Four days before Baidu reported, I told readers I’d sold a cash-secured put on the stock. The reasoning was in my August note: the valuation already priced in a lot of bad news, the AI pivot was real, and my strike sat below the 52-week low. Then Q2 landed. The stock fell as much as 13% intraday. The price alert on my phone went off enough times that afternoon that I half-assumed the tracker was broken. It wasn’t.
So let’s walk through what was actually in the report, why the market reacted the way it did, and, since I have money on the line here, what I’m doing with the position now.
What was in the numbers
Baidu reported on August 18. Revenue came in at RMB 31.3 billion, about $4.62 billion, down 4% year over year and short of the roughly $4.65–4.74 billion analysts expected. Earnings were $1.06 per ADS against expectations in the $1.35–1.46 range, a miss of more than 25% and wider than the revenue shortfall. On its own, a 4% revenue miss doesn’t explain a 13% drop. The segment detail does.
Online Marketing Services, the search-ad business that has been Baidu’s cash engine for two decades, fell 19% year over year to RMB 13.1 billion. That’s not a one-quarter wobble. It’s now several consecutive quarters of decline, and the cause is structural: Chinese users are moving from a search box to AI chatbots and short video, and a weak property market and cautious consumers have companies trimming ad budgets on top of that. Management said on the call to expect the pressure to continue through the second half. That is about as direct as a company gets about a problem it can’t fix quickly.
General Business revenue, the wider operating segment, was down 4% at RMB 25.2 billion. iQIYI, the streaming arm, slipped 5% to RMB 6.3 billion but managed a small sequential gain.
The AI side is strong, which is what makes this hard
The easy narrative right now is “Baidu is dying, sell it.” I don’t think the data supports that, and the reason is the other half of the report.
The AI-powered businesses grew 25% year over year to RMB 12.5 billion. Within that, AI Cloud rose about 50% to RMB 7.3 billion, and the piece that rents out raw compute for training and inference, GPU Cloud, was up 283% off a small base. So the report contains one line collapsing at 19% on a large base and another line compounding at triple digits on a small one. The question the market is trying to price is which of those curves matters more, and how soon.
There was one operational detail I flagged in my own notes. Apollo Go, the robotaxi unit, had ride volume “affected by operational adjustments in some domestic cities due to regulatory considerations.” Translated: local governments made them change how they operate in certain markets. Autonomous ride-hailing is Baidu’s most exciting long-term option, and it carries regulatory risk on top of the usual execution risk.
Why it fell this hard, not just that it fell
Two things turned a bad quarter into a 13% day. First, Fitch cut Baidu’s credit rating from A to A-, citing structural decline in search advertising. Rating agencies move slowly and deliberately; that downgrade is them saying the earnings base has permanently reset, not dipped.
Second, the sell-side reaction split in a way worth reading closely. Barclays cut its target to $124. Bank of America trimmed to $165 but kept a Buy rating. When analysts are lowering numbers and a chunk of them still think the stock is cheap after the cut, it tells you the disagreement isn’t about whether the ad business is deteriorating. It’s about whether the price already more than accounts for it. One fair-value estimate I saw put the stock around $132, roughly 27% above where it was trading after the drop; the average target sat closer to $166.
Where the stock sits now
The session was volatile from the open: down about 7% pre-market, off as much as 13% intraday, settling in the low $90s. Zoom out and Baidu is down roughly 35% in 2026, a brutal year for a stock that started it full of optimism about its AI story.
Even after the miss, the forward multiples are low: forward P/E around 14 and EV/EBITDA near 3.4. In a healthy grower those numbers would look like a bargain. The market is either pricing genuine structural risk or overcorrecting, and right now reasonable people are looking at the same figures and landing on opposite conclusions.
How I re-checked the position
I’ll give you the framework I actually used sitting with the report that night, rather than jump to the verdict. The process is the part you can reuse.
- Separate the headline miss from the segment story. A revenue and EPS miss tells you almost nothing by itself. You have to open the segment breakdown and ask which specific lines moved, and whether that’s information your thesis already contained or something new.
- Check whether management’s forward guidance changed. Baidu explicitly told investors ad pressure will persist into the second half. That’s a specific negative statement, not boilerplate caution. When a company volunteers that, take it at face value.
- Compare growth rates, not dollar amounts. GPU Cloud up 283% on a small base against ad revenue down 19% on a large one is a real race with a clock on it, not a vague feeling about “the AI story.”
- Go back to your entry logic, not your current mood. I sold the put at an $80 strike, below the 52-week low, for a breakeven near $74. Even on the worst tick of the day the stock was in the low $90s. My thesis was never “Baidu’s ad business thrives forever.” It was “the balance sheet and the AI optionality make this worth owning 25–30% below mid-August prices.” That thesis is being tested. It hasn’t been broken.
If you want to see how the payoff math works before selling a put yourself, run the numbers through the options profit calculator first.
What I actually think about the put right now
It wasn’t a fun 24 hours. Watching a stock you have strike exposure to drop 13% produces a specific kind of stomach-drop that no amount of “I calculated my breakeven in advance” fully cancels. I checked the position more times that afternoon than I’d admit in a column where I’ve told readers not to do exactly that with concentrated trades.
But this is where the discipline earns its keep. My strike is $80. The stock never got within ten dollars of it, even at the low. My delta going into earnings was in the 0.15–0.20 range; it’s ticked up with the drop but the cushion between the market price and where I’d be assigned is still wide. The report sharpened exactly the risk I said I was taking when I opened the trade, China-specific structural risk rather than a clean US-equity setup. It didn’t introduce a risk I hadn’t accounted for. I’m not surprised there was bad news. I’m working through what kind of bad news it is.
What I’m doing from here
Nothing, for now. The stock is well clear of both my $80 strike and my $74 breakeven, and the drop came from real, disappointing segment data rather than a liquidity scare. Buying the put back at a loss to make the anxiety stop, or doubling down to “average into conviction,” are both the kind of emotional move I’ve spent years warning readers about. I’d rather not become the cautionary tale in my own archive.
Over the next two quarters I’m watching two things, not the daily quote. One: whether GPU Cloud’s growth holds anywhere near triple digits or decelerates hard. That’s the number that decides whether the AI segment can backfill the ad decline before my June 2027 expiration. Two: whether management’s “pressure persists” language turns into an actual floor in the ad business. A floor changes the whole risk picture on this position. A second leg down changes it the other way.
I still think Baidu is worth owning at an effective $74 cost basis: strong balance sheet, AI Cloud compounding fast against a shrinking ad base. That view got tested on August 18. It didn’t get knocked down. It just got more complicated, more uncomfortable, and, honestly, more informative. Which is roughly the best you can hope for from an earnings report on a position you sized to be able to sit through.