Zhao Jian: Every year there are ghost stories about a US stock market crash. Will this time again end with more fright than harm?

I. Ghost Stories Every Year: A Crash Chronicle of the Past Five Years After spending enough time in financial markets, one discovers a pattern: every year, the market tells a terrifying ghost story. The story changes annually, but the ending is remarkably similar. Looking back at the past five years, it has been almost one hurdle per year: 2022 was the worst year aggressive rate hikes plus the Russia-Ukraine war. The outbreak of the Russia-Ukraine war and the violent dollar rate hikes arrived simultaneously a geopolitical black swan overlapping with a broad tightening of liquidity, a textbook macro version of a "Davis double kill." The Nasdaq fell more than 30% from its peak, global stocks and bonds were crushed together, and countless balance sheets were permanently reshaped that year. The hole dug that year was the deepest in five years. In 2023, the ghost story shifted to Silicon Valley Bank. A bank with US$200 billion in assets collapsed within three days, Signature Bank followed close behind, and Credit Suisse was hastily sold off at a bargain price. The market briefly believed that "Lehman Moment 2.0" had arrived, and autopsy reports on regional banks flew everywhere. The result? The Fed rolled out the BTFP tool overnight, and once the liquidity valve was opened, the panic dissipated within three months. In 2024, it was the yen's turn to hike. The Bank of Japan ended the negative interest rate era, carry trades were unwound on a massive scale, and on that "Black Monday" of August 5, the Nikkei plunged 12% in a single day, with global markets bleeding alongside. The speed at which the fear index spiked at the time was comparable in history only to March 2020 and October 2008. Yet three weeks later, the market behaved as if nothing had happened. In 2025, the ghost story came in two parts. In the first half, it was Trump's tariff war "reciprocal tariffs" burst onto the scene, and doomsday rhetoric about global supply chains ran rampant; in the second half, it was the exposure of financing fraud within AI tricks involving computing power leasing and circular revenue were uncovered, and the market began to question the financial authenticity of the entire AI narrative. Both episodes made the Nasdaq shudder for a while, and neither ended the bull market. This year, 2026, the story is again a two-part serial. In the first half, it was the US-Iran war decapitation strikes, retaliation, the shadow over the Strait of Hormuz, and oil prices soaring; in the second half, it was the midterm election upheaval Democrats took control of both chambers and began to settle accounts with data centers, and faith in computing power capex suffered a blow at its foundation. US stocks were once again pushed underwater. II. Why Has Every Time Ended with More Fright Than Harm? Every ghost story has dug a big hole in the market. Some holes were deep and long-lasting, like 2022, which took nearly two years to fill; some were shallow and short-lived, like August 2024, which was climbed out of in three weeks. But without exception US stocks ultimately climbed out, continued to surge, marched onward, repeatedly hit new highs, and extended the bull market. This is not luck. There are two hard logics behind it. The first logic is the new global investment cycle and innovation cycle brought by the AI revolution. This is not an ordinary inventory cycle, nor an ordinary equipment renewal cycle, but the superposition of the Juglar cycle and the Kondratieff cycle a decade-level capex wave nested atop a once-in-fifty-to-sixty-years technological revolution long wave. In human history the steam engine, electricity, the internet every diffusion of a general-purpose technology has set off epic investment and growth, and AI is the culmination of them all. Productivity is rising, consumption is expanding, and corporate earnings are genuinely climbing steadily this is the hardest foundation of the bull market. As long as this underlying narrative has not died, every crash is merely a deep squat at halftime of the bull market. The second logic is the Fed's massive liquidity injection since the pandemic, which has raised the water level of global asset pricing across the board. Enormous liquidity sits idle in the financial system with nowhere better to go. Every panic, which appears to be capital fleeing risk, is in fact merely capital briefly changing posture within risk assets. If the water is deep enough, the boat cannot sink. China also joined this bull market in 2024. With the "924" policy package as the starting gun, A-shares and Hong Kong stocks embarked on one of their longest-lasting bull markets Chinese assets are no longer merely spectators of the global bull market, but formal participants. III. This Time: It's America's Own Turn Since the start of this year, the plot has shown a new and worrisome feature: major asset classes are crashing one after another. First gold war premiums and inflation expectations tore at each other, and the gold price rode a harrowing roller coaster; then semiconductors the aftermath of the AI financing fraud plus the shadow of data center account-settling sent the Philadelphia Semiconductor Index tumbling from its pedestal; then South Korea the most sensitive barometer of the global semiconductor cycle the KOSPI plunged; then China A-shares and Hong Kong stocks underwent a deep correction. Now, finally, it is America's own turn midterm election upheaval, policy account-settling, and the heart of this five-year bull market, US stocks, has begun digging its own hole. The transmission sequence is interesting. Over the past few years, every time it was America that had trouble and others that suffered; this time, it was only after global assets crashed one by one that panic flowed back to American soil. What does this show? It shows that what the market is pricing is no longer just a single event, but deep skepticism about the entire AI capex cycle and America's political account-settling is precisely the official confirmation of that skepticism. IV. Will This Time Again End with More Fright Than Harm? So the core question is: will the market, as in the past five years, again come through with more fright than harm and resume its bull run? My answer is: it depends on the market no blanket conclusion. The probability of the US resuming its bull run is high. Two reasons. First, the AI technological revolution itself is real; what has been killed is only valuation bubbles and disorderly capex, not the grand trend of technology diffusion data center construction will slow and become more standardized, but it will not stop, because real growth in computing power demand is there for all to see. Second, the Democratic account-settling is "for high-quality development," not to bury the market ordinary Americans trade stocks, Democratic elites themselves trade stocks, and the two parties share a fundamental common interest in maintaining capital market stability. Historically, every hole dug by "left-wing correction" has, in hindsight, turned out to be a golden pit; the exception at the scale of 1929 presupposed that the technology cycle had already exhausted itself, and today is clearly not that. But for other markets, one probably cannot be so optimistic. Markets like South Korea and A-shares will most likely gradually bottom out and grind upward after all, the broader environment is improving with global liquidity easing, the war ending, and tariffs receding. But note that what they experienced was a fierce bear; after a fierce bear, they can typically only enter a weak bull state. Breaking above previous highs will come under considerable pressure. The reason is not hard to understand. The US bull market has "global pricing power over the technological revolution" as its anchor, so when it falls there are buyers; while the valuations of other markets are largely parasitic on the chain of US AI capex South Korea's memory, China's computing hardware export chain, are essentially "supplying America's data centers." When the chain master begins to compress spending and settle bubbles, the earnings expectations of chain members must be systematically revised down. Valuation holes can be filled by liquidity; earnings holes can only be filled by time. V. Conclusion: Living with Ghost Stories Returning once more to the question in the title: will this time again end with more fright than harm? For the US, most likely yes the long cycle of the technological revolution is not over, the great tide of liquidity has not receded, and after the panic clears, the halftime whistle of the bull market will sound again. For other markets, it is "fright, yes, but the harm will not fully dissipate" they will bottom out and warm up, but to jump straight from a fierce bear back to a mad bull is unrealistic. Five years now, the market has told ghost stories every year, people have liquidated and exited every year, and then every year they have slapped their thighs in regret. What this should teach us is not "mindless bottom-fishing," but to distinguish what is a ghost story at a cyclical bottom from what is a death knell at a cyclical top. There is only one criterion for telling them apart: has the underlying technological revolution and earnings cycle died? As long as the answer is no then every ghost story is an opportunity dressed in terrifying clothing. But if one day the protagonist of the ghost story is no longer liquidity, no longer politics, no longer war, but the technological revolution itself running out of steam that is when the real bear market takes the stage. This article is sourced from "Zhao Jian's Humanistic Meditations," author: Zhao Jian; GMTEight editor: Liu Jiayin.