저자: 미상 · 외국인
요약 보기
AI 주식 성장 둔화(성장의 크래시)가 이미 진행 중이라고 주장. 엔비디아 매출 성장률이 전년 대비 2배→분기 60%→34%→전망 10%대로 둔화, 마진 압축 우려. 엔비디아 적정가를 PEG 3배 기준으로 550~600달러로 제시하며 현재가(약 730달러)는 고평가로 봄. 슈퍼마이크로는 마진이 낮아 성장 둔화 우려가 더 크다고 언급. 인텔은 파운드리 관점에서 분산 투자 후보로 언급. 단기 하락 예측은 명시적이지 않고 성장 둔화에 따른 주가 압력을 시사.
다듬은 글
AI stocks will crash in this video I'm going to explain exactly what is going to crash what that's going to mean for the prices of stocks and we'll specifically touch on Nvidia super micro computer and we'll apply this to the broader industry of where opportunities could be I want to be very clear upfront I am invested in ships and ship making Technologies both personally and through an actively managed ETF that I run now in this video I'm going to make some very clear concern concs known to the world the last time I discussed Nvidia on its own video I predicted that the stock would go to the $600 to $700 range I predicted that would happen in 2 years this makes this a $600 stock by 2026 with this sort of growth rate and if we include the other aspects of the business you're actually at close to an $840 stock you're looking at potentially a you know what 80% return over the next uh 3 years to potentially a double even at today's levels at 375 unfortunately it happened within 9 months or I guess fortunately depending on how invested you were it happened way quicker than expected so my timing on that was wrong but now given that we're sitting right up against the next Fibonacci retracement level which is serving as a resistance for NVIDIA it's about 7:31 we have to ask ourselves are we ready to move to 969 or are we going right back to where we were at 493 where we topped out in this video we're going to work to address this we're also going to consider super micro computer who has maybe an 8% net profit margin selling you server racks to basically put your AI chips on but currently has a relative strength index of 90 down from the 95 that it had this morning so in other words something's either got a relax here or the growth probably has to stop at some point in the stock evaluation because it's done a pretty dang number already and from a technical basis it is ready to either pause or come right back down so in this video we're going to present the potential for the AI crash now what we have to understand is that AI will crash but first we have to realize how it will crash so let's come up with an example to make this relatable and simple let's say you do not have a cell phone you have no phone at all so in order for you to check your email you have to go to a computer you have to go to a landline to make phone calls and maybe you're faxing things that need to show up quickly I I don't know whatever right you don't have a phone we all know we all have a phone everybody's like come on man how is this relatable well because I want you to think about for a moment not having a phone and then I want you to think about having a phone what you've done is you've thought about going from zero productivity in your pants in your pockets to a lot of productivity okay now I want you to sincerely tell me how much does your productivity go up when you go from that iPhone 14 to that iPhone 15 not much more you're probably about the same level of productive maybe you're 1 or 2% more productive but you infinitely more productive than before you had the phone well and of course we could argue about that we could be well you get distracted in the bathroom on Tik Tok or whatever this really a topic for a different video but the point is you went from zero to a massive level of productivity having a phone having the Next Generation phone your productivity probably didn't move that much that is the difference between an infinite growth curve and a marginal growth curve and what is that going to look like for AI and the AI moment well in my opinion and this is inspired by a course member lecture that we did this morning this AI had a massive growth impetus companies went from we have ai or or sorry we we have no AI to now we need AI servers okay so we need AI servers so what are we going to do we're going to call up super micro computer we call up super yo we need we need a th server racks as soon as possible hey AMD we need 1,000 CPUs Nvidia we need 8,000 h100s everybody's going from zero to a lot and that's how when you go to a company like Nvidia or super micro computer you look at their growth year-over-year in like q1 Q2 of this year you're double right these companies revenues doubled it happened to Nvidia it happened to Super Micro computer you've seen it at that companies like AMD I mean look at Nvidia for example you went from $7 billion of Revenue uh ending April 30th of 2023 to double that 13.5 well so roughly double that right uh that's pretty incredible that's actually really really good that's great because the quarter before that they they didn't grow much at all in fact they shrunk so what's remarkable is you had a PC hole where now Nvidia is trying to get you to install llms on RTX PCS to motivate you to buy more PCS so you can download your own RTX chatbot now and install it it's like a 35 GB file it's a big deal a lot of people are talking about it actually have it right here I just haven't finished installing the darn thing uh but anyway what what's remarkable is you went from a company that was really focused on gaming computers and yes to some extent helping in automotive and servers to having a product that was perfect for this moment this spark the AI the moment of companies going from we have cloud compute we have cloud storage but crap we have no inference we have no AI okay so what do we do well we create the AI servers now we have ai servers do we need to replace all of our h100s that we just overpaid for you know $35,000 for something with an MSRP of $8,000 do we really need to go replace all of those to get the h200 which is all the rage right now why because well Nvidia tells you it's all the rage it's so much better why would you buy the h100 or the cheaper a100 when you could have the h200 and soon we'll be on the B100 keep in mind that like these charts are designed to sell you the product the reality is all of these chips are incredibly powerful and if you actually showed the whole bar chart you're probably only looking at like nominal differences that they've zoomed in on but that's okay that's that's how product refresh Cycles work notify me when the product becomes available for the h200 tensor cord GPU point is that the next model of the GPU just like the iPhone 3G or the 3Gs thereafter is not the big deal the big deal is the moment you go from iPhone or from no iPhone to iPhone or no phone to phone right for you Android folks I have an Android too okay it's actually pretty good I just hate that I can't easily get files it's like FaceTime somebody oh here's your Google link what what's that oh I shouldn't I shouldn't go down this shouldn't go down that path uh I do have it it it takes great photos I will tell you that the pixel is pretty good but this is going to happen not only is it going to happen this crash is already occurring this is not necessarily a crash in the stock price however a crash in stock prices can happen we're going to talk about that in a moment this is a crash in growth so when I say AI is going to crash that is not speculation AI is going to crash it will crash these a company's growth levels will crash there is zero question about that because we already know it's happening go into nvidia's earnings sorry we got a bunch of notes written on here from what we do in our course member live streams but look we've gone from doubling Revenue quarter over quarter to Growing Revenue say 60% quarter over quarter to Growing Revenue oh 34% now that was a beat that 13.5 million uh uh billion was a great beat it's 13,500 Millions is the way this is written so therefore 13.5 billion that was a beat but we went from doubling our growth to Growing 34% over each quarter and that's still phenomenal growth but wait a minute now they're only projecting 10.3% growth ah it's slowing quarter over quarter of course it is because now everybody who needs an iPhone has the iPhone just to stay with the scenario right okay so so now Nvidia does have a tendency of beating right now they're forecasting $20 billion in Q4 their earnings are you know next week on the 24st of February they'll probably beat again they usually beat by a factor of 1.8 to 1.48 post AI don't worry about how I came up with that the point is they'll probably end up showing somewhere between 15 to 19% growth in Q4 Cal their their uh rather sorry their fiscal Q4 2024 in other words the next time they have earnings okay so they'll probably show some some good growth but it's not a double anymore it's not even a third growth anymore so the point is the crash of growth is happening so then we have to evaluate okay when we conduct a fundamental analysis on a stock we generally look at earnings okay so what what are earnings going to provide us how much cash flow am I going to get now in the case of Nvidia the cash flow generation is phenomenal but for that's for the company's cash flow your cash flow is earnings per share you get earnings per share growth then the stock either goes up or they pay it out in dividends so earnings per share growth is one way that we can it's just one of many ways we can conduct a fundamental analysis on a company and so what we want to see are earnings growing earnings can grow in a number of ways they can grow through more Revenue adding a new vertical like the AI inflection point that drives the Top Line way up and of course the botom line way up uh or you can become more efficient by potentially laying people off or spending less money on research because you maybe you don't need to as much anymore whatever right those are all ways you can grow earnings when it comes to the artificial intelligence Revolution that we've hit once companies have their artificial intelligence the question becomes how much is earnings uh per share going to grow at AI related in AI adjacent companies like super micro computer after we're done with the AI CR well part of this has to to do with a very special word margin if you were previously which is a pure definition of of absolute pricing power as a result of supply chain shortages if you were pricing h100 chips at $35,000 and your MSRP is $8,000 and your cost to build the chips is $44,000 so that's cost that's manufactured suggested retail price and you're able to sell them your actual sell price is $35,000 well the reason there's a wedge here is because there's a supply shortage so you raise the price it's kind of price gouging really rich companies is what it is and that's okay like whoever is willing to pay the most should get the product that's how capitalism works it's like an auction effect when there's a shortage okay fantastic so what happens when that shortage goes away because that boom the adoption of AI is gone well when the boom goes away then what also goes away well it's obvious your ability to charge $35,000 for a chip goes away now don't get me wrong it's so fantastic that you could sell these chips that you're not even making Nvidia you're just designing them that's not to like insult you which is like it's really impressive right um the point is you're going to see margin compression from AI because we are at insanely inflated levels of margin today so there will be margin compression as Supply equalizes and of course competition increases I mean don't count out AMD here or some of the other competitors trying to get into AI there there will be competition and eventually just like there's the Android and the iPhone choice or whatever there will be many equally capable some slightly better than the other chips sitting here and like comparing specs of different chips doesn't matter because what really matters is going from no AI to AI okay but that's happened so the price levels to some extent are Justified because we do now have a new Revenue vertical and we do have more earnings per share which is likely to be durable that earnings per share will will likely stay for a while so I mean Nvidia is now pumping out nearly $5 in earnings per share the question is when does that $5 per share change well it could change very slowly unless there's another AI impetus consider that by the end of 2025 will be at6 to $6.50 current forecast so and and before that just to go back another year we were sitting at like 60 cents of earnings per share so draw this for a moment right here you're at 60 cents of eps then you get to $5 of eps per quarter and then you're at $6 of eps per quarter of course the growth rate is insane over here and then of course you're going to get articles like you get from Barons they're being some Knuckleheads right now uh Baron's uh put together an article and they're like oh Nvidia is really really cheap look at this if you take the growth between 2023 and 2025 nvidia's PEG ratio is just3 whoa whoa whoaa wait a minute wait a minute you can't take last year's growth levels and average it with the forward growth levels because we just went through a massive adoption phase I mean it's the same thing as having like a giant one-time expense okay let's say on your house you spend $30,000 a year to live in your house that's your payment and maintenance okay the furnace goes out whatever $30,000 a year is what you spend but one year you spend $60,000 because you install a solar system oh my gosh your costs of living in your home just doubled well yeah but it it was a onetime adjustment then we're back at 30 right so now we're back at normal level so we have to adjust for those things it's the same thing for Revenue okay if we go from zero to say 60 and then we go to 66 we went through infinite growth to 10% growth so the growth is still going to be there it's just going to be more incremental as you go through product refresh refresh Cycles just like you do with the iPhone okay fantastic so what does that mean for valuations and how do you actually run a valuation on a company like this okay so this is what I like to do uh personally okay so this just put it all out there I'm going to make this very simple and we'll run through a few of these together so I like to discount for this AI segment future growth rates and what I'm going to do is I'm going to run Peg ratios based on what I think is personally reasonable so this is you know not a guarantee but we look at a company like Nvidia nvidia's growth that is projected over the next and it's so hard to look out right but the next four years of growth for NVIDIA are currently projected to be 70.5% PS growth 16.5 17.2 17.7 that's 121% but we divide that by four we have an average growth rate of about 30 so 30.5% EPS projected growth I personally think Wall Street analysts are probably being too bullish on this number why because product refresh does not mean the growth goes on forever A and B margin compression literally barren is screwing this up trying to tell you that Nvidia has a23 Peg using last year's momentous growth you can't do that it's stupid it's it's a great way to like delude yourself so what would I rather like to do well I would rather like to go in here and uh we're going to go slash that because I am worried I'm substantially worried about margin compression okay I'm going to say if we could just keep the earnings and grow it a little bit we're doing good so let's say we grow at 10% nominal product refresh cycle growth let's just say this could be way too bearish for the company okay but I'm going to go with it anyway I'm going to go with 10% the current price of Nvidia is sitting at about resistance so 730 divided by uh earnings per share for this year uh which is the earning cycle that uh is being announced next week we're going to divide that by 1236 which is the expectation puts us at 59 times so now I'm going to divide that by 10 my PEG ratio for NVIDIA is actually 5.9 that's pretty high with 10% growth so if I think that the PEG ratio should be 3% because sentiment is strong and people want to allocate to AI be part of the future whatever it's a stretchy PEG ratio but let's go with 3% then the stock is probably worth half of what it's trading for right now at 10% growth that might be too aggressive and and and I realize that I don't think 375 is the right number so what happens just so you can see how quickly this could change what happens if we assume Nvidia has 15% growth and we're okay with a three peg okay simple so we can back this in $750 divided by the 1236 1236 that gets us to that number about the 8ish right and in this case sparing you the math a reasonable price range for NVIDIA in my opinion is probably 550 to 600 that's probably where we ought to be and if growth ends up going lower than this then I think the stock will get substantially squeezed down possibly back to that 495 rage region it all going to come down to growth projections if wall Street's right and we end up getting this 30.5 EPS growth well by all means then that's fantastic now we're at a onepeg so that's great and if it's supposed to be trading if we're going to have 30.5% growth going forward and this stock is going to trade at a thre Peg then a fair valuation based on what Wall Street is saying is actually 1130 so where growth ends up the longer term 4year growth pra trajectory where it ends up is what is going to decide the value of these companies so now as an investor we look at let's go optimistic and say Wall Street is right we're going to go 30% okay what's the potential growth we have for the stock well 1130 divided by today's price 730 still 50% growth in it in that optimistic scenario it's also possible wall Street's too pessimistic that AI is going to grow even more than anybody ever thought possible in that invent well the stock could be worth a lot more what if what if growths 40% you know run that for example 40 % take to 1236 we'll take a three peg uh that's going to bring us to $1,500 a share so everything's going to come down to where what that terminal rate of growth is for these companies on an annual basis and that's why I'm saying the crash will come in growth and I'm concerned because we've gone through this momentous shift that analysts are being too optimistic on these companies for growing so that's my take on Nvidia now how does this Supply to companies uh that are AI adjacent so what about like a super micro computer well I would actually be even more concerned for a company like super micro because I think once the servers are installed you're not refreshing the server Rex you're refreshing maybe you're going from the h100 to the B100 right you do every 2-year cycle refresh unless you're expanding more servers you're not necessarily buying more racks and so I'd be more concerned for a company like super micro computer and I'd be more skeptical of their EPS growth going forward than I would be for NVIDIA like I I have less concerns about Nvidia than I do for super micr computer but then again their stock does look really beautiful for now uh so what about a company like Intel well Intel really hasn't even had an opportunity to participate here yet Intel's really just operating in my opinion I don't underwrite Intel as a company that's really a chip designer I underwrite them as a chip maker a Manufacturing Company a plant that's being subsidized by the US government via massive chip supporting stimulus checks that that game is still that still has room to play out I think this is a 6465 stock I think it's a diversifier it's it's a you know could be a value trap but I I see it as an opportunity to diversify with so then you have to ask yourself after growth crashes for the AI companies are people going to say all right I guess I have to go find another growth stock and then sentiment shifts probably sentiment will shift sentiment is very bubbly right now it's very fomo uh hugely momentum driven on these stocks that's why the rsis are as high as they are that almost Rhymes but anyway sentiment will shift at some point and that sentiment does not necessarily mean the sock will crash look nvidia's RSI went over here to 49 to 50 when the stock was flat the stock just has to stop skyrocketing and the RSI can come down you don't necessarily have to have these stocks crash in stock value so uh for me it's very difficult to short these companies certainly with uh like short-term options I mean you could go into earnings obviously and try to play a short here uh the volatility is a little high on Nvidia right now so you'd probably be better off selling options what I would personally like to do is make sure that I'm Diversified as well as possible into where's the next wave I don't know that we're ready or primed for like the next big AI breakout like that next inflection point I don't think it's here so then the question is okay well what other stocks would people go into once they start selling the AI sector or we stop having as many inflows probably at this point that becomes a better investment than chasing AI after what it's already done knowing that the growth will and is already crashing everything next week is going to depend on nvidia's guidance if Nvidia tells us oh my gosh growth is accelerating way faster than we thought all of our estimates are wrong all the orders that came in in January oh my gosh there are even more than we even thought possible we thought last year we had shortages now it's even worse the stock will double it's just not ending if you know they beat slightly and then they give their forecast they're like okay so we grew 18% this quarter next quarter we'll probably be at the same level and they go flat growth quarter over quarter try to do a PEG ratio on a company with no growth it don't look very good you don't use Peg ratios for companies with no growth it's not possible uh but uh but the point is you can't divide by zero but uh the point is at some point and we don't know when that is growth will flatten margins will compress and there is substantial risk that EPS projections are way too optimistic there is also the potential that you could end up with negative EPS on what I call Tough comps that's really bad because now you have negative growth you want to know what company had negative growth well there are two I could think of one is called end phase and it did that another's called Tesla investors hate negative or shrinking earnings per share but when you have companies and this is really important to remember too when you have companies with really really bad numbers for a year what you've done is you've set up an easy beat next year people like oh the numbers have stopped going negative oh wow look y'all are growing like crazy duh cuz you're comparing to a low number and then the question is can you sustain that growth how long can you sustain that growth right but now it's the opposite at Nvidia or or a lot of these AI companies it's like oh the growth is so hot oh why is it going down it's all going to come down to growth I think we had our AI wave I don't know we're going to have the next part of this massive wave for another 5 to 10 years so I'm becoming a little why not adver my oh
댓글 0