저자: Kevin · 외국인
요약 보기
딥시크 AI 모델이 저비용으로 유사 성능을 낸다는 소식에 미국 증시가 급락했다. AI 지출은 GDP 성장의 큰 기여 요인이라 AI 투자 축소 시 성장 둔화·스태그플레이션 위험이 있다. 엔비디아는 고평가이며 칩 관련주가 가장 큰 타격을 받을 수 있다. 애플 등 가격결정력 있는 기업이 상대적으로 방어적이다. 미국 밸류에이션은 역사적 고점 수준이고 10년-2년 금리 스프레드 변화를 경계해야 한다.
다듬은 글
oh boy the US economy just walked into the Wilderness sculled with all of its GDP oh man what do that mean for us why is it happening and is this going to be a buy the dip opportunity I'll let you know in this video especially since this is fed week and we've got a lot to buckle up on so let's get started with well first of all I'm going to take this sip of coffee but then I want you to know why why all of a sudden did the NASDAQ now it's recovered a little bit but in pre-market drop one essentially the entire Market sold off $ 1.2 trillion do of market capitalization 3.4% here on the NASDAQ it was even lower just a few minutes ago Tesla is down 3% and if you look at Nvidia it's down 11.56% in the pre-market what is going on and why is Apple only down one qu of a percent well all of it can be explained and that's what we're going to do in this video in this am edition of the me Kevin report news that makes new money all right what's going on well listen look yesterday and uh on Friday I made videos covering the Deep seek story that is exactly what's going on and in this video I'm not going to reiterate to you how basically there is a new artificial intelligence model that allegedly can perform as well as some of the best quality versions of chat Bots from uh Facebook b or uh open AI you know competitor to GPT with a fraction of the training expense potentially as little as 95% or as much as 95% less uh actual compute or training expense to have similar results now a lot of people are saying I don't know there's something sus about the Chinese they're lying to us but the problem is the code for this AI is open source so every hedge fund and institution well they can express Jade on the internet you know they're downloading the app in the app store which is now number one in the app store they're going to deep seek and they're making burner accounts and they're testing it they're running it against the money the hundreds of millions hundreds of I'm sorry billions of dollars that are being spent on compute every single year just for artificial intelligence models that now China is putting to shame consider this for a moment just to train a traditional model usually takes somewhere around half a billion that's the train one model that they just did the Chinese via deep sea allegedly for less than $5.6 million that means we're spending a lot of unnecessary Brute Force money on artificial intelligence chips and server centers but it's not just that money I mean consider the first half of last year we spent about 200 I'm sorry we spent about $120 billion on artificial intelligence and artificial offal intelligence related we think that by the end of the year our US economy probably spent about2 200 to $225 billion on artificial intelligence uh and that's just really on like chip making Investments right now you might not think okay well I mean come on Kevin we have a $23 trillion economy that's you know maybe 1% of our economy why does that matter well there are a few reasons that spending matters number one has to do with the velocity of money so this is an old economic principle that basically says when one person spends a dollar and it goes into the economy it Cycles through the economy and actually generates spending between three3 to4 for every dollar that is spent and there are different types of velocity of money right so when you put your money in a savings account the velocity of money is closer to one when you go spend your money the velocity of money is closer to 3 to four well AI spending is well a form of spending so we actually create a substantially strong contribution to GDP through uh quite frankly artificial intelligence spending and it could be why we've avoided a recession see again if you take let's call it 1% contribution towards GDP it doesn't sound like that much but then recognize oh my gosh GDP is growing at about 2 and 1 half% to 3% with AI and then apply a 3X multiplier to that 1% that means all of the 3% growth could be attributed to artificial intelligence and the additional spending that comes via the velocity of money this is a really big problem because it potentially means our GDP without artificial intelligence spending if all of a sudden we pair back the spending if you know that's a big if right we could keep spending but if all of a sudden corporations like Microsoft and meta are like damn we've been duped well you know what we've got Big Data Centers let's try to use the technology deep SE cast with all the infrastructure we have let's pause on new infrastructure spending and let's try to innovate with what we have which we have plenty and now all of a sudden you have a license basically for Amazon meta Microsoft Google to spend less money on artificial intelligence and say let's work with what we have well then all of a sudden you're in a place where maybe you end up getting the contribution to to GDP sitting instead of at 1% and then as a result the velocity of money at 3% you actually end up sitting at 0% GDP growth well what happens when you have 0% GDP growth and a Federal Reserve that is freaking out over a Resurgence in inflation well what you have is textbook stagflation and the Silence from people like Elon Musk actually speaks volumes to how much of a risk deep sea could actually be if if you scroll through Elon musk's latest tweets there is not a single mention of deep seek and he tweets a lot why all of a sudden are we going quiet now maybe I've missed a reply somewhere but usually Elon is very very vocal and there is a potential there is a chance that maybe elon's looking at this going oh my gosh this is the valuation of xai poof that's the extreme I'm not saying that's exactly what's going to happen I'm not saying the economy is going to go into stagflation but we're going to have a highlighter on a Fed meeting this week and we're going to be looking for something very specific we're going to all be going to meet kevin.com because we've got a trading challenge that starts in like four days here and a huge trumponomics coupon code so if you're nervous about any of this that's going on check out at's because if you want to get through the m Madness that 2025 is going to be in the market and Via Trump you want somebody on your side every day we are going to be conducting cour member live streams you get to be a part of them every single morning ask me any questions you have but you're also getting access to all of the trumponomics content including the trading Challenge and options alerts that begin February 1st in addition to that we're releasing content around Tax Strategies under Trump investing under Trump hedging under Trump real estate under Trump building an entrepreneurial brand Under Trump and marketing under Trump all aspects of essentially making money under the Trump Administration we are covering in Trump atomics and the amount of content that's out right now is a fraction of the new content that's coming which means if you lock in your price today you get all of that content going forward for the life of the course under Donald Trump so check out the trumponomics course join trumponomics I can't wait to see you there and get in before the price goes up because the price will probably be double where it sits right now and I can't wait to have you if you have any questions email us at staff meetkevin.com otherwise check it out over at meetkevin.com so what else do we have to know well we have the Federal Reserve this week this is very very important the Federal Reserve this week is expecting to tell us hey we are going to pause but markets are actually going to start pricing in the potential for more interest rate Cuts if GDP tumbles now remember I have been one of the people who's been basically called a clown because I've argued that at some point there is going to be an efficiency movement I have said time and time again that when you look at a chip like Blackwell that uses 25% of the energy as an h100 to perform the same type of training then you don't need as much energy and as much electricity or utility Investments as people think now what's actually happened is not only are you getting Chip based efficiency but now you're getting model or demand side efficiency which means your demand for you energy will fall under new efficiency models but also your demand falls under more efficient ship models this makes utilities specifically exposed to unfortunately the negative risks of efficiency from better training compute especially since you're in a place where you don't actually have that much profitability coming from artificial intelligence today A lot of people are wondering is this overblown spending now look I'm a big fan of artificial intelligence and the things that can come out of artificial intelligence consider for a moment that companies like Tesla or I hate to say it but Tesla's competitors might actually now be able to engineer full self-driving versions much more efficiently than they have previously the holdup is all going to come down to data Tesla has the highest access to data today but what happens when training gets so inexpensive now you could actually train using artificial intelligence simulated models plus collecting data or even buying data well at some point that Moe for Tesla's FSD will begin to shrink and I have concerns that at some point that moat shrinks before Tesla actually makes it to the level of Robo taxis allowing multiple entrance of Robo taxis at the same time this is not to be bearish Tesla I actually think I'm very very overall bullish long-term Tesla Optimus Robo taxi very excited the problem that we all face though right now comes down to valuation and I hate being that person that brings up valuation but let's just consider for a moment what valuations look like in the market market so when we jump into US versus global equities and valuations thereof take a look at what we have we have the Magnificent s sitting at the highest level that we've seen compared to the rest of the world ever now keep in mind and you know these These are obviously you know adjusted over time here uh the problem with this chart is it doesn't consider the fact that earnings are disproportionately showing up in the United States because of our artificial intelligence so I actually think you're better off using an inflation adjusted sort of PE multiple right a multiple of earnings how many times earnings on an inflation adjusted basis are US Stocks trading for and how does that compare to other points in history and this is actually where you come up with the buffet measure or the cape Schiller uh PE multiple uh the cape ratio uh for what do we have valuations in the US Stock Market and what you'll find is that us valuations right now sit at the high that we sat at in December of 2021 which was also the time the Federal Reserve became more aggressive again and that's somewhat what we're going to hear likely this week an aggressive fed in the face of inflation fears also while the market is going to be concerned about growth fears it's not an ideal Duality that's going to be very similar to what we saw in December of 2021 and the fears today about the Federal Reserve actually began on December 18th it sounds very similar to December of 2021 which is when we had a similar valuation Peak if we go back to thec recession or bubble we actually see that valuations were quite a chunk higher if today we're sitting at uh 3847 then we sat at about 44.5 so if I just divide those into each other uh we could see that valuations were about 15% higher in the bubble than they are now but we're still at the fourth most expensive time in the history of the stock market and what I bring up that the stock market is at its fourth most expensive time in history I tend to get comments that make me very nervous I get comments from people that say things like but Kevin everybody has an app on their phone today and everybody can buy the dip on stocks so what's going to end up happening everybody can buy the dip and stocks will never crash again and it's as soon as I hear people say things like the stock market's never going to crash again or the business cycle is dead that I start think or quite frankly when like you know random people at grocery stores working the checkout lines are like Hey Kevin which meme coin should I buy I heard if I invest $1,000 I could turn it into $100,000 I get a little concerned that we're close to a cycle top especially since now you've got oh okay micro strategy today announced the launch of STK a new convertible preferred stock offering ah for institutional investors right okay so at least micro strategy is going to buy the dip by diluting their stock more that'll be fine I'm sure UBS Cuts hundreds of Swiss jobs and latest wave of cuts all right whatever anyway so what do we want to pay attention to well not just valuations right because remember nvidia's valuation actually looks pretty decent if you assume the growth that they have been projecting nvidia's valuation looks like they're Trading for a 1.7 Peg it's actually pretty decent but that's based on 30% growth well what does Nvidia look like if you cut their growth down to 5% well it looks like a 10.2 Peg which is more expensive than paler is today just crazy so then you have to wonder okay well what stocks are most resilient to any kind of sort of AI selloff well certainly not advantest Corp which is an Nvidia supplier which was down over 8% in Tokyo trading uh is potentially the uh Chinese stock market I know some people are investing in the hangsang tech index which was up over 1.3% a lot of people are frustrated with SoftBank because they see them as dumping money into arm and they own a lot of arm and SoftBank stock was down over 6% in Japanese trade today and we'll see how arm performs uh today but a lot of folks Wonder okay maybe the place to actually go run and hide uh is one of two places you look for a pricing power style stock uh that isn't highly exposed to artificial intelligence spending but could be a beneficiary of more efficient AI models what kind of company does that sound like well to me it sounds like apple apple could potentially be a beneficiary here and I think that's why they're only down 0.13% in fact I mentioned this in my video yesterday that Apple has a big moat just like Facebook people are still going to spend with Facebook ads people are still going to buy iPhones and these products will be enhanced by the commodity that is a good chatbot now I got a lot of complaints yesterday that oh but Kevin you know this company secretly used a lot of Nvidia h100s that's fine and in fact I already addressed that in my videos that has nothing to do with why Nvidia should somehow not sell off in fact it has everything to do with why Nvidia should sell off even if deep seek is is using the best Nvidia chips available today they still only need 5% of the compute power 5% of the compute power is 12th so for every one black CH blackw chips somebody else is buying uh or I should say for every 20 Blackwell chips another company is buying they only need one I don't understand really how some people believe that that all of a sudden is going to create more demand for NVIDIA and I broke down my thesis on that yesterday so if you haven't seen that yet make sure you watch the full video yesterday and you'll understand some of the context around this now some of the comments that were coming in were things like oh but Kevin you know this just means that people are going to be able to research more and their demand will go up this um assumes that artificial intelligence models are going to be better than they are today we are not limited we're not practically limited with a supply of artificial intelligence capacity today but it's just sort of at a 40 level a GPT 40 level and then add you know the asterisk of reasoning for for 01 but that's not good enough to actually get us to AGI we're a long way away from that we quite frankly have a commodius dictionary That's What markets have today everybody now has a chatbot which is as good as a dictionary that actually tells you what you want when you have questions about it and so the second potential investment that could be very interesting to pay attention to in today's market would be bonds bonds are a bit more of a flight to safety tool and they're really only something that you should consider using because they could lose you a lot of money they should really only be something that you consider using if you think that inflation is going to go away and you need a downside Market hedge so take it with a grain of salt but the bond market is up quite a bit today yields are down which means bonds are up 8.7 bips on the 10year is great now what you want to do is you want to look at the difference between the 10 and the 2-year we're at uh 4.
53 here we're now at 42 here so we've got about a difference of about 33 basis points what you want to watch for is is this going to be the Catalyst that all of a sudden pushes that spread between the 10 and the 2-year up to about 50 to 90 because if it does that's when you trigger the recession alarm and you want to be very very cautious I'm not saying that's going to happen but I do want you to take away from this video that there are going to be great companies Apple Microsoft meta that yes maybe they overspend on AI but they're still going to have a lot of pricing power but your chip and Chip adjacent companies are probably going to get hit decently hard in this sort of environment if the data ends up being true and again since it's an open-source model it probably is the these efficiencies are going to spread to all of the other AI modeling companies and the demand for uh quite frankly new compute chips may fall off a cliff rapidly and therefore your chip and Chip adjacent stocks are likely to get hit the hardest including plans for Stargate really curious to see what someone like Elon Musk eventually says but I do think your highest pricing power stocks like again even a Tesla from from this point of view at this point with the FSD technology they have they should be more insulated yes they spent money on chips that maybe they don't need as many of right now but that's fine they have the tools they'll figure out how to use them but they don't necessarily need to buy more will this lead to a broad and continuous market crash I have no idea I hope not because if it does you have to remember that as soon as the stock market really and deeply and meaningfully sells off we're probably going to end up seeing a lot of layoffs and I do not want to see a layoff recession so with that said I encourage you to head over to meetkevin.com check out the courses on building your wealth especially that trumponomics course you're not going to regret that purchase promise too you're going to love it and I look forward to seeing you remember trade alert starts soon and we'll talk soon thanks bye goo
Comments 0