Hello everyone, welcome back to the U.S. stock analysis channel. I'm lorder. It's Tuesday, September 15, 2026, 8:25 PM Eastern Time. Today the four major indices closed lower again, with the Philadelphia Semiconductor Index opening higher but fading through the session. Of the S&P's eleven sectors, only energy and basic materials finished in the green; the other nine were down. The 10-year Treasury yield closed above 5%, its highest level since 2007. The market is trading cautiously overall ahead of tomorrow's Fed rate decision.

In today's episode, we'll first pick up the AI slowdown conversation from yesterday — it turns out quite a few people don't agree with that direction. Then we'll walk through what two investment banks' historical data show about where rates and the S&P tend to go after a rate hike, and where the S&P, QQQ, and the two semiconductor ETFs could end up in a worst-case scenario. Consider this a bit of psychological preparation in advance.

Let me give you the bottom line first: our stance right now is to stand aside. The market is trading cautiously ahead of the Fed's rate decision, the four major indices closed lower, and the 10-year Treasury yield closed above 5%, a level not seen since 2007. A 25-basis-point hike is already largely priced in, with the probability at 92.4%. If the hike comes as expected, the key focus should be on the rate path, inflation, and unemployment projections in the Summary of Economic Projections.

In the current rate environment, high-valuation growth stocks carry the most risk. That said, the S&P's medium- to long-term path still depends on earnings growth expectations. A short-term pullback of 10% is a normal correction, not a crash.

Let's pick up the AI discussion. Yesterday we talked about how several prominent figures in the AI model space have said they want to slow down AI development. But that view clearly doesn't sit well with everyone. Meta's Zuckerberg and NVIDIA's Jensen Huang see things differently. Zuckerberg believes AI development needs to accelerate, while Huang told President Trump that they won't let AI development slow down. Meanwhile, Trump views AI as a national security issue and believes the U.S. absolutely cannot ease up on AI investment.

Other countries are also making their positions clear. China disagrees. Australia's Prime Minister recently made a similar point, saying that if Australia can't develop its own frontier AI models, it will become a country that has to seek help from abroad, rather than a sovereign nation. ECB President Lagarde has also expressed a similar view, warning against relying on the American model and calling for Europe's own innovation.

It looks like the table for an AI slowdown hasn't even been set properly, and people are already dismantling the stage. So we'll just keep waiting and watching. As retail investors, we can pivot quickly, so we'll follow the data. If things really go south, we can always step aside.

Next, let's talk about what happens after tomorrow's rate hike — where the S&P and rates tend to go, and some things to keep in mind.

Let me lay out where the major instruments currently stand, their key levels, and where they could go in a worst-case scenario.

The S&P 500 is currently in the 7,500 to 7,600 range. Support sits at 6,800 to 6,900, which is the limit level. Resistance is at 8,539. The current stance is neutral, with the index near the middle of its reasonable trading range. A drop to 6,900 to 7,000 is still reasonable, with a maximum drawdown of roughly 10%.

QQQ is currently around 687, with limit support below 637. If it breaks below 687, our view turns bearish. QQQ's beta relative to the S&P is roughly 1.2 to 1.5, and this year it's been around 1.4. Based on that elasticity, if the S&P falls 10%, QQQ's maximum decline would be roughly 15%.

SOXX is currently at 500, with support at 486. Below that, the next level is 436 to 443, with the limit below 368. Resistance sits at 577 and 655. Our view is neutral-to-bearish. A break below 486 opens the door to 436 to 443, and in the worst case, 368, which is roughly a 26% decline from the current price. On the upside, a break above 577 points to 655.

SMH is currently at 542, with support at 534 and limit support below 421. Resistance is at 671. Our view is neutral. If the semiconductor space as a whole breaks down, SMH could head toward the low 420s, roughly a 22% decline. On the upside, it could reach 671, roughly a 24% gain.

Broadcom is currently at 339, with support at 324 and resistance at 365. Our view is bearish. It has already entered the 324 to 365 range and is drifting toward the lower boundary. A break below 324 would suggest institutions are pulling back, and the high-growth narrative would need to be questioned.

Finally, let me highlight several risk points to watch.

First, interest rate risk. The 10-year Treasury yield has already closed above 5%. If it rises another 100 basis points to 6%, that would weigh heavily on high-valuation growth and tech stocks. In a DCF valuation framework, the 10-year yield is a critical variable. A sharp move higher compresses the present value of future cash flows, which hits unprofitable or barely profitable growth stocks with forward P/E ratios of 80 to 150 times the hardest.

Second, Fed decision uncertainty. If the Fed doesn't hike tomorrow — a probability of only 7.6% — we could see a significant market rally. If it does hike, the key is the language around the future rate path.

Third, AI narrative risk. If discussions about slowing AI development continue to gain traction, it could trigger a repricing of the semiconductor sector and related stocks.

Fourth, S&P pullback risk. The S&P's maximum drawdown is roughly 10%, to the 6,900 to 7,000 range, which is a normal correction. But we need to watch for significantly lowered EPS estimates. Using a PE-based valuation range, with 2026 to 2027 EPS estimates and an 18 to 22 times multiple, the reasonable trading range is 6,986 to 8,539, with a midpoint of 7,762. We're currently near the middle of that range.

Fifth, individual stock breakdown risk. If Broadcom breaks below 324, QQQ below 687, or SOXX below 486, that could trigger further downside. On the historical side, Deutsche Bank data shows that in Fed tightening cycles, 10-year Treasury yields typically rise more than 100 basis points. Goldman Sachs' analysis of the past 7 tightening cycles shows the S&P averages a 2% decline in the first 3 months, followed by a 9% average gain over the next 12 months.

Alright, that's all for today. We'll wait for tomorrow's Fed rate decision and see whether the market makes some new moves that bring some excitement back to trading. That's it for now — see you tomorrow at the same time.