Why I Sold All My Stocks on August 12, 2025
This is not financial advice. I am not your advisor. I am documenting my thinking, the evidence I reviewed, and the conclusions I drew.
When I sold all my stocks on August 12, 2025, it was not because I believed I could predict the exact day the market would peak. The familiar refrain is that you cannot time the market, and in many ways that is correct. No one can consistently identify the precise bottom of a sell-off because bottoms are governed by human sentiment, panic, and policy responses that are impossible to forecast with certainty. What I do believe, however, is that market tops are a different matter. While the timing of the final push higher is always uncertain, the boundaries of what an economy can sustain are visible. When valuations are stretched beyond historical norms and the real economy begins to buckle under the strain, the probabilities shift. Tops are not accidents; they are moments when financial markets lose touch with reality, and reality eventually forces them back into line.
This is the lens through which I evaluated conditions in mid-2025. The U.S. economy was not simply flashing a few warning signals; it was exhibiting stress across every major pillar simultaneously. Labor, consumer credit, housing, commercial real estate, and equities all pointed toward unsustainable conditions. The convergence of these signals, each troubling on its own, created an environment that made remaining invested in equities less a strategy and more a gamble.
The labor market is the first place I looked, because it shapes both consumer spending and political response. On the surface, unemployment appeared manageable, reported at 4.25 percent in April 2025. But that number disguised more than it revealed. Adjusted for the decline in labor force participation, the effective unemployment rate would have been closer to 4.9 percent. That is not a trivial adjustment. Sahm’s Rule, one of the most reliable empirical recession indicators, triggers when the three-month moving average of unemployment rises by half a percentage point relative to its prior 12-month low. By that standard, the labor market had already begun signaling recession conditions. The headline numbers suggested stability, but the reality was hidden weakness. Making matters worse, the Bureau of Labor Statistics announced in August 2025 that it planned to cease publishing official jobs numbers in their current form. This threatened to remove one of the most important tools for assessing labor conditions in real time. Without transparent data, policymakers and investors risk flying blind, unable to detect early shifts that often precede recessions.
If labor was a crack in the foundation, consumer credit was a flashing siren. By mid-2025, credit card delinquencies had surged past 12 percent, auto loan delinquencies were nearing 5 percent, and student loan delinquencies had returned to double-digit levels after forbearance programs expired. This trifecta of stress is unlike what we have seen in previous cycles. Typically, delinquencies rise in one category at a time. The synchronized deterioration across all major forms of consumer credit suggested that households had exhausted their financial cushions. Once consumers are forced to prioritize which bills to pay, discretionary spending dries up, corporate revenues falter, and layoffs follow. The warning lights in consumer credit were not merely flickering—they were blazing red, indicating households were no longer resilient.
Housing told an equally troubling story. New-tenant rents suffered the sharpest collapse on record, a signal that demand was weakening even as supply remained elevated. At the same time, the Federal Reserve Bank of Atlanta reported that housing affordability had reached its lowest level ever, with the income required to purchase a median-priced home exceeding median household income by more than 50 percent. This is the worst affordability gap ever measured, and it left prospective buyers locked out of the market. Rising inventories of unsold homes and a growing number of price cuts confirmed the imbalance. Housing is not just another sector; it is the central transmission mechanism between credit conditions and consumer confidence. When affordability collapses, it reverberates through construction, lending, and household wealth. Falling rents may pull headline inflation lower, but that should not be mistaken for health as it reflects weakening demand and recessionary pressure, not relief.
Commercial real estate and construction were also flashing danger signs. Office vacancies pushed to record highs, commercial mortgage-backed securities showed rising delinquency rates, and construction spending turned negative on a year-over-year basis for multiple consecutive months. These conditions rarely reverse quickly. Instead, they worsen as refinancing walls approach, forcing property owners to roll debt at higher interest rates with lower collateral values. The echoes of the 2008 financial crisis were unmistakable, though this time amplified by the work-from-home shift that permanently reduced demand for office space.
While the real economy showed mounting stress, the equity market seemed to exist in another universe. Valuations were stretched further than at any point in modern history. The S&P 500’s price-to-sales ratio surpassed even the dot-com bubble, while the market’s leadership narrowed to a handful of mega-cap technology companies. Those names were buoyed almost entirely by the artificial intelligence narrative. Enormous sums were being invested in AI infrastructure on the assumption of multi-year exponential growth, yet evidence of widespread productivity gains or revenue lift was scarce. Instead, revenue in the sector was often circular, with companies selling compute resources to each other and booking it as growth. Never before had employment deterioration, surging consumer delinquencies, extreme asset valuations, and record housing unaffordability all flashed red at the same time.
Herd behavior added another layer of risk in that human beings rarely act alone in these environments.
Companies delay layoffs until their peers announce them.
Investors hold off on selling until price declines force their hands.
This collective hesitation creates parabolic rises that overshoot fundamentals, followed by overcorrections that erase years of gains in months. The COVID-19 crash in 2020, the financial crisis of 2008, and the dot-com collapse in 2000 all followed this script. Once the gap between valuations and fundamentals widens enough, the herd inevitably moves, and the correction is swift and severe.
History shows that major drawdowns rarely unfold overnight. From peak to trough, the process often takes around 18 months. Denial gives way to hesitation, then capitulation, and finally forced selling. Policy interventions can soften the slope but rarely change the destination once the starting point is extreme. That pattern was already forming in 2025. With valuations at record levels, credit markets breaking down, housing affordability collapsing, and the labor market deteriorating beneath the surface, the probability of a major correction over the next 18 months was high.
This is why August 2025 marked my exit point. It was not about calling the exact top. It was about recognizing that every major sector of the economy was flashing red at the same time. Labor force participation was falling even as unemployment looked stable. Credit card, auto loan, and student loan delinquencies were climbing together. Housing affordability was at the worst level ever recorded, while new-tenant rents collapsed. Commercial real estate was faltering under record vacancies and negative construction spending. And equities were priced as if none of these problems existed, with valuations above dot-com levels and concentrated in a single speculative narrative. It was not one indicator but instead all of them.
Critics often argue that selling too early risks missing the last 10 or 20 percent of upside. That may be true, but the final stretch of a bubble is the most dangerous. It is typically driven not by earnings or productivity but by speculation and herd momentum. If you miss that last climb, you forgo some profit. If you stay too long, you risk catastrophic losses when liquidity vanishes and prices reset violently. For me, the calculus was simple. Being early is inconvenient. Being late is devastating.
Looking forward, timing the bottom will be more difficult. Bottoms are sentiment-driven, and human psychology changes slowly. People do not abandon beliefs overnight. They deny, rationalize, and resist before capitulating. This is why historical drawdowns take more than a year to fully unfold. My expectation, based on prior crises in 1929, 2000, and 2008, is that the trough will emerge roughly 18 months after the peak and a major catalyst even which has not yet happened yet. I will not attempt to buy the exact bottom. Instead, I will re-enter gradually once panic sets in, valuations normalize, and fundamentals begin to align with prices again.
The lesson from August 12, 2025, is not that timing markets is easy. It is that tops are visible when valuation excesses converge with economic stress. In those moments, remaining invested is not prudent; it is reckless. For the first time in modern history, employment deterioration, surging delinquencies, collapsing affordability, and record-high valuations all appeared at once. This is the most obvious bubble we have ever seen to date.
