Congress Keeps Buying the Dip Nobody Else Wants to Explain

# Congress Keeps Buying the Dip Nobody Else Wants to Explain

Here's a fun paradox for you: Broadcom posted one of the most jaw-dropping quarters in semiconductor history back on June 3, and the stock got taken to the woodshed for it. Revenue up 48% year-over-year to $22.2 billion. AI chip revenue up 143% to $10.8 billion — a record. Guidance for the next quarter at $29.4 billion, blowing past what Wall Street was modeling. And the market's response? A 12-15% haircut in the days that followed. If you're new to AVGO, welcome to the strangest tradition in tech stocks: sell the blowout, buy the panic.

And apparently, at least 13 members of Congress agree with the "buy the panic" half of that trade.

The Politicians Like What They See

The disclosed trading data on AVGO reads like a heat map of conviction. Across both chambers, one tracker shows 34 buys against just 11 sells over a recent window; another puts the House alone at 12 purchases versus 18 sales but with buy dollars running nearly double sell dollars (~$3.4M vs ~$1.9M). Zoom out to the full historical record since 2014 and it's 145 buys to 95 sells — a persistent, if modest, buy-lean. Jared Moskowitz has been in and out six times. Ro Khanna's spouse reportedly picked up AVGO options back in March. The most recent disclosed trade landed July 23.

Now, the honest caveat, because this column doesn't do cheerleading: the STOCK Act gives members up to 45 days to disclose, so a "July purchase" could easily have been executed in early June — right in the window of that post-earnings dip. Which, if true, means Congress didn't predict anything clairvoyant. They just did what any disciplined investor should do: treated a 48%-growth, guidance-raising quarter that got punished by 13% as a gift, not a warning.

Why the Market Sold a Great Quarter

The bears aren't just being contrarian for sport. Q3 gross margins are guided down to 74% as the revenue mix shifts more toward lower-margin AI silicon — even as operating margin holds a healthy ~67%. There's also the small matter of the ~$69 billion VMware acquisition sitting on the balance sheet, and debt levels that analysts keep a close eye on for dividend coverage. And then there's the single question that decides whether this stock is a $630 name or a $300 name, depending who you ask: does custom silicon keep taking share from Nvidia's merchant GPUs, or does this cycle top out?

That's not a small spread. A bull case of $630 against a bear case of $300 is basically the market shrugging and saying "we genuinely don't know how big this gets." When your price targets have a 2x range, you're not looking at a mispricing — you're looking at a stock the market hasn't finished arguing about.

The Case for Believing the Bulls

Here's what tips it, for me: management isn't just riding a good quarter, they're guiding to ~$56 billion in full-year AI semiconductor revenue (+180% from 2025) and holding firm on a $100 billion AI revenue target for fiscal 2027. There's a reported $73 billion backlog behind that. Broadcom is the custom-silicon shop for Google, Meta, and OpenAI — the "if you're not buying Nvidia GPUs, you're probably calling Broadcom" company. The Apple chip news that popped the stock into the high-$360s/low-$380s in early July, plus the August 6 VMware vDefend and Avi Load Balancer rollout, are reminders this isn't a one-trick AI pony — the software side is still compounding at +31% guided growth.

What I'd Actually Do Here

The stock has been chopped from the $495 area down toward the high-$380s, sitting well off its highs but nowhere near its 52-week low of $281.61. That's not a broken chart — that's a stock digesting a monster move while the market decides if 74% gross margins on hyperscaler AI chips is a feature or a bug. Q3 earnings, expected in the first days of September, is the next real catalyst, and given the guide of $29.4 billion, expectations are already sky-high — which means the "blowout quarter, stock drops anyway" script could run again.

If it does, take the hint from the folks on Capitol Hill: dips into the $360s on a name with this backlog and this growth trajectory have historically been for buying, not fleeing. Just don't confuse conviction with certainty — this is still a stock where the bull and bear cases disagree by a factor of two, and no amount of congressional buying changes that math.