Running Hot Into Scarcity: Why Bottlenecks Are the Risk to the AI “Bubble”

[0:00] All right. Uh, back from Florida and uh
[0:05] getting ready to head up to uh to Maine
[0:09] for at least a week uh maybe longer and
[0:12] preparing for the uh move up there for
[0:15] the summer. This will be a different um
[0:18] video. Uh I I really do believe that uh
[0:24] the combination of some of the things
[0:26] we've been talking about for the last
[0:27] three weeks in terms of getting ready
[0:29] for the next regime shift, which would
[0:31] be about inflation and the pressure it
[0:34] would put on the Fed and yields possibly
[0:37] moving higher
[0:39] and the fact that now everyone has been
[0:41] forced in and we clearly have uh the
[0:45] next phase of parabas. And I want to
[0:48] remind people uh because I'm writing a
[0:51] paper on this.
[0:53] It seems like a distant memory, but last
[0:55] year the main parabas of the year uh
[0:58] were first palunteer.
[1:01] Uh you can go look at it having a
[1:04] similar move as micron has the last 18
[1:06] months. Uh you can go look at gold and
[1:12] when gold doubles in a year um that is
[1:15] volad adjusted uh uh whatever bagger you
[1:18] want and silver was a four or five
[1:20] bagger uh those things have been
[1:22] consolidating and in a world of bubbles
[1:25] parabas and speed crashes as I said
[1:28] these things shift quickly and there is
[1:30] absolutely no doubt that uh between the
[1:34] leverage ETFs the breath breaking down
[1:36] that the market is in a different
[1:38] position. And so for me personally, I've
[1:40] been scaling out of things. I've
[1:42] mentioned that I had taken twothirds of
[1:44] Micron off. I am effectively as far as
[1:47] I'm concerned um out of everything. Uh
[1:50] definitely out of all of Micron. And
[1:52] this is not a negative view. This is
[1:54] just and this is not me wanting to short
[1:56] this stuff. I will be looking to
[1:58] accumulate things as I've been doing in
[2:00] silver and as I've been doing in
[2:01] Bitcoin. those things are not at 52-
[2:03] week highs and they don't have any
[2:05] bubble signs to them right now. You're
[2:07] going to have rolling bubbles. But this
[2:09] agenda for today, hoarding bottlenecks
[2:11] in the next risk phase. U it's just too
[2:14] positive and in particular
[2:17] we're running hot into scarcity. Uh and
[2:20] that's where I want to start this off
[2:22] because uh I did write a paper this week
[2:27] on running hot into scarcity. And again,
[2:30] this is about the government that wanted
[2:31] to run it hot, but they're running it
[2:33] into scarcity, guys. And scarcity at the
[2:36] beginning of the year for me to be short
[2:38] abundance long scarcity, which has been
[2:40] the trade. It's not just the thematic
[2:42] portfolio that I have that's up
[2:44] significantly this year. The other side
[2:46] was to be short software. If you take
[2:48] the two of them combined, you're talking
[2:51] about a massive outperformance. And I
[2:54] think now after we're five months into
[2:57] the year, uh, the scarcity trade is
[3:00] becoming the dominant part. But scarcity
[3:02] brings bottlenecks. It brings shortages
[3:05] and it does bring changes in earnings
[3:08] and that's what I think is going to
[3:09] happen. So, let's go through the
[3:10] inflation data. CPI
[3:14] came in in line a 6 print, but again, it
[3:17] takes the year-over-year CPI up to 3.8%
[3:20] 8% and it continues to trend. I don't
[3:23] think the important part is that it's a
[3:25] 6. I think that aside from during COVID,
[3:27] it's the largest print basically since
[3:30] 2012.
[3:33] Yet PPI, which came in even hotter, and
[3:37] this is the final demand PPI. Um, I just
[3:41] think at this point people don't realize
[3:43] that this is not just an oil story. uh
[3:46] this is the scarcity and the hoarding
[3:49] that has gone on
[3:51] for the dramatic shift that we've seen
[3:53] in AI compute demand. I've talked about
[3:55] shortages. I've talked about the model
[3:57] changes. I have moved almost completely
[3:59] to chat GPT at this point. 5.5 is by far
[4:02] the best model. Codeex to me has
[4:04] replaced quad code. Um, I moved all of
[4:07] my open claw stuff at this point to
[4:09] codeex uh to 5.5 and I've seen just
[4:12] amazing abilities to it particularly
[4:14] while while I was in Florida. So I just
[4:17] think at this point the compute shortage
[4:18] and the hoarding and the leaprogging and
[4:20] all the things that are necessary and as
[4:22] I go through this you'll realize if you
[4:25] put together what has happened in the
[4:27] first quarter of this year this was a
[4:28] dramatic shift open 4.5
[4:32] open AI agentic world much quicker than
[4:35] anyone expected and it happens at the
[4:37] same time that the one big beautiful
[4:39] bill bonus capex depreciation was going
[4:41] on hoarding's going on leaprogging is
[4:43] going on the revenues went higher we
[4:45] have RPOS's all over the place. Guys, we
[4:49] have a lot of things built into the
[4:50] market um that can be delayed by not
[4:53] being able to build out the data
[4:55] centers, not being able to plug things
[4:56] in. And I think that's where we're going
[4:58] to be at. It's not a negative story, but
[5:00] you have to get used to the AI uh cycles
[5:03] compared to the traditional consumer
[5:05] versus manufacturing business cycles. AI
[5:07] cycles are constantly in this $90
[5:09] trillion buildout, which is going to
[5:11] happen. We don't have enough supply for
[5:14] this. So you're going to run into
[5:15] positions where there's an issue. Here's
[5:17] the import price inflation which
[5:20] significantly above these numbers have
[5:22] been tracking high even before the war.
[5:25] And that's the most important thing in
[5:26] here. We're going in at levels that are
[5:29] um absolutely much higher than what
[5:32] could have been expected. Here's the
[5:33] PPI. So the input cost side and this is
[5:36] overlaid with year-over-year CPI. So
[5:39] again,
[5:41] I don't see how this the only way this
[5:43] wouldn't happen is if this was a
[5:44] temporary spike. I think you're going to
[5:46] see headline inflation for CPI head up
[5:49] significantly. And again, these are
[5:51] overlaid on the exact same scale. We did
[5:54] go way above. So even right now, you're
[5:55] still talking about a five handle on CPI
[5:58] if that is all that happens. And I don't
[6:00] think that's all that happens. So as I
[6:02] talked about, we are back in a regime.
[6:05] Um there's no doubt that this is
[6:07] different.
[6:08] So since chat GPT
[6:11] uh or sorry yeah chat GPT was
[6:14] democratized and launched here is the
[6:16] beginning of 23. We've been in a
[6:18] situation where rates have been above uh
[6:21] CPI. Now we have CPI breaking above
[6:25] rates three-month bills. So we've got
[6:27] negative yields yields around the globe
[6:30] 30 year US 30 years JGB 30 years uh
[6:34] German uh
[6:37] 10 years and guilt 30 years. This is not
[6:42] just a US thing. We have rates around
[6:44] the globe following the inflation
[6:46] situation. There was no solution to
[6:48] Iran. So the straight is still shut.
[6:51] That is the major issue. And if it again
[6:53] you've you've we've talked about it
[6:55] before. Um, I think we're way past the
[6:58] point of ignoring uh the impact that
[7:01] it's going to have. It's obviously had
[7:02] an impact on oil prices, but I think
[7:05] we're now going to start to see that the
[7:08] longer this sits up here, the more that
[7:09] the inflation pressures go, the more it
[7:11] puts pressure on the central banks to
[7:13] respond and actually raise rates. All of
[7:15] these traditionally are very very big
[7:17] negatives for the market. But there's
[7:18] been one positive, well actually two
[7:20] positives that have uh superseded it.
[7:23] One is that the earnings growth which
[7:26] has been dominated by the AI sectors has
[7:28] been far better than expected but you
[7:31] have a momo chase happening and I think
[7:33] we're about the ending of that now at
[7:35] least of the bigger side. Uh supply
[7:38] chain stress that peaked in COVID heads
[7:40] higher again. This is the global supply
[7:42] chain pressure index and the World
[7:44] Bank's global supply chain stress index.
[7:46] They're reflecting fresh strains. You
[7:48] can see this in everything. The PPIs,
[7:50] the supplier delivery times. Uh we had a
[7:52] big ruling this week may not have
[7:54] reached most of your news but it is
[7:56] important. Um this was a unanimous and
[8:00] yes unanimous Supreme Court ruling to
[8:04] shield freight brokers or does not take
[8:07] away freight brokers are basically can
[8:09] be liable uh if they hire unsafe
[8:12] trucking companies. This could be an
[8:14] extinction event for 30 to 50% of all
[8:17] freight brokers. Uh again, the issue
[8:20] with this is uh we're going to have a
[8:24] supply issue
[8:26] uh within in there. An adverse outcome
[8:29] in this case could create supply chain
[8:31] inefficiencies and destroy hundreds of
[8:33] thousands of transport companies. Major
[8:35] industries players say this was in
[8:37] April. So the ruling came down and it
[8:39] was unanimous. It's impossible to
[8:40] imagine anything in the Supreme Court
[8:42] being unanimous. Uh which just means
[8:44] this is not going to be changed. Uh,
[8:48] spring drought out this week. Worst US
[8:51] spring drought on record since 1895.
[8:57] We got the crop report.
[9:00] The lowest level since
[9:04] these are all just bad situations. El
[9:07] Nino, you're going to hear more about it
[9:08] this year. Uh, one thing you can't do
[9:11] the straight horses, you can try to um
[9:14] change it. You can try to taco it. You
[9:17] can't do anything with weather. Um, this
[9:19] will be an issue based on history. For
[9:21] many countries around the globe, it
[9:23] usually leads to dryness and fires. The
[9:26] only reason I'm bringing it up in here
[9:27] is when you run Hermoose for as long as
[9:30] it is in inventories are drained, which
[9:32] is a fact. That is not a doomer uh call.
[9:35] That is a fact. We are seeing the
[9:36] inventories be drained. You just don't
[9:38] have a cushion for weather events the
[9:40] way that you would. And it's a global
[9:41] thing. Mosaic cretailing fertilizer
[9:44] production due to the sulfur shortage
[9:47] from Hormuz. Uh here are the Asian
[9:51] inflation numbers.
[9:55] Sulfuric acid is also important for
[9:58] critical minerals. We don't have
[10:00] anything going through. Plastic prices
[10:03] are pushing companies to the edge.
[10:08] We're going to be out of, believe it or
[10:10] not, motor oil in the next couple weeks.
[10:12] Exxon and Shell inform major suppliers,
[10:14] including Costco, to prepare for empty
[10:16] shelves.
[10:18] Oil changes will be an issue. Um,
[10:23] Ford CEO Jim Farley has been blunt about
[10:25] it. The data center boom is already
[10:27] mutating into an energy crisis. The US
[10:30] in his telling is in the second or third
[10:32] inning of taking this seriously. Ford
[10:35] CEO warns it's already a full-blown
[10:37] crisis and Goldman sees an AI bottleneck
[10:40] that can't be vibecoded away. This is
[10:42] all the constraints of the physical
[10:44] world.
[10:46] Semi analysis this week. A very obscure
[10:48] part of the semiconductor supply chain
[10:50] NAPA is potentially become a quiet
[10:52] constraint on AI chips. We've heard
[10:54] about Helium NAPA. The main point is,
[10:57] and this is the thing I just want to
[10:59] make sure, prochemicals and oil, which
[11:01] are all in inventory issues, and there's
[11:05] nothing flowing through the straight.
[11:08] If you've seen anything from the AI
[11:10] trade, if you've seen anything from my
[11:11] thematic portfolio, whether it's
[11:13] cooling, whether it's um advanced
[11:16] packaging now, power semis, uh the bloom
[11:20] energy, all of these different component
[11:22] pieces and powers and chemicals and
[11:24] semiconductors in the rack, servers,
[11:27] they all depend on something coming
[11:29] through the straight. All of them. It's
[11:31] in everything. And the problem is as
[11:33] time goes on, this becomes an issue.
[11:36] There was clearly hoarding going on in
[11:39] the first quarter and that is going to
[11:41] raise earnings because the capex side
[11:44] remember from an accounting basis you're
[11:46] not taking all of the expenses. So from
[11:49] an earnings accounting basis is very
[11:52] straightforward. You spend $100, it goes
[11:54] into revenue for those companies that
[11:56] are receiving it. If you're depreciating
[11:58] the expenses and you're able to do it
[12:01] because of the one big beautiful bill
[12:03] extending
[12:04] things and allowing them to frontload
[12:06] this, you end up with elevated earnings
[12:09] for now. The question is when the rate
[12:11] of change shifts, how will the market
[12:14] deal with it? Um, peak Q1 earnings. I
[12:17] put this together because I want to make
[12:19] sure that again you guys can sit there,
[12:21] you can go through it. We had insatiable
[12:24] demand. The agentic AI rise was not
[12:26] expected. The adoption realiz
[12:29] realization was not expected. So because
[12:31] we get agentic AI then all of a sudden
[12:33] we get a parabolic demand side. The
[12:35] parabolic demand means we need more
[12:37] chips, more everything, cooling, tubing,
[12:40] every optical fibers. So it sets off the
[12:43] strategic AI race. It coincides with the
[12:47] one big beautiful bill. So you get this
[12:49] massive amount of dollars that are
[12:51] flowing through. you get all the
[12:52] benefits and you get all these RPOS
[12:56] which mean that they have to go and
[12:58] they're competing with each other. So
[12:59] that leads to all of these different
[13:01] components that are ordered. The problem
[13:03] is there's been a lot I'm sure of over
[13:07] orderering, hoarding. And if it didn't
[13:09] start that way, every single week that
[13:12] went by, when you're getting the message
[13:13] saying there's nothing here, if you want
[13:15] it, buy it now. The pricing pressure
[13:16] goes high. And then you add in this new
[13:20] component. Remember the straight of
[13:22] Hormoose being shut down is an April
[13:25] event.
[13:27] So for everyone out there when there's
[13:29] one thing to be in a war where there's
[13:31] bombing thinking it could end, but now
[13:33] the straits there. This bottleneck is
[13:35] real. So I think it is a big mistake for
[13:38] people to not realize that there is a
[13:41] risk here now. And this is not a bearish
[13:43] thing. This is more of you have to be
[13:45] ready for speed crashes when you're in
[13:47] parabas. So, I did a uh a special video
[13:50] on Tuesday of this week to go through a
[13:53] growing set of warning signals that
[13:54] suggest the AI leaders may be entering
[13:56] its next phase of rising volatility or
[13:58] consolidation. What stands out is the
[14:00] weakness is no longer isolated and signs
[14:02] of contagion are showing up in the
[14:04] hottest market in the world, South
[14:05] Korea.
[14:07] So, I go through this whole thing and
[14:09] basically say I'd be looking for a
[14:10] potential momentum unwind. Now, the
[14:12] reason this was so important last week,
[14:14] I highlighted what I'm uploading every
[14:15] week, which I will upload again to the
[14:18] website for subscribers, which is this
[14:21] exhaustion model, which I put in to
[14:22] break everything down by extreme
[14:24] exhaustion, elevated exhaustion,
[14:26] somewhat exhausted,
[14:28] and low exhaustion. And to highlight how
[14:32] it worked on Tuesday, here are the
[14:36] movements
[14:37] linking everything together. Now, some
[14:39] of these extreme exhaustion are semis.
[14:41] In fact, I'll go back and show you. Some
[14:43] of them uh are optical and inter. So, I
[14:46] took all of the ones in the extreme
[14:47] exhaustion, all of the ones in the
[14:49] elevated, all of the ones there. I had
[14:51] open claw go out and just tell me what
[14:54] the moves were for both an average and a
[14:56] median for each of these. And what you
[14:58] can see for the average move 4.1, it's
[15:01] exactly there. So, the exhaustion
[15:03] model's been tested. If you guys want to
[15:05] go see the ones and go look for
[15:06] volatility for you option traders, this
[15:08] is the way I would use it is go see
[15:10] what's had the biggest move. Go through
[15:12] the list, find the names and see which
[15:14] implied VSS have not had the similar
[15:16] type move. These were average moves.
[15:18] This is the median uh except for this
[15:20] one point here. Everything again lined
[15:23] up. And when you're building analytics,
[15:25] that's the kind of stuff you want to
[15:26] look for. Now, here were the signs that
[15:27] I showed on Tuesday. The cost be 50. You
[15:30] had this. So when you have exhaustion
[15:32] signals, what I'm looking for
[15:33] candlestick patterns. Candlestick
[15:35] patterns basically show you visually the
[15:38] fight going on between supply and
[15:40] demand. So we had a day with a big
[15:42] reversal. Here is SKH. We had a day with
[15:45] a big reversal. I'm not going to show
[15:46] you the updated one, but let's just say
[15:49] the battle has continued inside there.
[15:53] If you use Micron, by the way, look at
[15:55] how many candlestick wicks we've had in
[15:57] the last eight days. Basically, seven of
[16:00] the last eight days have had large
[16:01] wicks. I don't care whether it's an
[16:03] upside wick, meaning a day that open on
[16:06] the lows and then closed on the high or
[16:07] open mid, sold off, and then rally back.
[16:10] We had a bunch of those. We also had
[16:11] ones where it opened high and sold off.
[16:13] When you have big wicks and
[16:14] candlesticks, that is a an exact
[16:16] definition about the buyers and sellers
[16:18] at that point fighting each other. Um,
[16:22] here is Soytech. I mean, this move has
[16:25] gone from 20 to 180, a ninebagger on
[16:28] something that I recommended to a few of
[16:30] you back when it was in the 30s. Um, I'm
[16:34] not stupid. These types of things don't
[16:36] happen. Nothing has changed enough for
[16:37] this to be a ninebagger in that short
[16:39] amount of time. So, it had a reversal
[16:42] day on the candlestick. So, I showed you
[16:44] Korea and I showed you France. Um, we
[16:47] saw one in Japan as well, which I didn't
[16:49] include in this. Um, and basically, and
[16:51] that's the Japan one shooting star.
[16:53] Here's the way I would upload the
[16:55] exhaustion model with whatever charts
[16:57] you have that look that way and have it
[16:59] go through. And basically, it gave you
[17:01] what the candlestick pattern is saying.
[17:04] And overall, the summary, the AI
[17:05] semiconductor momentum trade is still
[17:07] alive, but it is stretched. The next
[17:09] phase is likely more selective and more
[17:11] volatile. Um, so that was the point of
[17:14] it and I wanted to show that SKHENX has
[17:17] had such a big impact on all the AI
[17:20] trades. This is the overlay between my
[17:21] thematic portfolio
[17:24] wasn't as correlated or direct overlay
[17:26] as then. But in this parabolic move,
[17:28] basically there's been a clear move by
[17:31] people to get weighted into this. This
[17:34] was probably the most important part of
[17:36] the video. The thing that I think is the
[17:38] the most important for you to think
[17:40] about.
[17:41] So in turbulence models or any kind of
[17:44] indicator, I'm looking for things which
[17:46] there's a breaking correlation and
[17:48] volatility between things that are in
[17:50] there. If I could do a turbulence model
[17:52] right now of just the AI thematic
[17:54] portfolio, which I probably will do this
[17:57] week, the problem is we've seen now the
[18:01] oil sensitive things which were very
[18:03] correlated to SKHEX start to break. So
[18:06] SKHEX is this orange or this yellow line
[18:09] here. This is the Cosby machinery index
[18:14] which peaked and started to break down.
[18:17] The white one is the construction. When
[18:19] I start seeing things like that, that is
[18:21] like a warning signal that we're
[18:23] breaking off, that not everything is
[18:25] benefiting and that's going to be
[18:26] because of oil. This chart probably
[18:29] bothers me the most though. I wrote I
[18:32] highlighted a thing video on here about
[18:35] doing an edge device thing for someone
[18:39] and I did it back in early no it was
[18:42] even late March about power semis. This
[18:45] is a this is six names of power semis
[18:48] that I recommended this movement here in
[18:51] a matter of less than a month on names
[18:54] that haven't even really seen the demand
[18:56] yet. Most of this has been the fact that
[18:58] they are cheap but the demand hasn't
[19:01] come through yet because the power semi
[19:03] side which is going to be driven by a
[19:05] part partly uh edge devices but also
[19:08] partly the DC convert the DC
[19:11] 800DC
[19:13] uh movement for Nvidia which I
[19:16] highlighted was something they were
[19:17] looking for. A lot of these things are
[19:19] on the come guys. So when prices start
[19:21] moving ahead that fast, that definitely
[19:23] has the category of let's just ride the
[19:25] trend and go with momentum. And for
[19:28] people that have made money on it, it's
[19:29] fantastic. I think when you're trading
[19:31] parabas though, as I said, at some point
[19:34] here, you got to be ready for the turn.
[19:36] And I think we're increasing the
[19:38] probability of the turn for a lot of the
[19:40] reasons that I'm going through now. This
[19:43] bothered me, too. Uh it's Saturday
[19:45] morning here. I got in late last night.
[19:48] This is from Goldman Sachs um desk. Our
[19:51] trading desk. We have seen sovereign
[19:52] wealth funds and asset managers getting
[19:54] stopped into this tape as buyers since
[19:57] the calendar flipped to May. A clear
[19:58] momo chase which is highly concentrated
[20:00] in everything AI. Sell losers buy
[20:04] winners. This is the rotation that I
[20:06] wrote about guys. Um I was at an
[20:08] institutional investor. They're one of
[20:10] the preeminent uh publications
[20:12] especially for public funds and
[20:15] endowments and foundations. They
[20:17] published my art article which I spoke
[20:19] about at their event. Your capex is my
[20:21] opportunity. The benchmark arbitrage of
[20:22] the AI buildout. This is about being
[20:25] underweight all of those things. So the
[20:27] reason that there was a chase is highly
[20:29] because they've been underweight a lot
[20:31] of the things that are working and they
[20:32] needed to chase and get rid of the
[20:34] things that they're overweight. That is
[20:36] what the benchmark arbitrage that is
[20:38] what I am trying to get raas FAS to
[20:41] focus on. This will be a permanent part
[20:44] of the next decade. There will be
[20:45] temporary disruptions, but it will be
[20:48] permanent. Bubbles, parabas, and speed
[20:51] crashes, the end of human market
[20:52] structure. That was the title last week.
[20:54] Again, I think you have to start paying
[20:56] attention that momentum is going to be a
[20:58] part of our our life in this uh market
[21:00] because the numbers are going to be
[21:02] enormous, but I do think the rate of
[21:04] change from Q1 is going to be an impact.
[21:08] Um, we finished the week for the
[21:11] thematic portfolio down 46 basis points.
[21:14] Uh, I'm showing that not to say it was a
[21:16] good week for it. I'm showing it more to
[21:18] say we just finished explosive weeks.
[21:21] This is the first down week. I think
[21:23] there's a risk that this continues. Um,
[21:26] and I would be I already have, but I
[21:28] would be looking and be just one eye
[21:31] open. If you're momentum trading and
[21:32] you're going with stuff, shorting
[21:34] bubbles, uh, is a dangerous game. Uh,
[21:37] but I do think switching things into
[21:39] call options and being more involved,
[21:41] expecting that we're going to have more
[21:43] of an issue here. Um, I just want to
[21:46] highlight, I think you guys should read
[21:47] this. This is from April 8th. Gartner
[21:50] forecasts worldwide semiconductor
[21:51] revenue to exceed 1.3 trillion in 2026.
[21:54] Uh, which would be
[21:57] this is what it was last year. This what
[21:59] is this year. So there's two things
[22:00] about this. Number one, absolutely
[22:02] positively
[22:04] this is justified for what has happened.
[22:06] Now you can sit there and argue that
[22:08] next year it'll be 1 trillion and you
[22:11] think this is a bubble and that's what
[22:12] people are doing. I think that is
[22:14] foolish. I think absolutely it'll be
[22:16] higher next year because the capex
[22:18] numbers are what's driving this. I think
[22:20] what's interesting is they're saying 500
[22:22] so 60 plus% increase but then next year
[22:26] dropping off significantly. That is the
[22:28] rate of change thing that I think you
[22:29] have to be prepared for to see how
[22:31] markets will handle. Sometimes you
[22:33] extrapolate things that far and all it
[22:36] takes is a big change or bottlenecks
[22:38] because you can't sell things if they
[22:40] can't be made and that's where the issue
[22:42] is. It's a production risk not a price
[22:45] risk even though I do think prices are
[22:47] going to come down. So that means in Q1
[22:49] the biggest factor was heavy orders. So
[22:53] production numbers were volume numbers
[22:56] were up but at the same time prices were
[22:59] up dramatically. If you get a scenario
[23:01] now for the rest of the year where we're
[23:03] not able to meet the production numbers
[23:05] because of the impact from horm moves
[23:06] and the impact of buying everything in
[23:08] Q1, you're going to see volumes slow
[23:11] down and if prices actually start to
[23:15] normalize because volumes are slowing
[23:17] down then you end up getting the whammy
[23:19] of both of them. And that's what I think
[23:20] the risk is and it's the reason why I
[23:22] think at this point we've built in a lot
[23:24] of them. And when you have eight baggers
[23:25] and nine baggers as I shown in in a span
[23:28] of months that doesn't go on forever.
[23:31] Global semiconductor sales hit nearly
[23:33] 300 billion in Q1. They're on track.
[23:35] Volume is up 25% quarter over quarter
[23:37] and sales totaled 99 billion in March
[23:40] alone.
[23:42] Okay. So again I'm going to show you
[23:43] some charts as to why this is justified
[23:46] um for people. This is the imports of
[23:48] capital goods versus the imports of
[23:51] consumer goods. So think every single
[23:54] consumer name seems to be trading right
[23:56] now on a 52- week low related to
[23:57] household durables. I'm going to go
[23:58] through a lot of that. And everything
[24:01] related to the capital goods side is
[24:02] through the roof. Now this obviously
[24:04] does include semiconductors. So that's
[24:06] one of the big things. But you got to
[24:07] look this is back to 1992. This is out
[24:10] of nowhere. And that's why the earnings
[24:12] are justified in terms of the rotation
[24:13] as well.
[24:15] Here's the overlay with socks over the
[24:18] NDX. So you got the semis over the NDX
[24:20] following that trade.
[24:23] This is DRAM prices. And again, I'm
[24:26] showing this for two reasons. One is
[24:28] look at this here. I'll show more about
[24:30] this in a second. Um, but socks over the
[24:32] NDX makes sense. And so over the NDX at
[24:34] this point is mainly semiconductors
[24:37] relative to again software was a heav he
[24:40] heavier waiting. Uh, and also the
[24:42] hyperscalers, suspenders. So, it's
[24:44] justified for anyone saying this is a
[24:46] bubble. The bubble is the spending and
[24:48] the try to catch up. Uh there's two ways
[24:51] for bubbles to end. You can get the.com
[24:53] bubble, which I don't think is the case,
[24:55] or you can get it where we run into
[24:57] bottlenecks, which is what I think it
[24:59] is, and I think people are going to have
[25:00] to adjust to that. So, what I did was
[25:02] take the two charts that I just showed
[25:06] separately, the prices of DRAM and this.
[25:08] Can you match up the time periods in
[25:10] these charts and overlay them with them
[25:12] each other to see how they line up? make
[25:13] them one of them green, the other red.
[25:15] The first one is the six-month rate of
[25:16] change of DRAM prices, and the second is
[25:19] a chart of the socks relative to the
[25:21] NDX.
[25:23] Here's what we got.
[25:26] So, back to 2017, using Zcores, uh,
[25:30] they've tracked pretty well. So what
[25:32] that basically means is when DRAM prices
[25:36] start hitting the second derivative,
[25:39] meaning when they start to roll over.
[25:41] That's usually historically when stocks
[25:44] over the NDX has rolled over. It led
[25:47] here, it coincided here. It coincided
[25:50] here, they bottomed exactly at the same
[25:53] point. This one bottomed a little bit
[25:55] before. I just think this is important
[25:58] because now the DRAM prices the the rate
[26:00] of change has turned negative. So I
[26:03] would just be very wary about this in
[26:05] terms of semiconductors.
[26:07] Um and this is the way it goes. DRAM
[26:09] pricing momentum is fading but
[26:10] semiconductor relative performance is
[26:12] still rising. The risk is that once DRAM
[26:14] momentum rolls over the semi-trade
[26:16] becomes more sensitive to bad news
[26:18] because the market can no longer rely on
[26:19] accelerating memory prices to drive
[26:21] vision. This is why the bottlenecks
[26:22] matter so much. Again, any kind of bad
[26:25] news at this point becomes an issue
[26:27] because you have everyone in. And I
[26:29] think we're certainly going to have it.
[26:30] Remember, we've already ignored things
[26:31] like Turbo Quant, which was an issue. It
[26:34] helped drive uh Micron down from the
[26:37] high 400s to the low 300s in a short
[26:40] amount of time. We're going to get
[26:42] episodic moves at this point. Um, this
[26:45] came out in Korea, which I just want to
[26:47] make sure people understand whether or
[26:48] not this happens or not. In a world
[26:50] where we're taxing wealthy, an AI
[26:52] windfall tax, that is what started the
[26:55] whole thing in Korea
[26:57] to show how concentrated this is even in
[27:00] momentum. I don't know if I've ever seen
[27:02] this momentum has been isolated to the
[27:05] AI trade. So with inside other sectors,
[27:08] so this is basically healthcare,
[27:10] consumer discretionary, TMT, financials,
[27:13] industrials, energy and materials.
[27:17] This is industrials and TMT. So this is
[27:20] TMT. This is industrials.
[27:22] I I don't even know what to say other
[27:24] than this is all the AI trade the
[27:26] winners and losers. So this has really
[27:28] been concentrated into complete
[27:30] momentum. Now we hit the momentum on on
[27:33] the momentum factor on a weekly
[27:38] and a daily basis. Both of them hit 70.
[27:41] I'm sorry weekly and monthly. So, what I
[27:44] wanted to do is when you hit them both
[27:47] on this, what typically goes on
[27:52] and these are the periods of time that
[27:54] they both happen. So
[27:58] again, you don't get collapses. This one
[28:00] you did, this one you did later. For the
[28:03] most part, momentum when it gets this
[28:05] overbought, it starts to run into
[28:07] something and then you don't make money
[28:08] in momentum usually for another year.
[28:11] So, it wouldn't surprise me if we start
[28:14] running into issues. I think you just
[28:16] have to realize that this is very
[28:17] extended. The RSIs on both the monthly
[28:20] and the weekly are up there. And
[28:22] historically, when we've had this, it's
[28:23] gone. This is another way to look at it.
[28:27] Uh price momentum versus minal and we're
[28:30] at a five standard deviation overshoot.
[28:35] Here's another way. um semiconductors
[28:38] remarkably trading 62% above the 20 the
[28:41] 200 day moving average and he just gives
[28:43] the times in history. I don't even know
[28:45] how he's got that on this but good for
[28:48] him back to 1718. Um
[28:52] the reality is we're very extended on
[28:54] this in the short term. Um the reason
[28:57] this is an issue or not an issue in the
[28:59] long term guys, you could have called
[29:01] this a bubble from 2006 to 2020. This is
[29:06] Amazon over retail.
[29:09] Let's just take it from here to here.
[29:13] The issue is not semi doing this. I
[29:16] think this is going to continue. I think
[29:18] the issue is the time side that it did.
[29:20] The reason this one didn't get called a
[29:22] bubble was because it took five years
[29:25] for this to do what it did. But look at
[29:28] the numbers. These numbers are insane. I
[29:31] mean, you're talking about from 400 to
[29:33] here. So you are talking about a
[29:36] ninebagger. It just didn't happen in six
[29:40] weeks. So again, I remind you we're
[29:43] running hot into scarcity.
[29:45] This trade has been the call all year.
[29:48] Uh at some point you run into the issue
[29:51] here, which is bottlenecks and
[29:53] shortages. And that's the whole point of
[29:55] this video.
[29:57] And a reminder, you get the bubbles, you
[29:59] get the parabas, you also get speed
[30:01] crashes. That is the risk. And that is
[30:04] the thing you have to watch out for. Uh
[30:06] and remember people have been trying to
[30:08] call the top of IGV versus uh SMH versus
[30:11] IGV for some time now. In fact, all
[30:13] throughout most of here and we've just
[30:16] continued to go higher. But now again
[30:20] two and a half bagger in a short amount
[30:22] of time. Um I just wanted to highlight
[30:26] this too. This is Micron.
[30:29] So the reason I am out of uh Micron now
[30:32] is because I was lucky enough to buy
[30:35] from here to here. This is where my
[30:37] purchases were. I to be able to get any
[30:41] time in your life um you know a five to
[30:44] eight bag where I think the first part I
[30:46] sold out was somewhere in here and then
[30:49] by the time I was done I probably
[30:50] averaged in here and maybe it does go up
[30:52] higher and I miss it and maybe it just
[30:54] continues to go and I never get a chance
[30:55] to buy it again. That would not surprise
[30:57] me. But here are the Mag 7 x Nvidia. So
[31:01] this is uh well this is Amazon, Google,
[31:04] Apple, Meta and Microsoft
[31:07] and this is back to 2018. So just think
[31:09] about how much Micron has gone. It was
[31:11] the worst performer as of the first
[31:13] quarter of 2025 and now it is by far the
[31:17] best.
[31:19] Here are the winners over the last five
[31:22] years. If you guys have not seen this,
[31:24] so the S&P since December of 21, this is
[31:27] obviously before the rate hikes, was up
[31:30] 57 is up 57% over the four and a half
[31:34] years. I want you to look at the names.
[31:37] Every single one of these names, this is
[31:39] sorted by performance. Every one of them
[31:43] is an AI name except for
[31:47] a rocket ship name and a rocket ship
[31:50] name. every other one. So, if you've
[31:53] been avoiding this bubble over the last
[31:56] five years, I mean, this is a 37 bagger.
[32:00] I like I don't even know what to tell
[32:02] you in terms of how far these things
[32:04] have gone. You've obviously missed the
[32:06] move. So, this is why people don't want
[32:08] to be even talking about it in terms of
[32:10] it being a a real thing. But the reality
[32:12] is everyone benefited in this one. They
[32:14] didn't call this one a problem, Nvidia.
[32:17] But now you have all of these other
[32:19] names and there's no way they could
[32:20] possibly grow this fast. The issue is
[32:23] what is it going to look like a year
[32:24] from now? Is this going to still be
[32:26] higher than than than these numbers?
[32:28] Maybe it will, but the reality is at
[32:30] this point it's really hard to look at
[32:32] that performance and not realize that it
[32:34] is all part of the AI trade and it all
[32:36] started in 2021.
[32:39] This is from a code presentation. Um I
[32:42] think this is critical to look at. So
[32:44] this is the number of days with a one a
[32:47] plus or minus one and a half% move in
[32:48] the US momentum index.
[32:51] First of all the trend here since the
[32:54] end of 2020 is very clear. It's up now.
[32:58] We were mean reverting in 2021. Then we
[33:01] had the rate hikes. But now since then
[33:03] this has gone up. Just look at what the
[33:06] norm was in the world before. Especially
[33:10] in the QE world. We are in a period now
[33:13] where we are having a lot more momentum
[33:15] shifts plus and minus on a daily basis
[33:19] which means they can move violently in
[33:21] in in a much bigger way. This is the
[33:23] speed crashes and this is the unwinds
[33:26] and this is the stuff that goes and this
[33:27] is why when I do this I have a long-term
[33:30] thesis for the next 10 years but to
[33:32] maximize money during it I think there's
[33:34] going to be periods of time where you
[33:36] need to be very very cautious in the
[33:38] names that you have and think about
[33:39] things. Uh Jensen Yuang said this week,
[33:43] "Agentic AI requires 1,000 times more
[33:46] compute than generative AI."
[33:49] He's comparing the shift to the world
[33:50] suddenly needing a thousand times more
[33:53] cars. I've done that same thing for
[33:56] imagine if we had 8 billion more people
[33:59] in a year. We wouldn't have enough food.
[34:02] That's what he's talking about. And
[34:04] that's what's happening with digital
[34:06] agents. Digital agents don't eat food.
[34:09] They eat compute. Compute is chips and
[34:12] power. That's why these trades are going
[34:14] to continue to work. The only thing that
[34:16] can end that trade is total efficiency
[34:19] and algorithmic games gains which are
[34:22] going to take some time to get in there.
[34:23] But there will be at some point a
[34:26] massive discovery on how to do this
[34:28] without as much memory and everything
[34:29] along those lines. And when that
[34:30] happens, it will probably coincide with
[34:32] more supply coming in. So sometime over
[34:35] the course of the next year, I would
[34:36] imagine that that will be a story and
[34:39] then the earnings that have been brought
[34:41] up will come down so violently because
[34:44] you won't need nearly as much memory and
[34:47] the prices will come down and that's the
[34:48] diff that's the problem when you have
[34:50] earnings growing this fast and this is
[34:52] the risk. So when you look at earnings
[34:54] and everyone says well this is what it
[34:55] is. There is an assumption in 2028
[34:58] that all of these semi and and and
[35:01] memory names are going to be higher in
[35:03] terms of earnings than they were and a
[35:04] lot of them will be but I'm not sure all
[35:06] of them will and that's why when it gets
[35:08] very concentrated you have to look but
[35:09] this is the reason why people are
[35:11] capitulating because there's a
[35:12] fundamental reason for it. Um first
[35:15] quarter profits have surged 27%.
[35:18] The only time we've seen this is coming
[35:20] out of a recession. So I don't know if
[35:22] that's a good thing or not, but I do
[35:24] know that it is lumped up and that
[35:26] eventually you run into issues
[35:30] all 11 sectors. So it's broad positive
[35:33] earnings growth, but again when you look
[35:35] at this and you go through the numbers,
[35:37] it is very very isolated in terms of the
[35:39] high growth here. So these are positive,
[35:42] but not many are above 10. It's when you
[35:46] start getting into
[35:48] some of the things in here that you're
[35:50] getting them above. So, I would just
[35:52] keep track of that in terms of going
[35:54] through it. Here are the profit margins.
[35:55] Again, profit margins very easy to grow
[35:58] when you're able to extend price to some
[36:00] incredibly high number. The question is,
[36:02] is this going to continue or is this
[36:04] mainly about the hoarding that has gone
[36:07] on and the supply, the demand relative
[36:10] to supply that came out of nowhere? um
[36:13] revisions are still massively positive.
[36:15] The reason I wanted to show that is it
[36:17] is very hard to have a sustainable down
[36:19] move in the market without these red
[36:21] bars coming out of nowhere or a peak.
[36:24] This is what I would be more worried
[36:26] about is a peak in things that starts to
[36:28] go down and that a market that again
[36:31] because it had a parabolic move.
[36:34] This is what caused the fall and then
[36:36] this is what caused the rise. So these
[36:37] were the revisions move lower because of
[36:40] things like airlines and all things re
[36:43] resulting in oil. The problem is I do
[36:46] think you're probably going to have
[36:47] supply disruptions from hormos into the
[36:50] semiconductors that will impact the
[36:52] ability to produce. So again remember
[36:54] when you're selling something it's not
[36:57] just are prices and demand high. It's
[37:00] can you actually make the volume that's
[37:02] necessary? And at some point you're
[37:04] selling everything off the shelf and
[37:06] then you don't have anything to sell
[37:07] anymore. And if production can't keep
[37:09] up, it's an issue. But again, a data
[37:11] center and the ability of now going from
[37:13] a GPU, which was Nvidia, to where we are
[37:17] now, there's a reason why every single
[37:19] semiconductor has gone higher. And
[37:21] that's part of the packaging, the
[37:22] advanced packaging. That's what the
[37:23] basket is for. So you can get
[37:26] disruptions. And if one piece can't get
[37:28] done, makes it very difficult for the
[37:30] rest of it get done. So you have
[37:32] correlation to supply chain risk that I
[37:35] think is a real risk. And again, I'm
[37:37] talking about the riskreward changing
[37:39] because prices moved higher, earnings
[37:41] have gone up and the supply disruption
[37:43] and things that went on are happening.
[37:45] They're real. Hormuz has not been fixed.
[37:48] So if that's the case, I think you have
[37:49] to be worried that potentially there
[37:51] could be a shift. Here's estimate
[37:53] revisions. uh the estimate revision
[37:55] factor to the S&P. So again, justified
[37:58] for the S&P to go higher. Revisions have
[38:00] gone higher. If revisions start to move
[38:02] lower, that's where the issue will be.
[38:06] One other reason to not even listen to
[38:09] people saying this is a bubble. Um this
[38:12] is from this week. Since the start of
[38:14] 2026, the S&P has risen by 8%. Consensus
[38:18] forward estimates have risen by 13,
[38:20] meaning the multiple has gone down. In
[38:22] bubbles, multiples go higher.
[38:26] Nothing on the turbulent side. And I've
[38:28] been asked a lot of questions about
[38:29] this. The reason there's nothing on the
[38:30] turbulent side is because this has all
[38:33] been one trade. Rates had not moved
[38:35] until Friday. Oil had moved higher and
[38:38] that's what this line here was about oil
[38:40] and the yen. But overall, because this
[38:43] has got a lot of macro factors, we just
[38:45] haven't seen anything. I would imagine
[38:46] we will, but where we are seeing things
[38:49] is in this. So the S&P has had continued
[38:52] to extend higher. This is the S&P
[38:54] relative to the VIX meaning and I showed
[38:56] this on the weekly update for the
[38:58] subscribers which is when you start
[39:00] seeing this you're getting divergences.
[39:02] So the S&P is way above the highs here.
[39:04] What has not is the VIX is much higher
[39:06] than it was back then. Here's a way to
[39:09] look at it. Here's the S&P inverted
[39:12] going down. And here's the VIX for the
[39:15] sixth contract which went higher on
[39:18] this had been over the course of the
[39:20] last six months. So this rally here the
[39:23] VIX has stayed higher meaning you can
[39:25] take it as there's a lot of people that
[39:26] are still buying hedges but sixth
[39:28] contract takes you out quite a way and
[39:31] that says to me that
[39:33] the risk is growing. Uh Charlie
[39:36] McGillicut
[39:38] notes that while there has been relative
[39:40] calm at the VIX surface, the action
[39:41] underneath has been anything but with a
[39:43] warning that many others have mentioned.
[39:45] These sorts of things tend to collapse
[39:47] under their own weight. He's got a
[39:49] variety of different components in here.
[39:51] SMH at three-month at money V like all
[39:54] of these things are saying that the V
[39:55] market despite being quiet an issue.
[39:58] Breath is bad too. the S&P 500 breath.
[40:03] So, not the broader market, and I like
[40:05] to use the S&P breath as as a warning
[40:07] signal because it typically does peak
[40:10] before the market has a fall. Um, it did
[40:13] it here. It did it here. It started to
[40:16] break down here at the same time. Well,
[40:18] it's broken down and it actually broke
[40:20] the 50-day moving average, but if you
[40:22] look at it, it's basically unchanged
[40:25] since the end of January.
[40:28] We got Hindenberg Omens. Um, again,
[40:30] Hindenberg Omens, this is one of those
[40:32] signals that I always love to know about
[40:34] and I'll always show it when they
[40:35] happen. And the reason is because that
[40:39] means you have a lot of new highs at the
[40:40] same time. It is not a lot of new lows,
[40:42] but the breath is breaking down at the
[40:44] same time. So, a lot of new highs, a lot
[40:46] of new lows, breath is breaking down,
[40:48] and then you need the market to be going
[40:50] higher. So, it was triggered for the
[40:52] NYSC and the NASDAQ at the same time.
[40:55] This has happened only 19 times
[40:56] historically. The S&P win rate over the
[40:59] next five months is below 50%. Two of
[41:01] those signals happen before crashes. I
[41:03] don't care about that part, but I do
[41:04] care that we're seeing these. Uh the
[41:07] only precedent for when the S&P 500 has
[41:09] been at record highs while fewer than
[41:11] 60% of stocks were above their 50 and
[41:13] 200 day moving averages.
[41:15] Again,
[41:17] you start getting into uh issues. We
[41:19] still have the financials not only below
[41:21] the 200 day with the 200 day pointing
[41:23] down, but it is headed down again. And I
[41:24] told you this has never been a good
[41:26] thing ever in markets. Here's another
[41:29] one. This is the equal weight consumer
[41:31] discretionary versus equal weight
[41:33] consumer staples. This one historically
[41:36] is a leading indicator for correction
[41:40] correction correction. Um it typically
[41:44] does not happen. But look at this
[41:46] divergence. I've never seen anything
[41:47] like this. This is basically saying that
[41:49] the consumer is in trouble. Equal
[41:52] weight. This is not cap weight. So, this
[41:54] is not Amazon and Tesla in there. This
[41:56] is equal weight relative to consumer
[41:58] staples. We are now at the lowest level
[42:00] since basically around liberation day.
[42:03] Um, again, you look for these warnings.
[42:06] If you haven't seen this, here are the
[42:07] household durable names year to date.
[42:10] There's 26 of them. Four of them are up
[42:13] on the year.
[42:15] Here's the restaurants. Look at some of
[42:18] these names are down. Door Dash is down
[42:20] 34%. wing stop down 50. Um, here is the
[42:24] bottom end of the retail part. All of
[42:28] these names, I mean, these are just
[42:29] huge, huge, huge moves.
[42:34] Breath can be bad for a long time. But
[42:37] I've said I believe we're in the 1970s,
[42:39] which means we're going to see up down
[42:41] up down up down. But over the course of
[42:43] the next decade from the disruption of
[42:44] AI, I believe it's going to be very
[42:47] difficult particularly since the capex
[42:49] names in my opinion the spenders are
[42:51] going to continue to see multiple
[42:52] compression. This is part of the code
[42:55] report. I love the code reports. I think
[42:56] they have a lot of good charts in them.
[42:58] Uh and again fighting right now between
[43:02] the sellers which are the ones selling
[43:04] the tokens and the buyer, the Amazons,
[43:06] the Metas, the Microsofts, the Googles,
[43:07] the hyperscalers. My hyperscaler chart.
[43:10] Um the sellers again the new leaders
[43:13] cash flow today not in the future.
[43:16] That's the issue. These guys are getting
[43:17] tons of cash flow now. These guys are
[43:19] spending and it's not hurting their
[43:22] earnings because they get to depreciate
[43:24] those capexes. So you get this again
[43:27] earnings boom which isn't going away by
[43:29] the way. So significant pricing power
[43:32] all-time high margins.
[43:36] Here are the numbers. You're going free
[43:38] cash flow. So, capex, free cash flow,
[43:41] and it is penalizing them. Their
[43:43] multiples are going down slowly here,
[43:45] but they are going down. And here's the
[43:48] deal. Um, year-to- date return, and
[43:51] again, this was obviously a while ago
[43:52] before these were there, but you get
[43:55] 107% in these names, a doubling, and
[43:58] it's coming at the expense of these
[43:59] guys. And the difference in market cap
[44:01] does it here is now Microsoft having
[44:03] capex equal to 37% of its revenues. Now
[44:06] they now they have a ton of revenues to
[44:10] come seen through the RPOS and I think
[44:13] it's 700 billionish or something. So uh
[44:16] again they could get the revenues in but
[44:18] if the bottlenecks take too long they
[44:21] won't be able to realize them. Um here's
[44:24] what has taken the market higher in my
[44:26] opinion. The market would not have gone
[44:27] higher. The semis would not have worked.
[44:29] Momentum and retail would not have been
[44:30] involved to the degree it is if it
[44:33] wasn't for this. This here is the
[44:35] hyperscalers. So they bounced, the
[44:37] spenders bounced. The reason I have this
[44:39] up is this is a GARP software basket for
[44:43] Goldman Sachs. This typically is
[44:47] extremely correlated to it. Now it's
[44:50] broken apart. And so this is sitting
[44:52] down here. These rallied. Um I think the
[44:55] hypers scalers are the thing to watch to
[44:57] see when this is going to break. Here's
[44:59] the chart of the hyperscalers relative
[45:01] to the S&P. So they went up with the
[45:03] S&P. That's why I'm saying they still
[45:05] didn't do well. They did well from here.
[45:07] They're still down year to date relative
[45:09] to the S&P. I'd be watching that chart.
[45:12] Um, I wanted to now add Nvidia to this
[45:15] just to show that Micron is here, but
[45:17] since 2018, Nvidia is still the the the
[45:21] winner and king. Uh, I added now also
[45:25] Bitcoin. And I added Bitcoin because
[45:27] eventually in the course of the next few
[45:29] years, I think Bitcoin will be the one
[45:31] that is above them. And if I take it
[45:32] back further, it obviously is. But
[45:35] that's the story right now on that
[45:36] framework. Now,
[45:39] two-year yields,
[45:41] broke out,
[45:44] three-month bills. This is one of the
[45:48] differences I want to make sure you
[45:49] realize as you're thinking about where
[45:50] to rotate. I do not think uh we're going
[45:53] to go into a crash mode. I do think uh
[45:56] we're due for a correction. How far that
[45:58] goes, we'll just play it week by week
[46:00] and see where it goes. if it happens. Um
[46:03] the difference is this time and when
[46:04] people talk about what's going on with
[46:07] inflation
[46:08] back in 2022, uh this is the yield curve
[46:12] in three-month bills versus 10ear rates.
[46:14] So we have a steepening happening right
[46:16] now and this is the issue that comes in
[46:18] is the government I we we we don't have
[46:21] the same situation as we had back here.
[46:24] So remember back here the curve
[46:26] flattened dramatically. It was massively
[46:28] negative for stocks.
[46:30] This is the BCOM index. So this is
[46:33] commodities. It was hit in 2022 a as we
[46:38] sto saw the yield curve flatten out
[46:42] and this is the component of seeing it.
[46:44] So as the yield curve went down, you saw
[46:46] commodities go down. We're not raising
[46:48] short-term rates the way we did. Um and
[46:51] that's where I believe that this is
[46:53] important for things like the straight
[46:54] of hormuz. Uh copper prices are moving
[46:58] higher. Gold started to move or silver
[47:01] started to move higher. That's where
[47:03] I've rotated a lot of the micron money.
[47:05] Um it came right back down. This thing
[47:08] peaked here. It's taking a while. This
[47:10] is the way the next phase takes. I
[47:12] wouldn't be surprised if we spend more
[47:13] time in here, but eventually as we get
[47:15] through this when the government decides
[47:16] that it is not going to be raising
[47:18] rates, whether it's a dovish meeting or
[47:20] whatever, um I think you're going to see
[47:22] this go higher. Gold, same thing. But
[47:24] while the rate fears are top of mind,
[47:26] that becomes an issue. We did see one
[47:28] break and then it came back and what I
[47:30] wanted to say is crude was going higher
[47:32] at the same time gold and silver that
[47:33] was a break in this negative correlation
[47:35] and then as we saw rates break higher.
[47:38] So rates are the critical components why
[47:40] you want to pay attention to the Fed.
[47:41] What makes it very difficult to compare
[47:44] what happened back here? Here is CPI
[47:46] year-over-year.
[47:48] This is wages. Wages were going up with
[47:52] CPI. This is the critical component for
[47:54] the central banks which is the reason
[47:55] why you want to be long commodities.
[47:57] This is not going to stop things. We're
[47:58] going to have shortages. You want to be
[48:00] looking to buy commodities into weakness
[48:01] because they are not going to be able to
[48:03] use short-term rates to go because look
[48:05] at the wage pressure. We got the lowest
[48:07] wage print yet in median Atlanta Fed
[48:09] wage. They were correlated before. They
[48:12] are not going to be correlated right
[48:14] now. Here's the real wage component.
[48:16] This is a big move. So, you now have CPI
[48:19] year-over-year above real wages. This is
[48:21] going to be a midterm issue. You don't
[48:23] get this too often. This is what feeds
[48:25] into why Bitcoin is the endgame. Here's
[48:28] negative real yields. So, be patient
[48:30] with everything, guys. This is the
[48:32] regime that we're going to be in for the
[48:33] rest of the year. It may take time for
[48:36] the technicals to break out, but
[48:38] Bitcoin, those are its numbers and
[48:40] times. We got close to this moving
[48:43] average and we failed again on Friday as
[48:45] the market headed lower. If the market
[48:47] heads lower, you're going to have it
[48:48] again. I want to do a comparison now.
[48:50] Just remember, so this is building
[48:53] something and it's going to sit here. I
[48:54] think it's going to break higher. And
[48:57] part of the reason I'm being patient is
[48:59] because we had this. So think of this as
[49:01] Bitcoin. Maybe we need to have one more
[49:02] move lower in Bitcoin. I don't know. But
[49:04] this was Micron. And like I said during
[49:06] this period in here, nobody wanted to
[49:09] touch it. It was a $100 stock. Nobody
[49:12] wanted to be involved in it. I am going
[49:14] to be watching Dogecoin now. XDG and
[49:17] Bloomberg, guys. Um, this is the retail
[49:19] fire. So, when this breaks out above
[49:22] this level, that's when I believe that
[49:24] you're going to have retail involved
[49:25] with crypto. And I think it needs the
[49:27] energy. It's going to look for its next
[49:29] parabola. And that's where I think over
[49:31] the course of the next year, the
[49:32] parabola will start. It took a while for
[49:34] Micron to become a retail name. It
[49:36] basically took a year, so it might take
[49:37] a year for a lot of this stuff to go,
[49:39] but Dogecoin will be the signal to me.
[49:42] Um, you should read the Coinbase
[49:44] earnings report just to see the reason
[49:46] why the digital economy is going to be
[49:49] the next par parabola. I believe that we
[49:52] are starting the next parabola. I don't
[49:54] think it's going to be the AI trade. I
[49:56] think we're going to see these things
[49:57] consolidate. There'll be es and flows
[49:59] within the different names that are
[50:01] going to do well. They'll do better than
[50:02] the S&P, but I think as a whole, the
[50:06] thematic part of AI is going to have to
[50:08] consolidate some of the moves. And
[50:09] you'll have some names go up, some names
[50:11] go down. But crypto to me is going to be
[50:13] the next phase that you are going to be
[50:15] forced to be buying into. And that's
[50:17] because we have parabas in here too.
[50:20] Here's tokenization. Here's stablecoin.
[50:22] The parabas are happening here just like
[50:26] anthropic. And it is the very early
[50:27] stages. So you will have to get involved
[50:29] and it's all coinciding with AI agents.
[50:33] Introducing Amazon Bedrock agent core
[50:35] payments with Coinbase and Stripe.
[50:38] Everything is happening. We got Clarity
[50:40] Act movement. We're up to 72% before the
[50:44] end of the year, which I view as a big
[50:45] positive. You've got the UK relaxing
[50:48] stable coin restrictions.
[50:50] And again, here's Micron. Here's Nvidia.
[50:54] This is over since 2018.
[50:57] Bitcoin is sitting here. I believe the
[50:59] next phase it will look like a parabola
[51:01] as well. And here's what Bitcoin has
[51:04] done over the course of the last 15
[51:06] years. The next parabola in Bitcoin will
[51:08] come with an ETF. It will come with
[51:10] sponsorship from the government. It will
[51:12] come from things that you can do
[51:13] discounted cash flows on like Ethereum
[51:15] and Circle and all of these different
[51:16] components. So, the narrative will
[51:18] shift. Software will still be under
[51:20] attack. That's it for this week, guys.
[51:23] I'll see you next week.