Broadcom, a US supplier of the chips and networking technology that underpin the global artificial intelligence boom, missed revenue expectations last week. Not by much.
On Monday, South Korea’s stock market had fallen more than 8%, Samsung Electronics had lost 10%, SK Hynix almost 8%, Taiwan’s market had been hit, Japanese semiconductor stocks were under pressure and roughly $1.8 trillion had been wiped from the value of the S&P 500.
Savvy investors should be paying close attention to that chain of events.
Not because it says something alarming about AI, but because it says something alarming about investors.
Broadcom’s results were disappointing. They were not disastrous. The company didn’t report a collapse in demand. It didn’t tell the market that AI spending was drying up, and it didn’t suggest data-center investment was slowing sharply.
Yet the reaction was brutal.
In my view, the sell-off exposed something that has been building for months beneath the surface of the AI rally: complacency.
The market has become so accustomed to positive surprises from AI-linked companies that even a relatively modest disappointment now triggers an outsized response across continents.
That’s rarely a healthy sign.
The dominant explanation for the sell-off is that tech stocks had become overextended and investors were taking profits. There’s some truth in that.
But profit-taking does not explain why a narrow revenue miss by a US company was capable of knocking more than 8% off South Korea’s stock market.
Complacency does.
For two years, investors have treated artificial intelligence as the closest thing markets have to a guaranteed growth story. The trade has been extraordinarily successful. Companies exposed to AI infrastructure have delivered exceptional earnings growth. Capital expenditure has surged. Demand has repeatedly exceeded forecasts.
The market gradually became conditioned to one outcome.
More spending, more growth and more upside surprises.
Every quarter reinforced the same belief. Investors who bought the story were rewarded. Investors who doubted it were punished.
Eventually, that creates a dangerous mindset.
The focus shifts away from risk and towards confirmation. Investors stop asking what could go wrong and start assuming the trend will continue uninterrupted.
This is what I believe happened here.
Broadcom exposed how little room for disappointment remains in the trade.
South Korea provides the clearest example.
Samsung Electronics and SK Hynix now account for roughly 40% of the KOSPI. Both companies have become major beneficiaries of the AI infrastructure boom. Investors around the world view them as direct ways to gain exposure to rising demand for advanced memory chips.
The same dynamic exists in Taiwan through TSMC. It exists in Japan through semiconductor equipment makers such as Tokyo Electron and Advantest.
As a result, a relatively modest earnings disappointment in the United States quickly became a reassessment of some of Asia’s largest and most influential companies.
This should concern investors for a simple reason.
Nothing material changed in South Korea, Taiwan or Japan.
What changed was sentiment.
A company headquartered thousands of miles away reported results that failed to satisfy a market accustomed to constant upside surprises, and investors immediately marked down some of Asia’s most important stocks.
As such, this isn’t a sign of a weak AI story, it’s a sign of a market that may have become too comfortable with a single narrative.
The irony is that the long-term case for artificial intelligence remains extremely strong. Businesses continue to invest heavily. Governments continue to invest heavily. Demand for computing power continues to grow.
Of course, I remain positive on the long-term outlook. But strong themes often create their own risks.
The stronger the narrative becomes, the more investors crowd into the same positions. The more investors crowd into the same positions, the more vulnerable markets become when expectations fail to match reality.
History is full of examples. The internet changed the world and investors still became too optimistic. China transformed the global economy and investors still became too optimistic. AI may prove even more significant than either, and guess what. Investors are capable of making the same mistake again.
It’s why I believe this week’s sell-off matters. Not because Broadcom missed estimates, nor that South Korea fell 8%, and not because AI suddenly faces a problem.
The significance lies in what the reaction revealed.
A market that has spent two years rewarding optimism is beginning to rediscover disappointment. And a market rediscovering disappointment can be a far more powerful force than a company missing expectations by a fraction.
A revenue miss in California should not be capable of wiping 8% off South Korea’s stock market.
The fact that it did should tell investors everything they need to know about where the real risk now lies.
Nigel Green is CEO and founder of the deVere Group.
