Market Update

Will Interest Rate Hikes Pop the AI Bubble? What 1999 and 2000 Can and Cannot Tell Us

By Corey HinklePublished September 30, 2026

The short answer

No one can say whether rate hikes will pop an AI bubble, and the video does not claim to. Corey Hinkle compares today with 1999 and 2000, when four Federal Reserve rate hikes, rising inventories, and customer debt preceded the Nasdaq peak, and names Nvidia's share price against its 200-day moving average as one early signal to watch.

Full video transcript

I got a very interesting question from a client yesterday, which was simply put as, are these rate hikes going to pop the AI bubble? And it's like, okay, there's a lot to unwrap there. So let's kind of talk through that one piece at a time. I've got some information to show you about that question today. And I'll give you a market update as well, because we got some interesting data that does actually have an effect on the expected rate hikes coming over the next 12 months. Before I begin, I must remind you, this is a financial education presentation. You must do your own due diligence before acting on anything you hear in this presentation. More disclaimer information can be found on Ankerstar Wealth.com. The opinions Express are mine alone. We do have the S&P 500 up about 1/4% today. with those yields kind of retreating a little bit, not a lot, but a little bit, and you've got oil back on the rise a bit. Here's why those yields are increasing or decreasing today. Because we do have the Fed's preferred inflation gauge with a much lighter than expected number. If you come over here, That brought it down to 3%. Now, what is 3% inflation? That's obviously too hot, but I think they were expecting 3.4% or so. And so kind of viewed as some good news. We also got some news that lots of private sector jobs were added in September, much more than August. And that was that was unexpected to some extent. And so you do see that employment market very strong. If we go over, I wanted to look at the FedWatch tool. It's like, okay, inflation data, employment data, what is this change on the expected rate hikes that we're going to be getting not only later this fall, but into next year? So they have Yesterday, we had about a 50-50 chance of a rate hike in October. Today, after this new data got digested by the market, you're looking more at a 63% chance of just a hold. So the Federal Reserve holding rates, you got to December, Here, actually, let's go out one year. If you go out to next September, we still have about a 50% chance, which didn't change much, according to this new data, about a 50% chance of four or more rate hikes over the next year. to five rate hikes, about a 50% chance. And then you've got, you know, a 50% chance that you may only get two to three more, but a 50% chance of four to five more is still in play for next September. So long term, this data doesn't change much. Why? Because inflation is still a problem. even if it's at 3% and you know employment clearly stronger than expected. So here we go, let's get to the question. Someone asked, are these rate hikes going to pop the AI bubble? That's what they said. So I told them, well, in the year 1999, and I had known this, you know, read the books and stuff, but in the year 1999, and the fact I knew was you got four rate hikes, between the May of 1999 and the NASDAQ peak, which was in March of 2000. So that would have been here. The NASDAQ peaked in March of 2000. The S&P 500 didn't peak until like September, I think. So you got 4 hikes between the whatever, the start of the hiking and the peak of the NASDAQ. Well, we've gotten one hike. that's still filtering through here, because we don't have the September data in this chart. But we certainly haven't gotten 4 yet, and it looks like we might. We might get 4. So that would be kind of history rhyming. Now, is there an AI bubble? I mean, that's a totally separate question. We've actually made a video on that. what that would look like and where it would show up. I'm not going to really speak about an AI bubble today, but they did have four rate hikes. And then the famous thing that happened to really pop the bubble was this Cisco. A lot of Cisco's customers, see, one of the differences is a lot of NVIDIA's customers are the hyperscalers and they're very rich companies. They've started taking on debt to do some of this data center stuff, in some cases pretty significant debt, but they are the biggest, strongest companies in the world. Back in 2000, in 1999, 1998, quite a few of Cisco's customers, some of them were big, like in AT&T and all that, but some of them were a little more speculative. So what started to happen to Cisco was They were enjoying the good times, obviously this is peak bubble stuff, but their inventory started to pile up and... You could see this coming out. So this was, yo, we're back in the year 2000 here. April 29th, 2000, they released this earnings report. It is also, can I say, it is really hard to find old either news articles or press releases in Google these days. It's really tough for me to dig back and find this, but I got this one from Cisco. You had inventories up 30, this is about 34%, 35% year over year in the 2000, April 2000 quarter. Remember, I think Cisco peaked, I don't remember if Cisco peaked in March of 2000. But it was in the peaking process here. Inventories were climbing rapidly. Here, let's go out to, this is actually the next one. This is in August of 2000. Here, we saw inventories up about 100% year over year. So inventories were piling up quickly. A lot of Cisco's customers were taking on debt to build some of this stuff out, the internet, and they were hiking rates. So, you know, do we have a trifecta like that going on today? I don't, well, let's look in a second. But one other thing I wanted to show you, this is kind of, this stuff fascinates me. You know, if you've watched the show, you know, this is my thing. I love this stuff. But I wanted to reread this statement from the CEO and just see if we could replace it with AI. So here we go. Globally, business and government leaders are beginning to dramatically transform their traditional business models into AI business models. That seems to fit, okay. These new AI-based models reduce costs, generate revenue in new ways, empower employees and citizens, and provide the agility needed for the AI economy's rapid pace. Customers are increasingly seeking Nvidia's expertise to help them through their transformation. So I just think that stuff's interesting. It's like there are some overlaps, right? If you replaced internet, the word internet with AI, it seems like, you know, some of these statements are almost verbatim. Working OK. Now, do we have debt problems today? Yeah, the poster child for that is going to be your Oracle. Oracle taking on significant debt to build out data centers for OpenAI. And Oracle probably not in a bubble here. I mean, you're down 52% year over year. If there's a bubble, it would be more on the semiconductor side of things, which which hasn't wavered yet. Now, we saw Cisco had inventory problems. I wanted to look at Nvidia's recent quarter. Is Nvidia showing the same thing? It looks like Nvidia's inventories in the recent quarter were up about 50% year over year. So are we going to get 4 rate hikes? I mean, maybe. Looks like it. That's what the market expects. Do we have inventory problems building at an Nvidia? Maybe. 50% year-over-year inventory climb. That might mean that things aren't flying off the shelves as fast as they were. Do we have debt problems from the customers like Cisco had? I mean, maybe Oracle, Oracle certainly doesn't have clean hands, but some of the hyperscalers are totally fine. So, you know, to answer that client's question, are rate hikes going to pop the AI bubble? It's an impossible question to answer, all right? But remember what we kind of keyed on in this show. And that is just going to be the health of Nvidia's stock price. Nvidia's stock price is going to be the tell, one of the first tells. If you start to get a break below the 200 day moving average, you start to get a softness, you know, from the buy side there.

Key Takeaways

• A client asked whether rate hikes will pop the AI bubble. Corey Hinkle's direct answer is that it is an impossible question to answer, but history offers a framework for what to watch.

• New data showed the Fed's preferred inflation gauge at about 3%, below what Corey recalls being expected, and private-sector hiring stronger than expected. In the FedWatch tool, the odds of an October hike moved from about 50-50 to roughly a 63% chance of a hold.

• Longer term, the tool still showed about a 50% chance of four to five more hikes by next September, because inflation remains a problem.

• In 1999 and 2000, four rate hikes, rising inventories at Cisco, and customer debt all preceded the Nasdaq peak. Today the market expects perhaps four hikes, while inventory and customer-debt pictures look mixed.

• Corey's stated tell is the health of Nvidia's stock price, particularly a break below its 200-day moving average.

What Did the New Inflation and Jobs Data Show?

The S&P 500 was up about a quarter of one percent, with Treasury yields retreating slightly and oil rising a little. Corey ties the move in yields to the Fed's preferred inflation gauge, which came in lighter than expected at about 3%. He notes that 3% is still too hot, but that he recalls expectations nearer 3.4%, so markets viewed it as good news. Corey covers the inflation reading at (1:19).

A second data point pointed the other way: private employers added many more jobs in September than in August, which Corey describes as unexpected to some extent. A strong employment market is one input the market weighs alongside inflation when it estimates where the Federal Reserve may go next. For more on how the labor market and inflation have been feeding each other this month, see Rate Hike Odds Rising: What Inflation Data Shows.

How Did Expected Rate Hikes Change?

Using the FedWatch tool, Corey shows that the odds of an October rate hike fell from about 50-50 the day before to roughly a 63% chance that the Fed simply holds. He walks through the tool's readings at (1:40). Markets reprice quickly, and these probabilities can change again with the next inflation or jobs report.

The longer-term picture barely moved. Looking out to next September, the tool still showed about a 50% chance of four to five more hikes and about a 50% chance of only two to three. Corey's explanation is that inflation is still a problem, even at 3% and with employment stronger than expected. Our 12-month market outlook discusses how the firm is thinking about policy uncertainty and inflation risk.

What Happened With Rate Hikes Before the 2000 Nasdaq Peak?

According to the history Corey recounts, the Federal Reserve raised rates four times between May 1999 and the Nasdaq peak in March 2000. The S&P 500 peaked later, around September of that year in his recollection. Corey discusses the 1999 comparison at (3:15).

So far in the current cycle, the Fed has raised rates once, and Corey notes the chart he uses does not yet include September data. The market currently expects that there could be four hikes in total, which he describes as history rhyming. A similar count of hikes does not mean a similar outcome. Rate increases are one factor among many, and each cycle unfolds under different conditions.

What Did Cisco's Inventory Show in 2000?

Corey points to Cisco as the famous example of how a bubble can start to unravel. In the video he shows Cisco's April 2000 earnings report, where inventories were up roughly 34% to 35% year over year, and a later report from August 2000, where inventories were up about 100% year over year. Corey walks through the Cisco reports at (4:21).

He adds that some of Cisco's customers were more speculative and were taking on debt to build out the internet while rates were rising. Corey asks whether all three conditions, rate hikes, rising inventory, and customer debt, are present at the same time today. His answer is that some appear to be, while others appear less clear.

Does 2000-Era Language Sound Like Today's AI Story?

Corey rereads a statement from the CEO in Cisco's 2000 earnings release about business and government leaders transforming traditional business models, and swaps the word internet for AI. He observes that the result reads almost verbatim like statements made about AI today. He covers the comparison at (6:39).

This is an observation about language, not a forecast. Similar wording does not show that outcomes will match, and Corey is explicit that whether an AI bubble exists is a separate question from the one he was asked.

How Do Inventories and Debt Compare With Cisco's Situation?

On debt, Corey says Oracle is the poster child for borrowing to build data centers for OpenAI, though he does not view Oracle as being in a bubble, noting its share price is down 52% year over year. He suggests that if a bubble exists, it would more likely show up on the semiconductor side, which had not wavered at the time of recording. Corey discusses Oracle at (7:42).

On inventory, Nvidia's most recent quarter showed inventories up about 50% year over year. Corey says that could mean products are not flying off the shelves as quickly as before, but adds a qualifier: it is a maybe, not a conclusion. He also distinguishes Nvidia's customers, many of them very large hyperscalers that have begun taking on debt, from the more speculative customer base Cisco had. This recap is not a view on any particular security.

What Is the Tell to Watch?

Corey's answer to the client is that whether rate hikes will pop the AI bubble is an impossible question to answer. What he says the show has keyed on is the health of Nvidia's stock price as one of the first tells. A break below the 200-day moving average, or softness from buyers, is the type of signal he would watch. He states this near the end of the video at (8:45).

For long-term investors, the practical point is that no single indicator settles the question. Market signals can be noisy, and a diversified plan that fits your time horizon and risk capacity is designed to work across many outcomes rather than depend on predicting one. Learn how the firm approaches this through its investment planning process, or review how rate and market risk may affect your retirement planning. To talk about your own situation, contact Ankerstar Wealth.

This recap is financial education, not investment advice. Investing involves risk, including the potential loss of principal, and past market cycles do not predict future results. Do your own due diligence before acting on anything you hear or read here.

This article is general information, not personalized investment, tax, or legal advice. Your situation is specific to you — talk to a qualified professional before acting on anything here.

Frequently asked questions

Will interest rate hikes pop the AI bubble?

No one can say for certain. Corey Hinkle calls it an impossible question to answer and separates it from whether an AI bubble exists at all. He suggests watching the health of Nvidia's stock price, particularly a break below its 200-day moving average, as one of the first tells.

How many rate hikes came before the 2000 Nasdaq peak?

According to Corey Hinkle, the Federal Reserve raised rates four times between May 1999 and the Nasdaq peak in March 2000. So far in the current cycle there has been one hike, and the market expects there could be four in total.

What are the odds of a Fed rate hike in October 2026?

As shown in the FedWatch tool in the video, the odds of an October hike moved from about 50-50 to roughly a 63% chance of a hold after lighter inflation data and stronger private-sector hiring. Probabilities like these change frequently with new economic reports.

How many more rate hikes does the market expect by September 2027?

The FedWatch tool in the video showed about a 50% chance of four to five more hikes and about a 50% chance of two to three more by next September. Corey says the new data changed this long-term picture very little because inflation is still a problem.

How does Cisco in 2000 compare with Nvidia today?

Corey Hinkle notes that Cisco's inventories rose about 34% to 35% year over year in April 2000 and about 100% by August 2000, while Nvidia's most recent quarter showed inventories up about 50%. He also observes that Nvidia's largest customers are very large hyperscalers, whereas some of Cisco's customers were more speculative. The comparison is illustrative and not a prediction.

Keep reading