Jordi Visser / VisserLabs
Pánico por la IA, desplome de Bitcoin, alertas de burbuja… y la oportunidad que se esconde tras el miedo
Pregúntale al tutor sobre este vídeo
Entra con un email verificado y pregúntale al tutor lo que quieras sobre este vídeo.
TL;DR
Resumen de investigación
▶ Pánico del mercado y decisiones como apuestas
Desde Maine, antes de Acción de Gracias, el presentador describe una semana que sus interlocutores vivieron como aterradora, aunque dice que no ocurrió nada grave y que la mayoría de las noticias fueron buenas. Expone que el S&P está 3,5% por debajo de sus máximos y su cartera, 35%; atribuye parte de sus pérdidas a compras de MicroStrategy y Bitcoin, que pesan mucho en su patrimonio, y recuerda compras que considera acertadas en Micron, Nvidia y Eli Lilly. Recomienda *Thinking in Bets*, de Annie Duke: como nunca se cuenta con todos los hechos, cada decisión es una apuesta y conviene recurrir a probabilidades y procesos en vez de dejarse llevar por emociones.
▶ Corrección, sentimiento y señales de crédito
El S&P cayó 1,8% en la semana y el Nasdaq 3,13%; el Russell 2000 bajó solo 70 puntos básicos y llegó a recuperar sus pérdidas el viernes. El presentador caracteriza el retroceso como correctivo: el S&P habría cedido solo 20% de la subida previa y cerrado bajo su media móvil de 50 días; cifra la corrección del Nasdaq en 24%. Señala que otros mercados globales superan a EE. UU. en el año y que el miedo, la volatilidad y la compra de opciones put alcanzaron niveles extremos, comparables con los de abril. A su juicio, el dato clave es que los diferenciales de bonos basura no se han ampliado: reconoce estrés en partes del crédito privado, pero lo considera aislado. Compara las caídas actuales —5,6% en el S&P y 8% en el Nasdaq— con retrocesos anteriores y dice que espera una recuperación, aunque admite que un estallido de la burbuja de IA podría traer una recesión. Cierra señalando que *The Economist* ha marcado mínimos con acierto en ocasiones anteriores.
▶ Debate sobre la burbuja de IA y evidencia de adopción
El presentador acepta el argumento de Michael Burry de que las acciones pueden alcanzar su máximo antes que el gasto de capital, pero sostiene que hay que considerar también los diferenciales de crédito y otros indicadores. Critica a Jeff Gundlach por advertir sobre préstamos de baja calidad y crédito privado, y por calificar la IA de exagerada. Menciona que, por primera vez en 20 años, gestores dicen que las empresas invierten demasiado; interpreta el gasto y la automatización como posibles fuentes de mejora de márgenes. Los CDS de CoreWeave reflejarían una probabilidad de incumplimiento de 40–45% en cinco años; el presentador considera alta esa cifra, pero estima que existe una posibilidad de 30–35%. Lo atribuye, en parte, a que la economía en forma de K y la incertidumbre sobre energía y eficiencia pueden afectar el crédito; dice que no ve esos CDS como prueba de una burbuja. Advierte contra pronósticos de innovación a más de cinco años. Como contrapeso al temor, recomienda atender a la visión de Eric Schmidt sobre la IA y comenta que Gemini 3 le pareció un modelo multimodal destacado. Reproduce la valoración de un participante de una mesa redonda: conversar con máquinas para crear software podría representar un cambio profundo, comparable al cambio que atribuye a la era del código y la concentración tecnológica posterior.
▶ Ingresos de IA y pruebas de una expansión real
Nvidia informó ingresos trimestrales récord de 57.000 millones de dólares, un aumento de 62%. El presentador considera que la empresa sigue barata dado un gasto mundial estimado de casi 5 billones de dólares en centros de datos durante los próximos cinco años, del que Nvidia podría captar 35–50%, posiblemente hasta 60%. Cita una expectativa de ingresos de unos 425.000 millones en cinco años y dice que vigilaría una valoración cara, la falta de compresión del múltiplo precio-beneficio o la caída de márgenes. También menciona el crecimiento de ingresos de Palantir, 63%, y de Meta, 26%, y sostiene que las compañías del S&P 500 registran crecimiento de ingresos y márgenes récord relacionados con la IA. A su juicio, la IA puede presionar el empleo y a las empresas que no la adopten, aun sin una recesión formal, y contribuir a una economía en K y a transferencias públicas. Señala que los precios de DRAM seguían subiendo en datos de septiembre y espera que los PMI aumenten el año siguiente por la infraestructura. Para evaluar una burbuja, argumenta que los diferenciales de bonos basura y las acciones bancarias suelen mostrar problemas antes que las acciones; afirma que los bancos suben 13% interanual ahora, frente a caídas cercanas al 20% en máximos anteriores. Compara el alza de cinco años del Nasdaq antes de la burbuja puntocom —800%, de 100 a 917 en su escala— con el avance actual de 95%, y pide comparar los episodios con esos datos.
▶ Gasto de capital y debilidad económica
El presentador coincide en que el gasto de capital en IA oculta debilidad económica: sin ese gasto, cree que el crecimiento del PIB sería más cercano al 1% que a casi 4% en el trimestre actual y alrededor de 3% en el anterior. Considera extraordinaria la inversión en infraestructura y compara su peso en el flujo operativo con el despliegue de 5G, pero juzga engañosa esa analogía: los hiperescaladores serían de alto margen, poco apalancados y con mucho flujo de caja, más parecidos a los auges de semiconductores, nube, ferrocarriles o electrificación. Sostiene que hay que entender usos posibles de la IA, como curas de cáncer o sustitución de tareas, para juzgar su impacto. También reconoce señales recesivas: el indicador económico adelantado permanece negativo desde 2022; señala recesión en inmuebles comerciales, debilidad de la vivienda y falta de creación de empleos, así como PMI manufactureros muy débiles. A la vez, dice que las ganancias por acción del S&P siguen subiendo mientras cae la contratación temporal, combinación que considera inédita. Recomienda una conversación que tuvo con Dean Kernut para escuchar su visión macroeconómica y sobre derivados.
▶ Bitcoin: caída, errores y apuesta alcista
El presentador reconoce que ha estado muy equivocado: esperaba que Bitcoin cerrara el año cuatro veces más alto y que el petróleo subiera; el petróleo, en cambio, siguió bajando hacia 50. Cree que el petróleo y la electricidad subirán el próximo año. Bitcoin cayó 36% desde sus máximos de menos de siete semanas antes. Tras duplicar y triplicar su posición durante la semana, admite que volvió a equivocarse al decirle a Anthony Pompliano que los mínimos habían llegado antes de otra caída de 7.000 puntos. Aun así, no cree que Bitcoin permanezca abajo mucho tiempo; menciona posibles descensos a 70 o 60, sentimiento extremadamente bajo y un RSI inferior a 21. Compara el pánico con abril, cuando publicó *The Light at the End of Turbulence*, y mantiene su apuesta alcista por la expectativa de PMI más altos, la tokenización y las stablecoins, aunque describe la caída como una liquidación y reconoce que otros pueden dudar de su tesis. Enfatiza que Bitcoin solo pierde 9% en el año tras dos años consecutivos de subidas superiores al 100%, y que un rebote por cobertura de posiciones cortas podría hacerlo superar al S&P. Atribuye parte de la presión a ventas de inversores minoristas y a la oferta de empresas y vehículos de activos digitales, incluidos los títulos de MicroStrategy; también ve correlación con acciones especulativas, factores de crecimiento asiáticos y un retroceso de Circle hasta el precio de su salida a bolsa. Recuerda que jornadas de apertura y cierre opuestos de 1,5% en el S&P han sido muy infrecuentes y compara episodios pasados, incluidos el Día de la Liberación y octubre de 2008.
▶ Probabilidades, revisiones y preferencias de cartera
Retoma la idea de Annie Duke: en la vida, como en el póquer, las decisiones se toman con información incompleta, variables ocultas y azar; conviene estimar probabilidades, evaluar valores esperados y actualizar las creencias con nueva evidencia. El presentador dice que se preocuparía si las revisiones de ganancias se volvieran negativas, pero afirma que siguen elevadas: la media móvil de 20 semanas fue 0,16 en la lectura anterior y está en 0,15. Sostiene que las revisiones mejoran globalmente y con más rapidez fuera de EE. UU.; por eso todavía prefiere acciones internacionales a las estadounidenses, aunque sea alcista en IA, y también favorece las pequeñas capitalizaciones. Cita una previsión de casi 60% de crecimiento del beneficio por acción para algunas empresas pequeñas y ve potencial de recuperación para pequeñas y medianas compañías tras tres años de mercado bajista, si suben los PMI o bajan los tipos. Indica que la amplitud del Nasdaq, NYSE y S&P está sobrevendida. La probabilidad de un recorte de tipos, que había bajado de 50%, volvió a superar 60% después de que John Williams, presidente de la Fed de Nueva York, señalara margen para otro recorte con el mercado laboral debilitándose. El presentador dice que no cree que ese recorte sea necesario, pero pide considerar para el primer trimestre del año siguiente los efectos de una ley de estímulo, los rezagos monetarios y la futura elección del presidente de la Fed por Trump.
▶ Tesis final, límites físicos y disciplina inversora
El presentador insta a no dejarse llevar por las advertencias de burbuja y sostiene que la IA y los márgenes seguirán siendo importantes, en un año de elecciones de medio mandato en el que la asequibilidad será un asunto central. Destaca una tesis de Craig Shapiro: la demanda de IA chocará con límites físicos, de modo que el control de terrenos, energía, agua y acceso a la red podría ser una ventaja; los hiperescaladores podrían pagar a industrias como aluminio, acero y química para que cedan sus asignaciones de electricidad. Dice que esa idea le parece valiosa y que usó ChatGPT para explorarla y obtener nombres y lugares para investigar. Vuelve a Duke y sostiene que las decisiones deben juzgarse por cómo se tomaron, no por si el resultado fue favorable. Recomienda basarse en hechos, estudiar la oportunidad y evitar el pánico; contempla aumentar la exposición durante caídas y reducirla cuando los precios se recuperen. Como ejemplo, recuerda que Micron pasó buena parte de comienzos de año por debajo de 100 y luego subió hasta 240 sin detenerse mucho; cree que podrían producirse movimientos parecidos en algunas acciones pequeñas el año siguiente. Cierra pidiendo investigar y aprovechar la turbulencia, y deseando una buena Acción de Gracias.
Transcripción completa
[00:30] most of the news was good. Uh but we'll go through things and uh see what's happening. Uh thinking in bets uh I'll be talking about Annie Duke on this episode and hopefully for the people who are looking for some kind of calm to set in uh as they watch this. Uh I want you to just think in bets. Uh capitulation signs, more bubble warnings from AI. The AI credit stress fears came out. I had to deal with those this week. Um the reality with AI uh Gemini 3 came out
[01:01] this week. I'll go through that. Try to remind people uh to stay calm and for investors who have benefited this year by every time there's an AI bubble fear uh which again I think I said it last week I'm going to say it. You're going to be climbing a wall of worry of AI for the next five years. uh Bitcoin crash. Uh I'll go through that uh painful uh thing for me and uh all decisions at the end of the day are always bets on the
[01:31] future. So let's start off this way. Uh this was a a post uh that got 625,000 views. S&P is down three and a half% from all-time highs. My portfolio is down 35% from all-time highs. That pretty much was about my thought process this time. I've I've uh I had a good year in terms of picking uh the right time back in the liberation time to put money to work. I have some great buys in Microns, some great buys in uh Nvidia.
[02:01] Uh lately've had some good uh Eli Liysies. But then on the other side, I have taken a bath uh on my Micro Strategy purchases which were done at the same time around Liberation Day plus some of the rally going and Bitcoin. And since uh a good portion of my net worth is in Bitcoin, uh it's been a uh it's been a a a bad week for my portfolio. And the reason I brought this guy up in terms of going through it, when you have 3,700
[02:31] followers and you put up a post that ends up going viral for 625,000, I think that sums up the way people are feeling right now. Thinking in bets. Uh, if you haven't read this, if you haven't purchased it for your kids yet, uh, I highly recommend it. Annie Duke is fantastic. Um, I had a, uh, a 90inute one-on-one conversation with her one day and we talked about so many different things. She's a brilliant lady. uh has a great mindset and because she's
[03:02] translated psychology and poker together uh this book I think for everyone to make it through the next five years I highly recommend thinking about it. Uh it's just a great way but it's really important for for kids. I've given it to um to many many people over the years. This is just a summary with from it. Um, just thinking in bets is a book about making smarter decisions when you don't have all the facts because in the real world you never do. It's really about the fact that every decision you make in life is a bet. Uh, and there you go.
[03:34] Every decision is a bet. Thinking in bets or decision tree. It really is just a great book for people to go through. But in particular, whenever you're letting your emotions run free because you think you know things that you don't, you have to get back to a process that's based on prob uh probabilities, uh and the outcome that'll come from it. So, um despite all the the fears that have gone on, uh and again, I I find it amazing. The S&P was down 1.8% for the week. Uh, and again, I I bring this up
[04:05] almost every week just to show you that since this week here in April, we've had such a massive move and every now and then we get a correction. And every time we get a correction like this, it's a bubble fear. It's a AI is going to collapse fear. It's a every single time. There's not one time that that doesn't occur. Uh, even if it's with the tariffs, it's because people believe we're in a bubble. the NDX little bit weaker down 3.13%. You can see that monthtodate when you
[04:35] add these three weeks in it's been a good size correction. So I don't want to minimize how much damage has been done. Uh but I'll put that uh a little bit more in perspective. Russell 2000 was only down 70 basis points for the week. So it outperformed uh and at one point on Friday it actually got back to unchanged for the week. uh nice close uh in terms of candlestick, but more importantly as an Aliot wave person, you have this day here which will you know is is is a day that people will talk about the fact
[05:06] that we gapped up uh a percent and a half and then we sold off and we're down a percent and a half and it's only happened four times in the last 30 years. Uh, this looks very corrective, but the reason I brought this chart up is with the S&P fall, we've only given up 20% of the gain that had gone on. We finally closed below the 50-day moving average, which had to happen at some point, and we had this nice reversal day today. I bring this up just because keep your head on straight if you're going to
[05:36] make money going forward. But I'm going to go through the reasons why you want to keep your head on straight now. But this looks like a corrective move off all-time highs. Uh in terms of the NASDAQ, we've done a 26% correction. So again, or 24% correction. So again, we haven't done anything. And to make it even look worse, here are the year-to-ate numbers across the globe. This isn't just a US thing. This isn't just an AI bubble thing. In fact, the US is underperforming.
[06:06] And to call this a bubble when the S&P is up 12 and the NASDAQ is up 15 while all these other markets are up a lot more and when you do it in currency terms uh in the dollar it's even more in terms of investing overseas. So fear and greed index the CNN one uh got down to six basically the same level it was in liberation day. The VIX climbed up to 24. This is the fiveday average of it. So this takes out just kind of the movements and we close right around the
[06:37] 24 level. So again, you know, except for liberation day, this is a big move in the VIX. This is the VIX vol 5day. We're actually at the highest level except for liberation daytime. Equity put call ratio for for single name options, highest level since then. put volume jumped to the second highest reading ever uh historically. It was a good time uh in terms of buying
[07:07] this stuff the last times that it's happened over these years. And I just want to remind people when you overlay the VIX with junk spreads, this is what I care about. Um if things were bad, junk spreads would be widening. Uh I always had this uh uh line I've used funding funding spreads or the funding markets lead the credit market which leads the equity market and right now we don't have any stress in the junk bond market. Yes, we have
[07:37] stress in some other parts of the credit side in terms of private credit, but it's nothing as of now. It's isolated to that area and I think that's where it's going to remain despite the brilliant work of many many people saying the contagion is coming and I'm going to hear my teenth recession call uh since the great financial crisis. The S&P here's the correction in terms of the draw down 5.6%. I mean, it's not even a blip on here. And yet again, if you take any 5% correction we've had and compare the
[08:07] amount of panic I heard this week, I think on a relative basis, I've never heard this much panic relative to the size of the draw down. NDX, I mean an 8% draw down. I mean, look how many we had in 2020 dur this is during bull market times. Forget the bare market times because this is when it stayed down. But all these others, we went right back up eventually uh very quickly. I think the same thing will happen this time. If the AI bubble burst, a recession could follow. Um, The Economist has had a
[08:37] really high hit ratio for marking bottoms. So, the fact that they pin this on their uh their X feed uh gives me good hope. Um, I don't understand why Michael Bur's getting all this uh press, but regardless, he posted a chart showing how stocks peak before capital spending did in past investment booms. Yes, it is correct. I'll go through some other things that peaked. Stocks are a leading indicator for the economy. So, I agree with the fact that if you want to make
[09:07] the argument that we're in a bubble and that stocks will peak first, I completely agree, but you have to look at some other things, too. I've already shown you one of them, which is junk spreads. Jeff Gunlock, um I I mean, this guy's been wrong all year, but regardless, um warns of garbage lending and private credit booms. He also said um AI was overblown. That's all I needed to hear to basically say why listen to someone when the tip of the spear is AI. If you say something
[09:37] like it's overblown, then just short the market and just continue because if you really believe that AI is overblown, that getting intelligence into everything is overblown, then yeah, if I were you, I would be focusing on this. So, for the first time in 20 years, AI bubble fields have fund managers saying companies are overdoing it. I bring this up because this is the percentage of investors that think companies are overinvesting. So they complain that there wasn't any
[10:07] investment and now they're overinvesting. This was during the post great financial crisis period when everyone was getting their balance sheets. This is why I talked about the fact that QE is basically artificial intelligence in terms of the fact that uh we're going to be getting so many profit margins from the fact of reducing expenses and there's operating leverage multiplier on it uh that I think people should focus on that instead of that. Well, going backwards off on the right one. Um,
[10:37] Coreweave CDS, this was brought up to me this week, building in a roughly 40 to 45% chance the company defaults over the next 5 years. So, let me say um that I don't have a problem with this at all. Uh, do I think this is too high? Yeah. But do I think there's a 30 to 35% chance the company will have defaults over the next 5 years? As someone who's as bullish as I am on AI, yes, this is part of the thing that I think people
[11:08] have to expand. Number one, we're in a K-shaped economy, which means I don't expect the credit situation, the job situation to get any better. That is part of the bullish outlook for me with stocks and profit margins. I also believe that there's a chance a good chance that at some point two years into this we will have some artificial intelligence driven solution to power andor efficiency in a big way. Now when I say two years the more we build out the power the higher the probability of
[11:38] actually reaching a point in AI where we can figure out some solutions. I think that's why making any bets longer than five years when it comes to innovation and when it comes to things related to the stock market, I think you're playing at risk. This is a belief I have with everything related to AI that it cannibalizes everything and that it is impossible to predict what the world's going to look like 3 years from now. So, I don't have a problem with the CDS being up there. It doesn't mean we're in a bubble. It just means the credit
[12:08] market is starting to get smart with the fact that they may not be able to justify all the debt they're taking. Same thing with Oracle CDS. Again, I bring this up because don't focus on this until we start seeing signs. So, do you believe Jeff Gunlock when he says it's overblown? Do you believe the bubble fears from the economist? Do you believe all of the bubble talk? Or do you believe Eric Schmidt? I I just go if you're going to do anything this weekend to just get your brain back into check in a in a way that you can understand what's coming our way, just watch the AI
[12:39] revolution is under hype with Eric Schmidt. And just remember, he's not the CEO of a company. He's not sitting there doing this. He's actually doing things for the government. He's heavily involved. He's in the know of everything going on and he's talking about this stuff. It is here. It is real. It is now. Gemini 3 comes out. Best model in the world for multimmodal understanding. It's amazing. Having used it now for the last two days. If you want to go get the details on it because you were too caught up reading about AI bubble fears
[13:10] and worrying about the losses in your portfolio. This round table discussion reviewing Gemini 3 was fantastic again from moonshots. Just to take out a full quote from Dave Blondon in this. I was over at MIT last night talking to a bunch of undergrads and I'm trying to tell them like look you don't know this but you know 40 years ago we started writing code as a species before I read the rest of this I just want to make sure you understand he's saying he went to MIT 40 years ago is when code started in 1992
[13:41] when Netscape basically went through its road show and was eventually launched the world changed Msei world went from being Japan heavy to USheavy and you ended up having a big shift that went on as the world and everything about it was gobbled up by concentration which we now see with the Mag 7 that was all built on code and I swear to God if you look at what happens today when you write code verse 40 years ago it's identical it's like a higher level language nothing's really changed all of a sudden you can build software by talking to the machine it's
[14:11] such a different world by starting today and moving forward this is like a different world starting today from the day that we lived in yesterday this is all about how he feels having used Gemini 3. And the reason I'm bringing that up is as you get negative and believe there's a bubble. This bubble has been building now for 40 years. Code has dominated what has worked. You are fading a 40-year trend of the winners being built on code and you're still looking at a world where the
[14:41] manufacturing side and all of this other side has become less relevant over time. Nvidia's Q3 revenues came out. These are more facts. This is not about AI bubble. This is not some old guy saying that AI is overblown. Nvidia Q3 revenues surged to a record 57 billion, up 62%. 57 billion in revenues. Here is the chart of their revenues. You're going to pick the top of that. If that was a falling chart, it would say
[15:12] don't catch a falling knife. This is a I've never seen a vertical chart like this go straight up. But people are trying to pick the top in it. Nvidia is cheap. I've done this before on this. I've done it on interviews. I'll say it again. There is going to be $5 trillion almost assuredly of buildout in data centers over the course of the next 5 years around the globe. Nvidia currently is gets about 35 to 50% of that buildout number. It's can even go up to 60%.
[15:44] Unless you believe they're going to lose their moat, which is almost impossible as I've talked about, they're going to get some percentage. And right now, we have an expectation out five years of their revenues being about 425 billion. They keep beating numbers. They keep raising numbers. Pay attention to Nvidia. When it starts to reach a point where it is expensive and where its PE is not compressing and where its margins are actually going down as opposed to being out, you can start to worry about things. Until then, just follow the facts. Here are more facts. Palunteer
[16:15] which basically said in the last earnings as as I highlighted that they are now benefiting from hundreds of companies that are using them for their adoption side up 63% revenue growth. Forget about the valuation and the fact that it got too far in front of it. The revenue growth is a fact. AMD's revenue growth is a fact. Meta, the companies you're worried about still going 26%. Revenue growth should be some multiple of nominal GDP. These are expenditures
[16:45] numbers yearover-year. Every single one of them is in some way tied. And here's the S&P 500 revenue growth. Just look at these related to it. These are facts. This is not a joke. S&P 500 profit margins made all-time highs because of what's going on with Nvidia. The rest of them AI is taking center stage in corporate America. Every single part of the S&P 500 is now mentioning AI. You will be out of business if you do not incorporate AI.
[17:15] AI will force a recession without having a recession. A recession is about cleansing out the weak, letting the strong to survive. That's what Amazon did against its competition back in the 2009 to 2019 period. You're going to see the same thing from public companies over any companies that can't compete with it. It's going to continue to put pressure on the labor market. You're going to continue to have a K-shaped economy. There's just no way to get around it. And the governments will be forced to continue to give transfer
[17:45] payments to the Kshape without a recession. DRAM prices, this is a fact. This is from September. So, as the bubble fears are growing, DRAM prices. So, every time you hear, oh, we we've got accounting issues with DRAMs. The price of DRAM keeps going higher. Here's DRAM, the old DRAMs, the ones that are basically should not be going up this fast. This is this overlaid with PMIs. I will still say PMIs are going higher next year because of the infrastructure
[18:16] needs for this buildout. If we were in a bubble, as Michael Bur said, the stock market does lead, but what also leads is junk spreads. Here is the dot bubble. LTCM starts there. We've got credit fears that go two years before we get it. In the housing problem, we had tons of issues going on before the stock market peaked. peaked in March of '08, which is even up higher. So, he's right, but junk spreads lead the stock market. What also leads
[18:46] is bank stocks. This is year-over-year bank stocks. Again, at the peak in 2000, we've got credit problems. We've got lending problems. That's a bubble. And guess what? The banks year-over-year, they're down about 20%. Same number they were down at the peak in stocks in ' 07. So, these are the peaks in stocks. This is where bank stocks were trading year-over-year. Right now they're up 13%. I mentioned this last week. So again, as people just throw this stuff out there. It's a bubble. It's a bubble. Get some facts. Here's the five-year
[19:17] move in NDX leading up into the dot bubble. This is normalized. This is 800%. So it starts at 100. We're at 917. That's five years. Here's where the current level is, 95%. So 10 times more or eight times more, whatever you want to go through last time. So when you're looking at a bubble, at least do it right. Capex spending on AI is masking economic weakness. I completely agree. I'll go
[19:50] through why capex spending is right now in my opinion masking significant economic weakness. No matter what people say about trying to figure out the math as to what it's doing, if AI capex wasn't happening, we weren't having the money going through and Nvidia wasn't getting the revenues coming through, I don't believe the economy would be growing at the level that it is. We've got GDP now up near 4%. The last quarter was up around 3%. We wouldn't be anywhere near there. We'd be closer to 1%. We'd be struggling getting through.
[20:21] It would be very difficult. Artificial intelligence is consuming capital faster than investors can recalibrate. Bank of America now sees global hypers scale spending blah blah blah blah blah. Now the reason I put this up if I agree with it that level of investment is extraordinary. At its peak the 5G telecom buildout consumed about 70% of operating cash flow. AI infrastructure is now approaching the same strength. I I I just can't emphasize how easy it is for you guys to go into chat GPT take a snip snapshot of that paste it in
[20:52] there and say give me your read on this. That's all I did in this. Why comparing the AI capex to the 5G telecom buildout is misleading. Telecom was a levered low margin crappy regulatory cap industry. I added my own uh adjective in there. Hyperscalers are the opposite. They're high margin, low leverage, enormous cash flow. This looks much more like the 1990 semiconductor capex boom. Do you remember how that exhed? No. Cloud buildout in the early AWS Azure era. Do you remember that
[21:23] ended? No. Railroads are electrification era industrial capex booms. It's not like telecom. Again, I I if you guys are going to try to lose money and just follow these people that aren't doing anything on this and are just using their history of I can tell this is a bubble because I'm a genius as opposed to what are the facts behind the case. Why am I not going to a Michelin star restaurant for dinner? I'm going to sit at home in AI and pay attention to things. You cannot sit there and actually know how or why AI is useful
[21:55] unless you have some idea as to possibility of cancer cures, possibility of replacing people with it and using in businesses, what's actually happening. But it is disturbing the signals in the economy. So I understand why people believe a recession is happening. This is the leading economic indicator. There has never been a time since the in the history since it's been created. So for people that are in their 80s right now saying AI is overblown, just go back over time. This is a time where
[22:26] historically we've been in a recession when the LEI the LEI has been negative now since 2022. Since we raised rates, we raised rates so fast everyone thought there'd be a recession. Well, technically there is a recession. Commercial real estate's in a recession. Housing is technically getting close to a recession. We've got no job games being created. We are in some sort of a recession the way you define it in the past, but it's not the world we live in now. PMIs, again, this is a 20month
[22:56] average. The 20-month average, we can't get out of our own way. We've always been in a recession when manufacturing has been this weak. Always until now. The job situation, I showed this before, earnings per share, S&P earnings per share just continues to go higher. While temp jobs, they're not hiring temp employees. This has never happened before. So we are in a recession. This is why I did in this uh podcast I did with Dean Kernut. We both come from the derivative
[23:27] area. Most of the podcasts I do these day have some relationship at this point to crypto uh just because those are the ones that have been asking. But if you want to go hear me talk in more of a macro perspective and talk about things that are going on on the derivative world and the equity world, you can go listen to Dean. I thought he did a great job in terms of bringing up questions and going through it and I went through it in there. Now, for the pain for me, as wrong as I've been this year, I thought Bitcoin Bitcoin would be four times where it closed on this. So, this is maybe the most wrong I've been ever
[23:59] in terms of forecasting things. I thought this year we'd also see oil prices go higher. Uh we didn't we didn't see any of that. In fact, oil prices are not only weak, they continue to march down towards 50. Uh, and I believe power power is going higher and I think oil will go higher next year. But these are the places where I've been wrong this year. And Bitcoin has now fallen 36% from the highs that it was less than seven weeks ago. Now, here are the draw downs on this. Uh, and again, you had
[24:29] close to 90, 80, 80, 78. We guess could go down here. I think we're doing another one of these. So again, as long as I've been in it, I doubled down this week and tripled down this week. I also texted Anthony Pompiano on Tuesday and told him I thought the lows were in before another 7,000 uh point fall. So catching this falling knife has become something that has not worked for me, but uh I like doing things when you've got
[25:00] this much fear and greed. So maybe we'll go down to 70, maybe we'll go down to 60. I don't know. I just don't think we're staying down long. That's fear and greed index down at levels below where we were in liberation day. Here is the DSI, the daily sentiment index for the futures low since it's been uh since uh Bitcoin had futures. RSI fell below 21. Here's what's happened afterwards if you want to play the numerology game and look for an area to go in. But more importantly, uh I did
[25:31] write a paper called the light of the end of the turbulence on April 8th of this year. that was about the stock market and about the fact that again I thought whenever it got like this where people were jumping out a window and thinking the government was actually going to let the stock market fall uh it turned up higher because I believe risk assets are going higher next year because I believe PMIs are going higher next year and because I believe tokenization is not understood by people and stable coins we're going through some sort of a uh a liquidation part in
[26:03] crypto and it does move on its own uh basis is again it also rallies faster than anything. So in this paper I go through that it feels the same to me as it did back on April 8th where I've got panic coming from everywhere spread throughout family members, friends, everything along the line in terms of what's gone on. I get it. It's been a big fall. I've been wrong on it. People can doubt it. They can go through it. But at least I'm putting my bet on the line and highlighting why I think it is about to go higher. Now, in terms of
[26:33] this fall relative to the S&P, uh, again, it's 33%, we've had bigger falls to put it in perspective, and it's still gone higher. Most importantly to me, it's only down 9% year-to date, and that's after back-to-back years of over 100%. So, regardless of everything going on, two good weeks in Bitcoin in terms of a short covering rally, it could be outperforming the S&P. That's how fast it moves and how quickly when sentiment gets this bad historically, you can go the other direction. I'll take it. One of the other things that continues to
[27:04] make me happy is I do believe that retail has gotten caught here. Um, a lot of good things and a lot of positive news in Bitcoin this year. You also had a lot of supply from the DATs and the companies that were out there uh issuing stuff. Uh, there was a lot of supply. Micro Strategy obviously brought a lot of supply. Micro Strategy. This is the chart of it versus my specame index. Again, when you have a correlation that's this high that really begins. So this fall here to me, this liquidation which came right after the
[27:34] Trump tariff situation, there was a massive liquidation. There's still an overhang. This consolidation here to me was OG. This was my silent IPO time. And then this was retail getting knocked out once we broke the range. And I think this has been a lot of selling that's gone on. I think there's an overhang from that period. But we also saw the spec names go again. The spec names are mainly on nuclear, they're in quantum, things like that. These are names that I don't buy into um that I think their retail was just trading momentum in.
[28:04] They've bailed out of them. Here's the chart of it again since past the October 5th period. Bitcoin falls first, then you get a bounce and then from that point, the white line is spec name basically moving tick for tick with it in terms of day for day. The other thing it's been correlated to, there's been a massive growth unwind that began around the same time as the Bitcoin high in Asian growth factor. Uh, Asia is very much into Bitcoin and you've had a straight line down in the Asian growth
[28:34] factor. Massive move. Uh, there's clearly an unwind going on over there. Here is the overlay since August in the Asian growth factor with Bitcoin. Circle has done a complete round trip. So, you've now taken it back to the opening price of the IPO. All right. You got to love to witness. This is the I highlighted this before. S&P opened at least one and a half% only to close one and a half% lower. There's only been three other times in more than 32 years.
[29:06] Um, Liberation Day, this is when I wrote the light at the end of turbulence. That ended up being a good time to buy. Uh this is October and of08. Both of those two. Uh that was a few weeks after Lehman Brothers and that was when we got to we're getting close to an original low which happened a couple weeks later before bounce before another low that happened. I'm going to bring this up again and this time I'm going to go through it. Annie Duke again thinking in bets.
[29:36] Life is more like poker than chess. You must make highstake decisions with incomplete information, hidden variables, and luck by framing every decision as a bet. You learn to assign probabilities, evaluate expected values, and challenge your own assumptions. Duke emphasizes constant updating, treating beliefs as hypothesis that evolve as new evidence arrives rather than the fixed truths. Ultimately, the book teaches that great decision makers aren't certain. They're disciplined, probabilistic, and adaptable in the face of uncertainty. You look for the scenarios where things are wrong. every
[30:06] single chart. If I wanted to get worried about the market right now, I would want to see this turn negative. Again, this is revisions. These green lines here are revisions going higher. Earnings revisions. They still remain high. These are periods, and you can go through them on your own where you want to be long stocks. A lot of these are coming out of coming out of the dot bubble, coming out of the great financial crisis, coming out of COVID. We're sitting up here. This is why I expect the PMIs to go higher. The PMI went higher in here. The PMI went
[30:38] higher in here. PMIs went higher in here. PMIs went higher in here. You can go look it up for yourself. Revisions are going higher. And that means a lot of good things. Here's the revision. Like I said, overlaid with PMIs. I've shown this chart before. This is the 20week moving average of revisions. Last one was 0.16. We are currently on a 20week at 0.15. It is still sitting up here, guys. revisions are going higher and the positive earnings news in the US has not been lost in the rest of the world. This is a
[31:08] global story global. It's actually going faster outside the US which is going to help the US as well. And to go back in time, this is the biggest relative earnings growth differential from the US rest of the world in the last decade. So owning international stocks, which I still like over US stocks, even though
[31:39] I'm bullish on a AI guys, see I can have a view that is focused more on else. I also like small caps more. And guess what? And h how can you be looking at this? Realizing consensus EPS growth for select small caps forecast point to nearly 60% growth. If we get anything close to 60% growth, do you think small caps are not going to have a good year? Do you not think the market's going to have a good year? Um especially since what they're
[32:09] coming out of negative. Negative. This again is what PMIs look like when you go from negative midcap, same thing. So for those of you in the midcap and small cap space, there are good things to come. You've been in a bare market for three years. If you get PMIs moving higher or if it's just because rates are moving lower and they're getting the benefit of the tailwind of rates moving lower, it's still good. Oversold breath. We are oversold, but in particular the
[32:39] NASDAQ on the breath. Oversold. I also highlighted this. Oh, didn't make this bigger. We're oversold on NYC breath. We're oversold on S&P breath. We got the Fed move down to less than 50% earlier in the week. And then an important news item came out. New York President John Williams sees room for central bank to cut interest rates again as the labor market softens. This is a big deal. Um, and again, I say it's a
[33:09] big deal. jumped the probability back to above 60. I don't think this is necessary, but I say this because remember in the first quarter of next year, you're not only going to have the one big beautiful bill positive side. You not only have the monetary lag, but you're going to have Trump pick his Fed chair. These are all news items to come. I want you to keep your head in the game. AI is strong. Don't fall for the trap of the bubble stuff. Don't fall for the trap as we're going through one of the most potent and unbelievable times where margins are going higher. This is
[33:41] going to remain a story and it's going into a midterm year where affordability is a major issue. Craig Shapiro put out a fantastic um post. Uh and I've used it in meetings this week. Uh I believe in it. It's he put it in a way that was great. I just want you guys to think about this. If you're focused on uh industrials, if you're focused on old commodity related names, AI's demand colliding with physical limits, the next moat is control of land power, water, and grid access. In a world where we
[34:11] need all electricity, the legacy industries use up a tremendous amount of electricity, aluminum, steel, all kinds of stuff. They've got he's got chemicals listed here. if we need it, the hyperscalers can pay them to go offline and use their electricity allocations. This is a great great call. I liked it so much I put it into chat GPT. It gave me a whole bunch
[34:41] of ideas on it. Again, as I do my videos coming up that'll be on the payw wall, you guys are going to hopefully be able to do this yourself and all of a sudden it comes with names. It comes with places to go look. But I would keep that in the back of your mind because I do think that is what's going to happen. I'm going to finish it off again with Annie. Uh, all decisions are bets on the future and they aren't right or wrong because they turn out well or poorly. They're right or wrong because of how we made them. I couldn't agree more. I want you guys to remember that as you go
[35:11] through this. It is about making decisions right now, not based on the fear. It's got to be based on facts. And even though we don't know the future, all decisions are bets on the future. You put all the information together, my job here every week is to try to give you guys a story that involves both the positive and negatives that are going on with AI to help you guys be strong in these points when people down because a lot of the alpha that's being made is during these times. Even if you just increase your waiting at this point and
[35:42] then reduce it when it goes back higher, we're going to have a lot of these moves. And just remember, Micron was under a 100red for the better part of the year at the beginning. Couldn't get above and then when it finally went, it went up to 240 without really stopping at any point. I think you're going to get similar moves next years on some of these small cap names. So, do your homework now. Don't panic. Use this opportunity and just remember the light at the end of the turbulence is the time to be focused. Have a great Thanksgiving, guys. I appreciate
[36:12] everything there. Subscribe if you haven't already. Give it to your friends, your family. Buy Annie's book. See you.