The Warning Lights Are Flashing Again

Markets do not ring a bell at the top.

They do something far more annoying.

They give you a series of warnings, one at a time, from places most investors are too busy ignoring to notice. Valuations get stretched. Credit spreads get too tight. Retail traders discover leverage again. Everyone decides that buying the dip is no longer a strategy but a constitutional right. Earnings start looking better than the underlying business reality. Wall Street produces a fresh stack of PowerPoint decks explaining why none of this matters because artificial intelligence, productivity, innovation, and whatever phrase the Internet Experts are yelling this week.

That is usually when I start counting my fingers and checking my wallet.

The current market has several of those warnings flashing at the same time. None of them says the market must crash tomorrow morning at 9:37. Markets can stay expensive, silly, and overconfident longer than we can stay entertained by people explaining why old rules no longer apply.

The warning is not about next week.

It is about future returns.

The setup today suggests investors should lower expectations for broad market returns, demand a larger margin of safety, avoid paying heroic prices for heroic stories, and remember that credit markets usually tell the truth before stock investors are emotionally prepared to hear it.

Let us start with earnings.

Owen Lamont at Acadian Asset Management recently published a terrific piece called “Waiter, there’s a P in my E.” The title is clever. The message is not cute at all. Lamont points out that the S&P 500 already looks expensive, with the market trading around 27 times trailing earnings. That is bad enough. The bigger problem is that the “E” in that P/E ratio may be getting flattered by accounting treatment rather than operating reality.

Lamont focuses on the effect of accounting rule ASU 2016-01. Under GAAP, companies must include certain unrealized gains and losses from equity holdings in reported earnings. That means rising prices in public and private markets can show up in net income, even when the company did not actually sell the investment and collect cash.

That matters right now because reported S&P 500 earnings growth looked spectacular in the first quarter of 2026. According to Acadian, S&P 500 earnings growth was said to be running around 28% year over year. Lamont cites work showing that unrealized gains at just Alphabet, Amazon, and Nvidia added $69 billion to reported S&P 500 earnings in the quarter. Remove those unrealized gains, and earnings growth falls from about 28% to roughly 16%.

Sixteen percent earnings growth is still good.

Twenty-eight percent earnings growth is the kind of number that makes people say things like “new era” while looking for a camera.

The Amazon example is especially important. Amazon owns a stake in Anthropic. When Anthropic completed a February 2026 funding round at a much higher valuation, Amazon’s stake increased in value. That helped Amazon report $16.8 billion of pre-tax non-operating gains tied to its Anthropic investment. Amazon’s first-quarter net income was $30.3 billion. In other words, a very large chunk of reported earnings came from a mark on a private company investment, not from selling more cloud computing, shipping more packages, or squeezing another nickel out of the retail business.

That is not fraud. That is not a conspiracy. That is accounting.

The problem is that markets do not always treat accounting as accounting. They treat it as proof.

Stock prices rise. Private company marks rise. Reported earnings rise. Higher reported earnings make the market look less expensive. The market then decides higher prices are justified. Higher prices create more gains. Everyone claps. CNBC gets another segment. Main Street media discovers a new bull market theme. Internet Experts post charts with rocket emojis.

That is not fundamental analysis.

That is a mirror pretending to be a window.

Lamont’s warning is simple. If rising prices are helping create rising earnings, then investors are using price action to justify price action. That is a feedback loop. Feedback loops can be powerful on the way up. They can also be remarkably unpleasant on the way down.

Citadel Securities’ first-half 2026 market structure review adds a second warning. The market is not merely expensive. It is being driven by flows, concentration, retail activity, passive buying, and leverage in a way that looks increasingly mechanical.

Citadel notes that the ten largest companies now account for nearly 40% of the S&P 500. That is near record concentration. Semiconductors now represent nearly one-fifth of the S&P 500, the highest share on record, and their index representation has quadrupled since June 2020.

That is not diversification.

That is a handful of giant technology and semiconductor stocks wearing an index costume.

The passive flow data is just as eye-opening. Citadel says ETFs have already taken in $1.2 trillion of net inflows year to date, 45% ahead of last year’s record pace. In just six months, investors allocated about 2.5 times what historically represented a full year of ETF inflows.

Passive flows are not evil. Indexing is not evil. ETFs are not evil.

They are also not valuation sensitive.

Money goes in. The index buys. The biggest names get the most dollars. The biggest names go up. Their weights rise. The next dollar into the index buys even more of them. Wall Street then explains that this is all rational because the stocks went up. Thank you, professor. Very helpful.

Retail behavior is even more interesting. Citadel says May and June set new retail activity records, with average daily retail cash equity volume running 65% above 2025 levels and more than double the 2024 average. Nine of the 10 most active trading days ever observed on its platform occurred in the prior two months, including seven in June.

Retail has become what Citadel calls a structural bid. June was tracking as the strongest month in its history, with daily purchases running nearly four times last year’s average. June 12 was the largest single day of retail net buying ever observed on Citadel’s platform, beating the prior record by 50%.

The buy-the-dip crowd is not just buying dips. They are buying everything.

Citadel reports that retail investors bought nearly 3.5 times the average daily amount on S&P 500 down days during the first half of 2026, the strongest buy-the-dip behavior in its dataset. Even on rally days, they still bought nearly 1.5 times the daily average.

That is a great description of confidence.

It may also be a great description of complacency.

The options market is where this gets more troubling. Retail traded a record $6.8 billion of options premium per day in June on Citadel’s platform, more than double the historical average. Retail is also chasing the same leadership that dominates the index. In June alone, retail traded about $1.9 billion of semiconductor options premium per day, 6 times the historical average, with about 75% of that activity in calls.

Nothing says “thoughtful long-term capital allocation” quite like everyone buying short-dated call options on the same stocks after they have already become the market.

The leverage picture is no better. Citadel says one out of every three listed options traded in the United States now expires the same day. Nearly half of retail options volume executed by Citadel Securities is now in 0DTE contracts, up from 30% in 2025 and just 13% in 2021. Leveraged ETF assets reached a record $218 billion, more than 4.5 times the level from June 2020. Since the end of March, leveraged ETF assets rose roughly $82 billion, or 60%, led by technology and semiconductor exposures.

We have seen this movie before.

The costumes change. The actors change. The explanation changes. The plot is always the same.

Investors make money. They decide they are smart. They add leverage. They confuse liquidity with permanence. They mistake a crowd for confirmation. They assume the exit door is much wider than it actually is.

Citadel’s volatility data also suggests this is not a calm market. It is a market with strange internals. Three-month implied correlations have fallen to their lowest level in more than 15 years, creating what Citadel calls one of the strongest stock picker’s markets in history. Semiconductor implied volatility has more than doubled over the past decade, rising from 32% in 2016 to nearly 72% today. Nearly 70% of Nasdaq rallies during May were accompanied by higher implied volatility, the highest monthly frequency since 2005.

That last point matters. In a normal market, stocks go up and volatility goes down. In this market, stocks go up and investors pay more for upside exposure. That is not calm confidence. That is a chase.

Then there is credit.

Credit spreads are one of my favorite market indicators because bond investors are paid to worry. Stock investors can tell stories. Bond investors have to ask whether they are getting paid enough to take the risk of not getting their money back.

Right now, high-yield spreads are tight. The ICE BofA U.S. High Yield Index Option-Adjusted Spread was 2.75% on June 30, 2026, according to FRED.

That is not panic. That is not even mild concern. That is credit markets walking around in flip-flops with a margarita.

Tight spreads do not mean disaster is imminent. They do mean investors are not being paid much for credit risk. When high-yield spreads are this tight, future returns from credit tend to be less attractive because the coupon has less room to absorb mistakes, defaults, downgrades, or a shift in investor mood. Credit can stay tight for a long time, but tight spreads are rarely where the great bargains live.

The better opportunities come when spreads are wide, people are scared, and Wall Street is suddenly very concerned about risk after encouraging everyone to ignore it for the previous three years.

That is why I prefer a credit-first framework. Credit tells us when the market is paying us to take risk. Right now, broad credit is not paying much. Equity valuations are not offering much either.

The S&P 500 Shiller CAPE ratio was 41.02 for June 2026, up from 40.16 the prior month and 36.11 a year earlier. YCharts notes that CAPE is used as a valuation metric for long-term returns, with higher CAPE ratios generally associated with lower future returns over the next couple of decades as valuations revert toward the mean.

That is the point.

This is not about predicting whether the S&P 500 goes up or down next week. I have no idea. Neither does anyone else, although many people with excellent haircuts will pretend otherwise.

This is about the likely return from buying the broad market at high valuations, with tight credit spreads, concentrated leadership, record ETF inflows, record retail activity, rising leverage, and earnings that may be flattered by unrealized gains.

That is not my favorite setup.

It does not mean investors should sell everything and move to a bunker stocked with canned beans, ammunition, and gold coins advertised during late-night television.

It means investors should be selective. It means price matters again. It means balance sheets matter. It means cash flow matters. It means insider buying matters. It means tangible book value matters. It means credit spreads matter. It means we should not let Wall Street convince us that valuation discipline is an old-fashioned hobby practiced by people who still balance checkbooks.

My approach does not change.

I want assets selling below a reasonable estimate of value. I want strong balance sheets or at least balance sheets where I am being paid for the risk. I want cash flow. I want dividends that are covered. I want management teams buying stock with their own money. I want community banks below tangible book value with excess capital, low credit problems, and M&A optionality. I want REITs where the assets are real, the debt maturity schedule is manageable, and the market has already priced in too much misery. I want BDCs and credit vehicles where the portfolio is seasoned, marks are credible, and the yield compensates me for the risk. I want small and mid-cap stocks where the business is improving and the valuation still leaves room for being wrong.

I do not want to pay 40 times cyclically adjusted earnings for the privilege of owning the same crowded mega-cap index everyone else owns, while pretending that a mark-to-market gain on a private AI investment is the same thing as recurring operating earnings.

The warning signs are not screaming that the world is ending.

They are telling us that the easy money has probably been made in the broad market. They are telling us that future returns are likely to be lower from today’s starting point. They are telling us that the next several years may reward security selection, patience, and valuation discipline far more than blind exposure to the indexes.

That is fine with me.

A market with low correlations and high dispersion is a stock picker’s market. Citadel’s own data says three-month implied correlations are at the lowest level in more than 15 years.

Good.

Let the index buyers chase the same 10 names. Let the 0DTE crowd turn the market into a casino with better graphics. Let Wall Street sell the story that valuations do not matter because this time the earnings are different.

We have seen this before.

The answer is not panic. The answer is discipline.

Buy assets with a margin of safety. Favor cash flows over stories. Favor credit signals over television commentary. Favor tangible value over adjusted fantasy metrics. Keep some dry powder. Be willing to look boring while everyone else is confusing activity with intelligence.

The market is giving us warnings.

We should listen.

And in the meantime, happy Fourth of July!