When the Market Turns Down: Read the Evidence Chain First, Then Set Your Stop-Loss

1. Rallies Run on Capital — and So Do Selloffs

A rising market is, at its core, driven by capital flows, and behind those flows sits the main chain of logic: an industry that keeps getting better. Once money has poured in for long enough and prices have already climbed sharply, the market becomes fragile. The slightest gust of news can trigger a big move.

The reason is that the market's original supply-demand balance gets broken. On the way up, demand is strong and everyone is rushing to buy. But once prices are high, holdings are full of accumulated profits, and a single disappointment can flip the market from strong demand to strong supply — far more people want to sell than want to buy.

2. This Time's Trigger: Google's and Intel's Q3 Reports

The immediate trigger for this selloff was the Q3 earnings of two heavyweights in U.S. tech.

Alphabet (Google): Q3 revenue came in at $127.58 billion, up 25% year over year, with Google Cloud revenue growing 103% — a striking growth rate. But the profit side clearly missed: earnings per share were $2.12 versus analyst estimates of $2.99 (partly due to an $8.7 billion quarterly loss on its equity investment portfolio). More importantly, look at cash flow: free cash flow was negative $9.1 billion, the second consecutive quarter in the red, and analysts broadly expect it to stay negative for several more quarters. The company also issued $28.7 billion in senior unsecured debt during the quarter. After the report, the stock fell as much as roughly 11% in after-hours trading.

Intel: This one is a bit different. According to public reports, Intel's data center business grew strongly and earnings per share beat expectations, but the company's forward outlook — especially for its PC business — clearly disappointed the market, and the stock dropped sharply, with reported declines ranging from roughly 20% to 30% depending on the source. In other words, what hit the market wasn't just the quarter's numbers, but the guidance for what comes next.

Together, the two reports cooled investors' expectations for the AI industry's future, and a wave of holders started selling. The selling pressure was heavy enough to drive prices sharply lower.

3. There Are Always Two Voices in the Market

At moments like this, the market usually splits into two very different reactions:

• Some investors sell first and ask questions later, worried the trend has changed;
• Others remain convinced the AI industry is still growing strongly, and see this as short-term capital being overly sensitive to sentiment, producing a temporary wave of selling rather than a real fundamental problem.

Who is right can't be decided by emotion right now — it has to come back to evidence.

4. What Actually Changes the Narrative: The Evidence Chain in Leaders' Earnings

In my view, what truly changes an industry's narrative logic ultimately has to show up in the earnings of the industry's leading companies. If the following signals appear together, they form a short-term inflection point for the whole industry:

1. Capital expenditure keeps shrinking — suggesting giants are losing confidence in future demand;
2. Leverage quietly rises — suggesting expansion is increasingly funded by outside financing rather than internally generated cash;
3. Short-term operating cash flow drops sharply, or even turns negative — suggesting the core business's ability to generate cash is weakening.

Any one of these on its own doesn't necessarily mean the trend has reversed. But when several stack up together, that's a fairly complete evidence chain.

5. Checking Against Reality: Where Are We on That Chain?

Hold Google's results up against this evidence chain and you find an interesting picture:

• On debt, Google issued $28.7 billion in bonds this quarter — debt is rising, and that one is already visible;
• On cash flow, free cash flow has been negative for two straight quarters — that one is already visible too;
• But on capital expenditure, Google's earlier guidance put full-year capex at up to $205 billion — there is no sign of capex shrinking. The current tension is precisely "spending too much, recouping too slowly," not "afraid to spend."

So my read is: the evidence chain is only partly in place, and it's too early to call this a turning point for the entire industry. Whether the Q3 shortfall is temporary, or whether the giants' capex and operating health are all actually starting to decline, needs to be confirmed by tracking the other major players' earnings. If other cloud providers keep raising capex and AI hardware orders keep climbing, this is more likely an emotional overreaction. If multiple giants cut capex at the same time, that's when to be on guard for a genuine change in trend.

6. If a Downtrend Really Does Form: The Playbook

Suppose the trend has indeed turned. Within my own trading framework, two things need to be done:

First, set a stop-loss. This is the most important discipline in a downtrend. Decide in advance: "at what price am I proven wrong?" — and then execute it strictly, rather than finding new reasons to keep holding as the price slides. A stop-loss isn't admitting defeat; it's about controlling risk and preserving capital, so you're still around when the next opportunity shows up.

Second, wait patiently for panic selling. The tail end of a downtrend usually comes with a round of emotional, panicked selling — everyone rushing for the exits regardless of fundamentals. Once that wave has been released and most of the money that wanted out has already left — when selling pressure is essentially exhausted — that's when a better entry point may be forming. In other words, don't try to catch falling knives on the way down; wait for the signal that "everyone who was going to sell has sold."

Closing Note

This post records my thinking on how to approach a market entering a correction. It reflects a personal view and is not investment advice. Earnings data and market reactions change quickly, and any actual decisions should be grounded in your own ongoing research and independent judgment.