The data arrives without a warning label. S&P 500 earnings closed the 2025 reporting cycle 14% above the long-term trend line. First time since 1955. That is not a forecast. That is a ledger entry. Equities shrugged. Crypto kept looking for the next catalyst. My job is to audit the gap between the number and the story.
I have been doing this since 2017. That year I audited the OmiseGO token sale line by line, found the exchange-rate logic favoring early whales, published a 15-page risk report, and watched the market blow up without me. The process is the same now. The number says something is stretched. The question is not whether the stretch corrects. The question is what breaks first.
Context: The Trend Is Not What You Think
What does 14% above trend actually mean? If the long-run trend for S&P 500 operating EPS is roughly $220-235, a 14% overshoot puts trailing earnings near $250-268. That matches public data: TTM operating EPS around $250-270 in 2025. So the statistical claim is plausible. But the word "trend" is doing heavy lifting. Linear regression? HP filter? Moving average? The source does not say. That matters. Different filters produce different overshoots.
What we can verify independently is the margin story. S&P 500 net margins are 12-13%. The 1990-2019 average was 7.5-9.5%. That is the real anomaly. Margins are the raw material of this earnings overshoot. Revenue growth has been moderate, but profit per dollar of sales sits at a historical extreme. This is not a debate about valuation. It is a debate about the durability of margins.
Four forces built this margin gap.
First: fiscal impulse. The U.S. federal deficit has run at 6-7% of GDP for consecutive years. The CBO projects roughly $1.9 trillion in red ink for 2025. That is a direct transfer into corporate top lines. Kalecki was right: budget deficits feed gross profits. Remove the fiscal pulse, and this earnings overshoot loses one of its main oxygen lines.
Second: the rate-lock effect. Companies refinanced at 2020-2021 interest rates when the Fed was pinned at zero. The Fed then hiked to 4.25-4.50%. But existing debt costs stayed low. That is not productivity. That is a time arbitrage. The lag between monetary tightening and corporate interest expense is longer than most models assume. The overshoot is partly a survivor dividend from refinancing windows that have now closed.
Third: asymmetric pricing. PPI is running around 1.5-2%. CPI is running around 2.5-3%. Input costs fell, output prices stayed sticky. That spread is an accounting transfer from consumers to shareholders. It shows up as margin expansion. It also shows up as political risk. Call it pricing power or greedflation; the mechanism is the same. The window is closing as wages and rents reprice higher.
Fourth: AI concentration. Information technology and communication services now account for roughly 40% of S&P 500 earnings. That is above the dot-com peak. AI-related capex is running above $300 billion per year across the major hyperscalers. Remove the Mag 7 and the companies around them, and the remaining 490 stocks are barely above trend. The overshoot is not broad-based. It is a leadership phenomenon.
Core: The Ledger That Matters for Crypto
Ledgers do not lie, only analysts do. And the analyst consensus right now is that record earnings mean risk assets are safe. I read the opposite. The S&P 500 earnings overshoot compresses the market's margin for error. If earnings mean-revert, equity estimates drop and multiples compress at the same time. That is the Davis double-kill. Crypto will not be immune.
The transmission chain is simple. Strong nominal earnings keep the Fed from cutting. If the Fed cannot cut, real rates stay higher for longer. That means stablecoin yields remain competitive with on-chain yield. It means the opportunity cost of holding risk assets stays elevated. It means the dollar stays bid. Every one of those factors acts as a headwind for speculative crypto flows.
But there is a second transmission channel that retail does not price. When equity markets start to question earnings quality, portfolio managers de-risk. They sell the smallest, most liquid positions first. Crypto is not sold because it is flawed. It is sold because it is the most efficient way to raise cash. Liquidity vanishes; principles remain. That ordering is consistent across every cycle I have traded.
I built this playbook in 2020. I put $50,000 into Harvest Finance to stress-test yield decay. The APRs looked incredible on day one. They decayed exactly as capital flooded in. The same math applies to macro earnings. A 14% overshoot is a yield above its sustainable equilibrium. It attracts capital until the marginal buyer is exhausted.
Here is the table I keep in front of me when I trade this setup.
Driver | Evidence | Lifecycle Stage Fiscal impulse | Deficit 6-7% of GDP | Mature, politically constrained Rate-lock effect | 2020-21 refinancing wave | Fading, debt rolls over Asymmetric pricing | PPI-CPI spread positive | Late stage, consumer resistance AI capex | Mag 7 capex growth >30% YoY | High momentum, unproven payoff
Each driver is at a different stage of exhaustion. Fiscal expansion cannot grow forever. The rate-lock benefit disappears as bonds mature. Asymmetric pricing attracts regulators. AI capex is the only pillar with forward momentum. But AI has not yet shown a clear productivity payoff in aggregate TFP data. That makes the entire overshoot a leveraged bet on one narrative.
I have lived this setup before. In May 2022, when Terra collapsed, I had a pre-planned liquidity protocol. I converted stablecoin positions into USD within minutes. I did not write a eulogy; I wrote a post-mortem. That is the difference between a trader and a tourist. Risk is not a rumor, it is a variable. You calculate it. You position for it. You move when the trigger hits.
Contrarian: Strong Earnings Are Not a Buy Signal
The retail read is straightforward: record profits, strong economy, buy the dip. The smart money read is different. The same number that proves strength also proves that the market has priced in perfection. CAPE is near 35. That is only a few points away from the 2000 peak. The forward P/E is 21-22. That leaves no margin for missed estimates, rising costs, or falling pricing power.
History does not give you comfort. Earnings ran above trend before the 1968 inflation breakout. They ran above trend before 2007. They ran above trend in 2017, just before the 2018 Q4 carnage. Sometimes the trend line catches up through slow price appreciation. Sometimes it catches up through a crash. The difference depends on whether the overshoot is structural or cyclical.
AI could be structural. But structural changes are rare, and they show up in hard data. Productivity growth above 2.3% is a start, but it is not conclusive. Until AI revenue appears across the board, I treat the 14% overshoot as a repricing risk, not a productivity dividend. Trust the contract, doubt the community. That applies to blockchains and macro narratives alike.
There is also a hidden contradiction. The inflation narrative says the Fed is winning. Core inflation is sticky near 3%, but official CPI is cooling. Corporate pricing power says the opposite. If companies truly have enough pricing power to generate 14% excess earnings, why is inflation falling? The answer: the aggregate is masking a K-shaped economy. Large firms have pricing power. Small firms do not. S&P 500 earnings are not the economy. They are the share of the economy captured by the top 500 firms.
That distinction is the single most important blind spot in the bull case. Investors are trading an index that claims record profits while the underlying real economy is cooling. The divergence cannot last. It will resolve through either a recession or a profit recession. Both outcomes are bearish for speculative crypto leverage.
Takeaway: Trade the Triggers, Not the Headline
I do not issue price targets. I issue trigger lines. The market owes you nothing. Precision kills emotion in trading.
Watch these levels:
- S&P 500 quarterly earnings growth. If it drops below +3% for two consecutive quarters, the mean-reversion trade is confirmed.
- Net profit margin. If the index margin falls below 11%, the margin compression story is live.
- Mag 7 capex. If combined capital expenditure growth turns negative, the AI pillar cracks.
- Core PCE. If it stays above 3%, the Fed will not cut. Rate relief is off the table.
- Oil above $95. That triggers a margin squeeze through input costs.
For crypto, the actionable levels are simple. If Bitcoin loses its 50-week moving average, the macro bid is gone. If Ethereum fails to reclaim its prior quarterly range, the altcoin cycle is over. I am not calling for a collapse. I am calling for an audit. Volatility is the tax on uncertainty. The earnings overshoot has increased the tax rate.
Position for the audit, not the story. The ledger always settles.