Why DeepSeek Triggered $1 Trillion Market Crash (Uncut) 01-28-2025

Why DeepSeek Triggered $1 Trillion Market Crash Today | Keith McCullough

The tech sector is leading a broad market sell-off today on Monday as the S&P 500 is down about 2% and the Nasdaq is down almost 3.5% intraday. China’s DeepSeek AI company is taking Silicon Valley by storm and it has a lot of American companies spooked. So DeepSeek is a Chinese AI company which just a week ago launched its latest AI model called R1.

The model is supposedly on par if not better than some of OpenAI’s latest iterations at problem solving and this has investors worried because importantly not only have they done this with a fraction of the cost but they also don’t need the most expensive GPU chips. As a response, NVIDIA is down about 16% on the day leading the charge downwards. Is this the end of American tech dominance? We’ll find out with our next guest, Keith McCullough, CEO of Hedgeye.

We’ll also be talking about his outlook for Q1, the Trump trade and monetary policy. What will the Fed do this week in response? Welcome to the show. Welcome back, Keith.

Good to see you. Thank you. I much appreciate it.

Looking forward to the conversation. A lot of volatility today. NVIDIA dropped 16% as China’s DeepSeek AI model sparks a global tech sell-off.

DeepSeek is a new open source AI model that is topping in number one of the App Store with a fraction of the budget as OpenAI and some saying just as powerful if not more. What can you tell us about this tech sell-off, Keith? Is this a micro-related event or is there something else bigger going on here? I think it’s a very big deal when it comes to NVIDIA because it’s got the gun pointed right at the head of the main exposure that many people own when it comes to tech, large gap growth, the factor exposure that is momentum. So if NVIDIA didn’t break and actually weaken throughout the day, I wouldn’t consider it and break being defined as my intermediate curve trend signal level of $134.

It broke that and it was like a hot knife through butter, as we like to say, and it’s all the way down to $118 now. And there is fundamental reason. My analyst, Felix Wong, our global tech analyst, immediately and essentially downgraded NVIDIA for the first time in his career, saying, look, this is a game changer.

And if you really want to get into the details on that, I’m not going to give you the specific deep-seeker details on what’s the difference between that and chat GPT, but you nailed one of them, which is far lower cost. And the other one is an efficient search on a narrower and one would say a more kind of expert field. So we’ve said like instead of just looking at the entire hospitals, chat GPT, you’d be asking specific specialists on what it is that you’re asking about within the medical profession.

Well, let’s just see what some people are reporting on this and whether or not it’s impactful to the markets long-term. So DeepSeek launched a free open-source language, a large language model in late December, claiming it was developed in just two months at a cost of under $6 million. We talked about that.

The company released a reasoning model that is reportedly outperforming open AI’s latest and many third-party tests. Mark Andresen, tech VC called DeepSeek’s product one of the most amazing and impressive breakthroughs I’ve ever seen and a profound gift to the world. So the tech race between China and the U.S. is heating up.

What does this mean for the tech sector overall, you think? I think generally speaking, there’s probably a lot of opportunity. I mean, today, this is again, this is talking about super expensive GPUs from NVIDIA becoming much lesser of a cost in terms of people having to have those chips, needing to have those chips. And we’re having an ability to point just to create this brand new thing that everybody’s excited about.

And the other thing is that the next new exciting thing can be ostensibly created quick and at a cost-effective weight as well. So technology is a broad sweeping term. Growth is a broad sweeping term.

I think that you get a lot of other growth-related opportunities on sale today. But I’m not buying NVIDIA, in fact, on any balance. I’d short it.

Why short NVIDIA? I mean, it’s the most expensive stock in human history. Let’s just start with that. But I mean, that’s not the real reason.

I mean, that’s just going to augment the view, which is you have a rate of change slowdown, as we like to call it, pod one or revenue growth to begin with at NVIDIA. Before this happening today, revenues were going to decelerate throughout the year at NVIDIA. Therefore, we would say that it should show multiple compression off its most expensive multiple, you know, like again, like any mega cap has ever seen.

So it’s an expensive stock with growth slowing now as a huge competitive threat staring right at them that the market’s taking quite seriously. And that’s good enough for me. Let’s just take a look at the heat map today.

So as you pointed out, NVIDIA down, Microsoft, a bunch of chip stocks are down, AMD, Qualcomm. So if you look on the other side, Amazon down a bit too as well, Tesla down, Google down, Apple is up and a bunch of com stocks are down. Other sectors are relatively unchanged.

So maybe this is just a chips and semiconductors sell-off today. It’s not a broad market sell-off, if you can call it that. How are you interpreting today’s move? Well, a part of today’s sell-off, don’t forget, David, came from last week’s all-time closing highs for SPY.

So, I mean, it didn’t come out of nowhere, as I like to say, like when you get a market that’s that complacent after a rally, there’s some fragility, right? And there’s some susceptibility to a correction. So you’re seeing to your point, you’re seeing that very specific and focus on on semiconductors and NVIDIA specifically. But you’re seeing a broadening, if anything, of strength in places where the market has not had strength on a relative basis, the most obvious being health care.

So you can see some of the growth year or prior growth names that people got smoked in, like Eli Lilly or Novo, for example. Those stocks are having a pretty good day. You’re seeing consumer staples, which has been a terrible place to have your money, is having a big bounce.

So there’s a big rotation underneath. The market’s basically saying, I’m going to take this market cap away from NVIDIA and I’m going to give it to this part of the market. And that’s in some ways healthy.

Now, if the VIX were to break out into the mid-20s, I wouldn’t call that healthy. I’d call that more of a concern. But for now, it’s actually in the order of magnitude.

If you’re telling me NVIDIA is going to be down 15, 20 percent on the day, I wouldn’t have guessed that the market wasn’t down a lot more. Overall, are we going to start seeing over the next year or perhaps two years a rotation of of capital from U.S. tech stocks into perhaps Chinese tech stocks with DeepSea just being the onset of this rotation? Anything’s possible. China’s an unmitigated disaster right now.

Ironically, or maybe not ironically, David, you had the Chinese printed their worst PMI numbers last night, both on the services side into a contraction there and on the mainline manufacturing PMI with sub 50 to 49. So if China’s in what we call quad four, they have more demographic issues than Time Magazine. There’s nothing that’s really resuscitated their growth from a stimulus perspective, despite every single attempt.

So the last thing I’m buying today is going to be something out of China. Again, I’d be looking for U.S. growth opportunities and or related sector exposures that would do well in what we call quad two, which would include the financials, for example, or industrials on sale. If you get it.

The problem is, like I said, you’re not getting a broad based sell off in every single stock. All right. Well, how much of the sell off today is just anticipation for what the Fed may do this week? Well, there’s that too.

I mean, I was joking this morning that, you know, Jensen Wong is giving the Fed a rate cut all of a sudden because, you know, bond yields went right to what I call the low end of the risk range. And anytime bond yields go to the low end of the range, in particular on the short end of the curve, the two year, which is really my front runner for Fed policy, you have to ask yourself whether or not Powell’s going to be more dovish or the data is going to be more dovish and make Powell more dovish. Now, the problem of course is that the data has been hawkish and that’s why we’ve eliminated basically every single one of these rate cuts but one.

I still think that it’s like a token one at that. And then the other thing to think about with the Fed meeting is that right after that, you’re actually going to get a Fed. We won’t get a Fed cut this week.

I don’t think you will. But if NVIDIA loses another 15% tomorrow, maybe they’ll have an emergency rate cut. I’m joking, but really the S&P 500, the quite sometime.

But right after that, you get the ECB rate cut on Thursday. So lots to think about this week that have nothing to do with NVIDIA. What happens if the Fed does not cut this week as actually the markets are anticipating? Are we going to get a sell off or is that already priced in? I think it’s priced in.

If you go back to when inflation bottomed, we’ve had an inflation cycle bottom back in September of last year when headline CPI hit 2.44%. Since then, CPI’s got up a full 50 basis points. The Federal Reserve completely missed that. Most consensus pundits who are looking for 3, 5, 6, 8 interest rate cuts completely missed that.

So it’s been really since October that we’ve been making the argument that number one, the US dollar figured that out. It went up 10% over that period. Number two, the bond market figured it out because it melted down and has been rate short.

And number three, it’s because of what we thought it was. It’s an economy that’s actually stronger than people think. And actually in January, our data was suggested it got even stronger.

How many rate cuts are you expecting over the course of 2025? I know that’s a tall order. A lot can change in the next couple of months. But what’s a reasonable base case assumption here? I just saved one.

Okay. All right. 10-year yield, are we getting back to 5%, you think? I think you took a good run at it.

Now, of course, the 5% threshold is a round number, which I don’t like to always use. It’s actually 4.99 was the top of the cycle for the 10-year yield. So if you go back and you can see that, that was back in 2023.

Of course, we’re coming out of an inflation environment back then where inflation got up towards 9%. So now we’re talking about inflation accelerating from 2.44 on our numbers closer to 3.5 by the time we get into the third quarter. We still have the street high on inflation in terms of what we call our CPI now cast.

So yeah, I think that combined with U.S. growth accelerating. Don’t forget that today’s news on, while it’s bad for Nvidia, it’s actually very good for human beings. I mean, you get a cheaper, more readily and accurate bot, if you will, to discover things and learn from things.

I mean, our own company is going to do better with that and all the learnings from that than having to use chat GPT or whatever else my team’s using. So the growth component plus the inflation component gives me, yeah, I got the 10-year buying yield can go back up towards 4.9%. I don’t, at this juncture, have it breaking up beyond that. Let’s take a look at some news coming out of the Bank of Japan.

They raised their key interest rate to about 0.5% on Friday. This is the highest interest rate in 17 years. Are we going to see a reverse carry trade again, meaning flows out of the U.S. into Japan? I think that was a one and done.

What happens, and to your point, what happened technically here was that it’s called interest rate differentials, right? If the Fed’s not going to be able to raise rates on Wednesday because the data’s too strong and the BOJ themselves, who’ve never raised rates, or at least in my career, on a relative basis, they’ve never been more hawkish than the U.S., that’s your trade. Your trade is rear view looking. I think to superimpose that into the future, you’d have to have an outlook for Japanese economic acceleration further, which we do not have.

We actually had Japan starting to slow already. And you’d have to have the U.S. slowing in kind, which we don’t have that either. We have, like I said, we have the U.S. economy, what we call Quad 2. So that would be on the margin from here, as opposed to looking backwards, which was different into the BOJ rate act, that would be on the margin hawkish relative to yen policy.

And stepping away from tech stocks and NVIDIA for just a minute, on the macro front, the bigger picture, are you cautious or optimistic about equities overall this year? What’s your positioning or sentiment? We’ve been very bullish on growth as a factor, exposure and momentum. So S&P Moe, SPMO, for example, or different growth categories that not everybody thought through post the Trump trades or post the inauguration, like space. So, for example, the ETF of all things is called UFO that I’m long.

And it might sound like, you know, crazy or whatever, but I don’t care about that. I mean, we’re long quantum computing. We’re long growth where we can find it.

And that can be consumer discretionary as well with XLY, for example, which obviously includes Amazon. Any growth story, we’re willing to buy it. There’s actually been a broadening of the rally into small mid-cap stocks, mid-caps we like a little bit better right now.

And at that level, I think we’re very specific. Like we have, I think, 58 names right now on what I call my signal strength buy list where my analyst likes it and so does my signal. A lot of those stocks are what we call quad two stocks.

Like I said, financials have plenty of growth stocks right now that are working quite fine for us as well. Can you just explain what quad two is? From my understanding, it’s more of a risk on play. Am I correct? Yep.

So when you have, so we have our four quadrant model. Quad one, two, three and four. Quad two is when you have the rate of change of both growth on a year over year basis, real GDP growth and inflation accelerating at the same time.

Those are typically the most buoyant markets. You know, go back to 2021 when we called quad two, things went completely haywire obviously with all the smidgaps and meme stocks. This time, it’s actually not that different.

If you look down the hood at like things like your rocket labs, which is in the UFO ETF, for example, or we’re getting computing, which is the biggest name in the quantum ETF. You know, these things are absolutely ripping right now. And you know, despite you can get one or one or two data corrections, I don’t think many people called this one on deep seeker, but you know, there’s a, there’s a lot to do out there.

The other big, the other big thing in quad two that you’re long, don’t forget David is commodities as an asset class and energy stocks more broadly. We’re still bullish on both oil and natural gas and the related equities. Okay.

So commodities, natural gas, but what about other sectors in the tech in the equity sectors? You mentioned tech earlier in the interview, how a lot of them are really overvalued like Nvidia. So how do we find growth that is not already very richly priced in? Yeah, I don’t, I typically, uh, I’m laughing at myself because they shouldn’t have started with that. I mean, the fact of the matter is I never start with valuation.

They start with the quad signal, right? So your valuation is not a catalyst for something to go down. The catalyst for something that’s expensive to go down is that the rate of change of their revenue growth slows, which is endemic for now, uh, to Nvidia. It doesn’t seem to be the case for Amazon, for example, um, or plenty of growth stocks that were currently long.

Uh, if our analysts gets it wrong and they show a deceleration in growth, then you’re going to get absolute multiple compression. So, you know, valuation, if it, if this is going to sound ridiculous too, but it’s, it’s true, uh, which makes it not ridiculous. According to me, uh, if some, if, if something wasn’t expensive, I wouldn’t be buying it because the market certainly wouldn’t, wouldn’t, wouldn’t be believing in the rate of change growth story that’s on the come.

And another thing to think about too is cyclicals like energy stocks. Of course, if you’re buying them now, they look expensive, but if you’re using, you know, my WTI number of like closer to 79 or 80 bucks a barrel, you’ll be like, ah, that’s cheaper. So you also have to be careful with that.

How would you bet on the American consumer for 2025? Would you be long staples or discretionaries or neither? Uh, I’m short staples and I’m long discretionary. I’d be long big time, the small business owner in America. Uh, you and I talked about this last time we were coming out of a borderline, not recession, but a depression.

Uh, and now you’ve seen every single NFIB small business survey that we measure map has gone vertical, like in a way that it’s never gone vertical before. Uh, those are north of 99% of American businesses. They’re almost half of the hiring.

So there’s quite a buoyant, you know, again, some people will say, well, look, that’s just Republican. And I’ll say, come on, man. Like a lot of people that run their own businesses do not like getting regulated.

They don’t like getting taxed. And they certainly, uh, don’t, they didn’t like the environment coming into the election. We could show you plenty of data to support that curtailing, or at least the delaying of hiring cap backs, you know, all these decisions.

Um, so, you know, I do think that that’s a bull case for small caps too. Uh, you just have to be careful on which, which specific small gap you’re getting long. You’ve also written about Trump trades.

So, uh, walk us through some of the Trump trades that maybe you haven’t talked about in regards to some of his policy changes. And we’re not just talking about taxes or fiscal reform here. We’re talking about, uh, the deportation of illegal immigrants.

So immigration policy is going to change. Uh, H1B policies may change, um, the regulation initiatives on the banking and crypto fronts. A lot is changing.

Yeah. I mean, in, in, in some cases, our best shorts are based on some of those changes. I mean, uh, not withstanding that I, I’m an immigrant, uh, the, the, the immigration policy is not bullish for housing.

I mean, it’s, we’re short housing, uh, ITB, we’re short REITs, XLRE. I mean, the immigration that we had that was unstated, um, uh, but, but known in this country was a huge buoy to, to, to, to, to plenty of, um, to plenty of housing demand that is, that isn’t as since God. And we also have obviously a view of higher interest rates.

So there’s that there’s, there’s RFK, you know, there’s the Fed on Wednesday, David, but there’s also like the main show. This is gonna be really interesting to watch RFK testify, you know, like he is not messing around with consumer staples companies, things that are bad for your kids, bad for your family. There are plenty of things, you know, go down the line.

I’m sure everything from PepsiCo to you pick a general mills. There’s a lot of, of crap out there that, that people consume, uh, that is, that, that has any to share for town. I mean, there’s a lot of disclosure that needs to happen.

You’d know this, this is a little different in Canada. It certainly is in Europe where they already know the truth. So we’re just waiting for somebody to have the courage to put their hand up and save for our kids.

So I think there’s a, there’s also plenty to do on the, on the short side of healthcare. So a lot of the Trump trades, well, while they can be exciting and I’m not, I’m not really, you know, concerned about the tariffs in particular, uh, that seems to be more of a tourist, a macro tourism type exercise. There’s plenty to be long and plenty to be short.

What’s your outlook for the yield curve and, uh, which sectors are most, or most sensitive to, uh, changes in the yield curve? Yeah, I’m, I’m, we’re long a steepener. Uh, the ETF for that is eyeball IDOL. Um, so we’ve currently, you know, we’ve seen the curve go from inverted to, to positive finally, tens minus twos today, I think was at plus 34 basis points.

But as you well know, as the curve continues to steepen, that would be a signal in addition, uh, will long high yield, um, high yield credit and spreads are, are tightening towards the tights. So that’s an explicit when you get a yield curve steepening and you get credit spreads tightening, that means that you’re in quad too. Okay.

That’s the best environment for high yield, the best environment to be long a steepener. And it shouldn’t surprise anybody if the yield curve pens minus twos, you know, could, could be 50 to 75 basis points, uh, in the positive. So that’s, that’s been a sneaky, good position.

That’s making money for us on a day like today, where, you know, if you’re as long as six, whatever the consensus portfolio, it’s not a good day. All right. Well, tell us about your, uh, currency outlook.

So the DXY is going to be stronger or weaker under Trump, at least for the first two years. Do you think? We think, uh, the first part will be strong. Um, it’d be a continuation of the trend.

Again, a lot of these things, the market discounted, uh, and has been discounting the whole way. So into the election, out of the election dollar strength, I think reflects, um, a lot of different things. One big one, David is actually, uh, having the courage to address the deficit.

Now there are many countries that are doing that, uh, if, if any at all. So, um, you know, you’ve got countries that are being pushed around Canada in particular, uh, with Trudeau to, to, to, to resign post, uh, these kinds of deficit spendings and the social kind of inflation that people feel. But I, I think that this is going to be something that is going to surprise people is that there’s, uh, and there’s an ongoing strength in the U S economy, ongoing strength there, therefore, and also an inflation that will keep the dollar, uh, dollar up and fed policy, uh, not in rate cut mode anymore.

Do you think that, uh, capital flows will favor the DXY because of tariffs? Or do you think tariffs will not have any impact whatsoever on currency pairs? Uh, it’s good. It’s always going to have an impact. It’s not the main mainstay of, of, of why I’m, I’m long the dollar.

I mean, what you’re finding around the world today is that the fiat currency, uh, experiment that is hundreds of years old is failing miserably from South Korea, you know, to, to what happened in Argentina, obviously, um, you have, you have extremely, uh, low levels of trust in government and their ability to either print fiat currency and, or combine that with deficit and stimulus spending. So I think that that’s a huge thing. I mean, you have, like I said, between South Africa and Latin America, you have about 70, 80 currencies that nobody wants.

Um, so the big winners have been dollars, gold and Bitcoin. And these are the things, if you go, like you can go from, you can go to Botswana or you can go to Bolivia and I can guarantee you that they’re going to understand the value of like Mr. T’s gold chain if it was around your neck, uh, or a handful of $20 bills or Bitcoin. Uh, what, what these companies, these countries have left in terms of currency credibility, uh, to me is maybe it’s, it’s probably the darkest hour.

And I’ve been doing this for 25 years that I’ve seen for the global currency market. And I don’t see what the catalyst would be to change that darkest hour for whom, for all these, all these governments, you know, there are a lot of them went to school in America, right? The South Korean central bankers went to my school, you know, I went to these different Ivy towers. They’re all doing the same, same dumb stuff.

I mean, there’s a reason why squid game too is so popular. I mean, the people in South Korea are actually that pissed off. They don’t have any money left.

Uh, there’s demographic issues. And if you keep devaluing the currency of the people on the order of double digits, you know, they’re just not going to take it anymore. So I, I think that that is something that, you know, and there’s plenty of work that’s been done on this, you know, through Reinhart Rogoff in particular, where once you get to certain debt as a percentage of GDP thresholds or deficit as a percentage of GDP, there’s a point of no return.

And a lot of these countries have Russian ruble. Would you be long that there’s a possibly end to the war inside the U S Trump just stopped aid sending aid to Ukraine, by the way. So we’ll see what happens there.

But anyway, I, what about the Eastern European countries, Russia in particular? I wouldn’t belong to anything Russian. That doesn’t mean it won’t work. I’m just buying, I’m definitely not buying rubles today.

No. Yeah. How do you, okay.

Let’s, let’s end on that. So you said that it’s the darkest hour for some of these other countries. And, and a lot of these countries in emerging markets in particular are seeing high levels of inflation.

They recognize gold. They recognize a dollar. They recognize the value of Bitcoin.

Do we in an OECD country like Canada or the U S recognize the value of Bitcoin? Is there a need for Bitcoin? How do you feel? Yeah, I think that there’s an ongoing diversification exercises in play. I mean, I’m not the big, you know, as you can tell, I’m not the Bitcoin maxi with laser eyes. I would never do that with anything that ticks.

But that said, you know, within the context of what I was outlining, the failure of the global fiat currency system, the failure of the elites, the failure of the state to tell you when and how you should spend your money and, or how they can spend your tax dollars. I think that that is, it’s crumbling on itself. Now it can only last so long in Canada.

Again, even they even let it go there for what we have Trudeau for nine years. That’s a long time. And if you look at the other side of it or the alternative, which could be Argentina, which we’re also long ARGT, which is Argentina.

What Mila has done is, is phenomenal. I mean, he routinely mocks these governments and all these bureaucrats and he’s cut out all the spending, which are certainly a big, big piece of it. And other countries like Brazil don’t get it.

Like we’re short Brazil, we’re long Argentina, for example, Lula is going to run that place into the ground, just like any, just like Trudeau was trying to. So you have, I think that this is like Bitcoin’s time to shine when it comes to the replacement value or at least the alternative value of a currency than the local one that is being abused. You think Bitcoin will replace gold? No.

I think it’s a good kind of marketing for them, but you know, gold is not Bitcoin and Bitcoin is not gold. If you really love Bitcoin or a gold bug, I mean, I wouldn’t try to conflate the two. I mean, they’re different.

You know, one is the technology and one is, you know, it’s, it’s sitting right in my safe. I mean, I don’t need to, like, I don’t need to simplify the complex by saying that two different things are the same thing because they’re not even remotely close to the same thing. Okay.

So what’s your outlook for gold then? Gold has been surprisingly strong, David, given that we’ve had, it was weakened as interest rates broke out towards the cycle highs, but since obviously bond yields have come back in and gold has recovered. Again, gold has proved, most central banks around the world are accumulating gold. They’d much prefer to buy gold than they are U.S. treasuries.

So you have, you know, natural incremental buyers. Like I said, it’s one of the stores of value or actually money alongside, I would argue, the world’s reserve currency, the dollar and Bitcoin that people are certainly interested for now. I don’t see what the catalyst would be.

Other than a breakout, like you mentioned, the 5% level on the 10-year yield, that’s where you got to be careful with gold. Gold does not like it when interest rates break out to new cycle highs. Excellent.

Thank you very much, Keith. Where can we follow you and Hedgeye in your work? Just hedgeye.com. And if you can put up with me, you can follow me on Twitter under my name. All right.

We’ll put the links down below. Well, enjoy your talk. As always, Keith, appreciate it.

Speak next time. Thank you. Appreciate it.

Thank you for watching. Don’t forget to like and subscribe.