the correct word — has been building for three quarters and is now acute. Across the major non-traded BDCs and private credit funds, $19.5 billion in redemptions were requested in Q1, according to a Business Insider analysis of SEC filings. Only 53% were fulfilled. The remainder are queued, gated, or quietly being discouraged through the kind of polite but firm investor relations communications that amount to: we have your money and we are not giving it back right now.
The fund-level data illustrates the dispersion. Blue Owl Technology Income Corp saw 40.7% of shares tendered for redemption in Q1, per Bloomberg — four out of ten dollars trying to leave. Blue Owl Credit Income Corp saw 21.9% of shares tendered, with total requests of approximately $4.2 billion, of which $3.2 billion was blocked. Blackstone’s BCRED, the largest non-traded BDC, processed a record $3.8 billion in redemptions in a single quarter, representing 7.9% of assets, according to CNBC. Goldman Sachs’ private credit fund reportedly saw requests come in just below the 5% gating threshold — close enough to the contractual limit that the precision raises questions about whether the number was entirely organic.
The performance data explains the pressure. FS KKR Capital — a major BDC co-managed by two of the most storied names in leveraged finance — was downgraded to junk (Ba1 from Baa3) by Moody’s on March 23, citing worsening asset quality and non-accrual loans at 5.5% of the portfolio. Fitch followed with its own junk-territory downgrade on April 9. The fund had reported a net loss of $114 million in Q4 2025. Eagle Point Credit, a CLO equity fund that sits further down the capital structure, cut its dividend by 57% and reported that its net asset value had declined 31.8% over the year.
Morgan Stanley strategist Joyce Jiang published a note in March projecting that private credit direct lending default rates could reach 8% — approaching pandemic-peak levels, versus the 2-2.5% historical average. (Morgan Stanley characterized the projected spike as “significant but not systemic.”) The underlying distress is concentrated in software companies — the darlings of the 2020-2021 private equity boom, when every SaaS business with a pulse was leveraged 7x to fund a take-private. Software now accounts for $25 billion in distressed private credit, representing 31% of all distressed loans but only 13% of the market. That concentration is a tell. It means the pain is not yet dispersed. It is gathering.
Commercial real estate tells a parallel story. The CMBS distress rate hit 12.07% in March — an all-time high, per CRED iQ. Office vacancy rates are at 21%, also a record. An estimated $875 billion in commercial real estate loans mature in 2026 and must be refinanced at rates that are, in many cases, double or triple the original coupon. The math on many of these properties does not work. And the banks holding this exposure — particularly regional banks that loaded up on CRE during the low-rate era — are looking at a wall of maturities with underlying collateral that has declined 30% to 50% in value.
The bank-to-private-credit nexus is the contagion channel that connects all of this. Major banks have extended approximately $300 billion in credit facilities to private credit funds, per Moody’s. Goldman Sachs alone has $118 billion in exposure to non-bank lenders, per Reuters. If the underlying portfolios deteriorate, the banks face a choice between enforcing covenants (triggering forced liquidation) and extending and pretending (increasing eventual losses). Banks are historically bad at making this choice well.
Meanwhile, high-yield spreads sit at 2.83% over Treasuries. For context: this is comparable to the levels that preceded the 2007 financial crisis, when HY OAS hit an all-time low of approximately 240-260 basis points before blowing out to 2,000+ basis points by late 2008. It is significantly tighter than the ~350 bps that preceded the 2015 energy credit crisis and the ~340 bps before COVID. The last time spreads were this tight or tighter, the worst financial crisis in 80 years followed. The market-wide price of credit risk is at one of its lowest readings in history, even as the actual incidence of credit stress is at one of its highest.
The HYG put-call open interest ratio has reached 5.16, an all-time extreme. One-week put skew is at 41.8%, the most bearish near-term reading on record. The options market on high-yield bonds is the most bearishly positioned it has ever been, while the bonds themselves are priced for perfection.
This is the disconnect within the disconnect. It is not just that equities and credit disagree. It is that the credit market disagrees with itself. CDS volumes say institutions expect trouble. Spread levels say the market hasn’t priced it yet. Both things cannot be true for long.
The Historical Record
There is a comforting belief among investors that historical analogies are inherently misleading and that “this time is different” is not just possible but likely. This belief is comforting because it allows you to ignore the historical record, which is not comforting at all.
The closest analog to the current configuration — equity markets at all-time highs while credit stress builds underneath — is 2007. Not the September 2008 that everyone remembers. The earlier part, the part nobody remembers because it didn’t feel like anything was happening.
The ABX index — a basket of subprime mortgage-backed securities — began its sustained decline in early 2007, accelerating sharply in February when HSBC reported $10.5 billion in unexpected subprime losses. By June, the index was implying significant losses on subprime mortgages. Two Bear Stearns hedge funds, both heavily exposed to subprime, blew up in June and July. The financial press covered it. The credit market noticed. And the equity market went up. The Dow Jones Industrial Average hit 14,000 for the first time on July 19, 2007 — one month after Bear Stearns’ hedge funds began collapsing. The S&P 500 didn’t peak until October 9, roughly eight months after the ABX started flashing red.
Ben Bernanke told Congress in March 2007 that “the impact on the broader economy and financial markets of the problems in the subprime market seems likely to be contained.” The people buying stocks in July 2007 were not fools. They had a reasonable thesis: subprime was a small part of the market, the economy was strong, corporate earnings were growing, and the Fed had the tools to manage any disruption. Every part of that thesis was true. The S&P 500 still fell 57%.
The 2014-2015 energy credit stress episode is instructive for a different reason. High-yield energy spreads began widening in mid-2014 as oil prices collapsed. For fourteen months, the S&P 500 largely shrugged. The equity market’s thesis was familiar: energy is a sector-specific problem, it won’t spread. Then in August 2015, the index dropped 12% in six trading days. Not because anything new happened in credit. Because the accumulation of stress finally broke through the surface tension of passive flows and momentum and dip-buying. The specific catalyst was a Chinese yuan devaluation, but the underlying vulnerability was the fourteen months of ignored credit deterioration.
In these two clearest precedents, credit stress preceded equity declines by 8 to 14 months. In other historical episodes, the relationship is murkier. The 1998 LTCM crisis revealed systemic fragility and leverage — but the equity decline that followed 18 months later was driven by the distinct dot-com valuation bubble, not by the credit stress LTCM had exposed. The Fed’s rate cuts in response to LTCM arguably extended the equity bubble rather than averting it. The 2019 repo market crisis exposed plumbing fragility — overnight repo rates spiked to 10%, the Fed intervened within hours — but the equity crash that arrived six months later was triggered by COVID, an exogenous pandemic that no credit indicator could have predicted. In 1973-74 and 1990, equities and credit declined simultaneously in response to oil shocks rather than one leading the other.
The honest reading of the historical record is this: when credit markets diverge sharply from equities, the divergence deserves serious attention as a leading indicator. In the cleanest cases, credit led equities by the better part of a year. In other cases, the credit signal was real but the equity catalyst came from somewhere unexpected. The 2022 episode — the counter-example, where credit stressed and equities corrected 25% without a systemic crisis — required conditions specifically absent in 2026: consumer balance sheets flush with stimulus savings, no external energy shock, and a Fed with credibility and a clear reaction function.
What the historical record does not support is a tidy “credit predicts equity with an 8-month lag” formula. The lags vary from months to over a year. The mechanisms differ. The catalysts are often unpredictable. What the record does support, consistently, is that credit-equity divergences of the current magnitude do not resolve in equities’ favor. The credit market may not tell you exactly when the equity correction arrives. It has a strong track record of telling you that one is coming.
The current divergence is approximately two months old. If the historical range holds, we are early.
The Trigger Map
Markets do not transition from complacency to panic because of gradual, evenly distributed revelations. They transition because of specific events that serve as coordination points — moments that give everyone permission to acknowledge what they already suspected. The macro environment provides the kindling. The triggers provide the match.
Here are the matches, in chronological order.
May 5-11: BDC NAV Reporting Cluster. The major publicly traded BDCs report net asset values on a quarterly lag. The May reporting window provides the first comprehensive look at Q1 marks across the private credit universe. FS KKR already trades at 45% of book value. KKR Real Estate Finance Trust is at 47%. These are prices that imply the market believes roughly half the stated asset value is fictitious. Fresh NAV marks will either confirm that suspicion — in which case, the redemption crisis accelerates — or reveal that reality is even worse. A third possibility, that the marks come in clean, would require either genuine improvement in underlying loan performance or aggressive mark-to-model accounting of the kind that, in 2007, delayed recognition by two quarters and doubled the eventual losses.
Approximately May 13: April CPI. The probability that April CPI exceeds 4% headline is, based on the mechanical passthrough of $125 oil and current food inflation trends, above 80%. A 4% print eliminates any residual expectation of a Fed rate cut in 2026. More than that, it forces the market to confront the possibility of a rate hike under a new chair who has already abandoned forward guidance. The significance is not just the number but the narrative it destroys. The equity rally has been underwritten, in part, by the assumption that the Fed is a backstop — that if things get bad enough, the chair will cut rates and support asset prices. A 4% CPI print turns the Fed from a potential ally into a potential adversary.
May 19-21: Consumer Earnings — Home Depot, Target, Walmart. These three companies are the most comprehensive real-time read on the American consumer. Home Depot tells you about housing-adjacent spending. Target tells you about the middle-income discretionary consumer. Walmart tells you about the broad consumer base, including the lower-income households most sensitive to fuel and food inflation.
The setup is precarious. The personal savings rate has fallen from 4.5% to 3.6% in two months — 90 basis points of cushion gone. Excess savings from the pandemic stimulus era are fully depleted. Gas prices are approaching $5 per gallon nationally. Consumer sentiment has been dismissed as a vibes indicator that doesn’t correspond to actual spending behavior, and that dismissal has been correct. But there is a gasoline price at which sentiment and spending converge — at which the consumer who has been saying “I feel terrible about the economy” starts acting on that feeling. If all three companies miss and guide down, the “strong consumer” narrative collapses in a single week.
June 16-17: Warsh’s First FOMC Meeting. Kevin Warsh takes the chair having signaled that forward guidance is over. No more dot plots. No more carefully calibrated language. The market will, for the first time in over a decade, go into an FOMC meeting without a clear expectation of what the Fed will say or do.
This is a structural increase in the distribution of outcomes. For the past decade, the Fed has effectively capped volatility by telegraphing its intentions far in advance. That dampened uncertainty, compressed option premiums, and facilitated the entire ecosystem of risk parity, vol-targeting, and short-vol strategies that currently supports equity prices. When that dampening mechanism is removed, option premiums should be structurally higher, vol-targeting strategies should carry structurally less equity, and short-vol strategies should demand more compensation. None of these adjustments have happened yet. When they do, the repricing will be abrupt.
The 30-Year Treasury at 5%. Bank of America chief investment strategist Michael Hartnett has described 5% on the 30-year yield as the “Maginot Line,” warning that “should 5% Maginot Line break badly, then the door to doom starts to open.” The 30-year is currently at 4.7%. At the current pace of inflation prints and Treasury supply, 5% is a plausible destination by late summer. Above 5%, CRE refinancing math breaks for a wide swath of maturing loans, mortgage rates breach 7.5%, and the equity risk premium — already at generational compression — enters territory from which there is no gentle exit.
Oil at $140 and the Behavioral Threshold. Consumer behavior around gasoline prices is not linear. It is stepwise. The historical evidence suggests that $5 per gallon is a threshold where behavior changes abruptly: consolidated trips, deferred travel, spending substitution that shows up in retail data within four to six weeks. At current refining margins, $140 Brent produces $5.10 to $5.30 at the pump. The GDP drag from $100 to $140 oil, based on standard estimates, is 1.2% to 2.0%. Applied to a baseline 2% GDP growth rate, that puts the economy near stall speed.
The most likely scenario is not a single trigger but a sequence. No individual event on this calendar is likely to crash the market on its own. The equity market has demonstrated its ability to absorb individual shocks — ceasefire collapses, hot inflation prints, single-company earnings misses. But it has not been tested on absorbing a rapid succession of them in a six-week window, with the equity risk premium already at generational lows, institutional credit protection at all-time highs, and every mechanical amplification strategy at maximum exposure.
The market can shrug off one of these. It is unlikely to shrug off all of them.
The Verdict
The market is mispricing risk. I want to be precise about what I mean by that and what I don’t.
I don’t mean that AI earnings are fake. They are the strongest reported numbers in a generation, and they are growing. I don’t mean that the economy is in recession. It isn’t, not yet. I don’t mean that the five companies driving this rally are overvalued in isolation — on a growth-adjusted basis, you can build a case for every one of them. What I mean is that the price of the S&P 500, taken as a whole, embeds a set of assumptions about credit conditions, consumer resilience, interest rates, and geopolitical stability that are collectively implausible. Not individually — each assumption is defensible on its own. Collectively. A market priced for perfection needs everything to go right, and the base rate for “everything goes right for six months during a shooting war with $125 oil” is not a number that inspires confidence.
The equity risk premium is at generational compression. Institutional credit positioning is at all-time extremes. Private credit redemptions are being gated. The CMBS distress rate is at a record. Office vacancy is at a record. High-yield spreads are as tight as they were before the 2007 crisis. The savings rate is declining at 45 basis points per month. And every mechanical force that amplified the rally — CTAs, dealer gamma, risk parity, vol-targeting, retail conditioning — is loaded to maximum long exposure with symmetric reversal risk.
The S&P 500 is not “the stock market” in any meaningful sense. It is a market-cap-weighted index in which five companies account for roughly 30% of the total value. When people say “the market is at all-time highs,” what they mean is that five technology companies with genuine AI revenue growth are at all-time highs, and their size is large enough to drag the index higher even if the other 495 companies are, in aggregate, struggling. And many of them are struggling. The equal-weight S&P 500 has underperformed the cap-weighted index by a wide margin. Small-caps are down on the year. Financials are under pressure. Consumer discretionary is facing the headwinds described above. The “average stock” is not at an all-time high. The average stock is having a mediocre year and hoping that the five giants at the top of the index continue to paper over the weakness below.
This is a tech fund with 465 hostages. The hostages are exposed to $125 oil, 3.2% inflation, tightening credit, and a consumer whose savings buffer is vanishing at an observable rate. The generals at the top of the index are marching forward. The army behind them is already in retreat.
I think the disconnect resolves via equity repricing, not via earnings growing into the valuation. I think the repricing is 15% to 25% from current levels, concentrated in the non-tech, credit-sensitive, consumer-exposed portions of the index. I think the timeline is months, not days or years — the trigger calendar runs from early May through September, with the densest cluster in the next six weeks. I think the Mag-5 hold up relatively better, because their earnings are real and their balance sheets are fortresses, but they do not escape entirely — a 20% decline in the equal-weight index drags the cap-weight index down 10-15% even if the top five names are flat. And I think the probability distribution is roughly: correction exceeding 10%, around 60%. Correction exceeding 20%, around 30%. Continued rally or flat, around 40%.
The single variable that determines whether this is a manageable correction or something worse is the growth trajectory of AI revenue at the Magnificent Five. The ERP is compressed because investors are implicitly betting that earnings growth will be fast enough to transform the current P/E into a future P/E that justifies the price. If Microsoft’s AI business continues at 123% and Google Cloud at 63%, the math self-corrects through the numerator — earnings grow into the valuation over two to three years. If AI revenue growth decelerates to 30-40% — still excellent by any normal standard but not enough to close the ERP gap — then the denominator has to adjust. Prices fall until the math works.
The honest assessment is that AI revenue growth will probably decelerate from current rates — 123% growth is not sustainable indefinitely at any scale — but probably remains high enough to partially validate the earnings case. That partial validation is why I expect a correction, not a crash. The Mag-5 earnings provide a floor under the cap-weighted index even as everything beneath them reprices. The passive autopilot bid provides a second floor, permeable but real. The result is something more like 2001-2002 — a grinding, rotational repricing over months — than 2008.
Now: the direction has a trade.
