$AVGO initially dropped about 6% after the earnings report.

Then Hock Tan started speaking, and the stock recovered a significant portion of its losses.

I think that's exactly the story of Broadcom's earnings tonight.

The numbers were good. But the call said much more important things.

First, the earnings report:

▪️ Revenue: $29.6B (+86% YoY)
▪️ Non-GAAP EPS: $3.32 (+96%)
▪️ Free cash flow: $13.7B
▪️ FCF margin: 46%

Broadcom is literally printing cash. There's not much to discuss on the operational side here. The balance sheet is very strong.

The main event is again AI:

▪️ Q3 AI semiconductor: $16.7B
▪️ +221% YoY
▪️ +54% QoQ

Q4 expectation is: $21.7B

So, a business that has grown this much is expected to grow another 30% in the next quarter. I think this is one of the most important figures of the night.

So why did the stock initially fall by 6%?

Q4 total revenue guidance is $34.8B.

Is that bad? No.

But in a stock like AVGO, where expectations are at their peak, just a beat isn't enough anymore.

The market wanted to see a big beat, a big raise.

It didn't come.

I attribute most of the initial sell-off to this.

Then Hock Tan opened up the long-term AI mathematics in the call.

Broadcom's discussed path for AI semiconductor revenue is roughly:

2026: $56B
2027: $115B
2028: $230B

Wait a minute... If this happens, the AI ​​business almost doubles for two consecutive years.

This is where the market is listening again during the call.

Another issue is margin. The non-GAAP gross margin has fallen to approximately 75%.

But I don't see this as the main problem today.

Because management had already stated that the consolidated margin would decrease as the AI ​​semiconductor mix grows.

So this isn't another bad surprise tonight.

AI is growing faster, the revenue mix is ​​changing.

The price of this is blended margin.

I think the real risk to watch is something else:

Custom AI silicon competition.

Of Broadcom's $20.8B semiconductor revenue, $16.7B is now AI.

So, AI is driving semiconductor growth to a very large extent.

That's why Google starting to work with Marvell isn't insignificant news.

The question is no longer:

"Will AI grow?"

"How much of this growth will Broadcom get?"

One more detail:

Broadcom had a $0 buyback this quarter.

But at the same time:

▪️ Paid off $5.6B in debt
▪️ Cash outflow $19.6B → $24B
▪️ Generated $13.7B in FCF

I don't interpret this negatively.

The company is reducing its post-VMware debt and strengthening its balance sheet before the massive capital needs required by AI growth.

My post-call summary:

✅ AI demand isn't slowing down, it's accelerating
✅ Q4 AI: $21.7B
✅ FCF is exceptionally strong
✅ Software business continues to grow
✅ Long-term AI outlook is much stronger than post-earnings fear

But:

⚠️ Expectations are set too high
⚠️ Margin mix will put downward pressure
⚠️ Google/Marvell and custom silicon competition must be monitored

Initial reaction after earnings:

“Guide isn't good enough.”

Reaction after call:

“Okay, but the AI ​​story might be much bigger than we thought.”

For me, there was no break in my $AVGO thesis tonight.

On the contrary, the long-term math on the AI ​​side has become even more interesting.

This is not investment advice.