Ceasefire, Inflation, Mythos and a Scarcity vs Abundance Moonshot — Jordi Visser (12 abril 2026)

[0:00] All right.
[0:00] Up in Maine this week.
[0:02] Uh
[0:03] doing this just before the close on
[0:05] Friday. So, if some of the prices don't
[0:07] seem to match up with the closing
[0:08] prices,
[0:09] uh just because uh
[0:11] I'm doing this so I can get out of Maine
[0:13] and get back to Brooklyn. Uh lot to go
[0:17] through. It's amazing how many things
[0:19] happen in a week.
[0:21] Uh obviously,
[0:23] this is uh definitely taco week, but I'm
[0:26] going to go through and and just again,
[0:30] get everyone to stay away from the S&P
[0:33] 500 as a gauge. Uh 95% of things that I
[0:37] listen to and watch in podcast related
[0:39] to the market, it's mainly about whether
[0:41] the market's going to go higher or
[0:42] lower. And I think at the end of this
[0:44] year, you're going to destroy yourself
[0:47] playing that game. Uh
[0:50] I'm going to stick with the same theory,
[0:51] which is that this is going to feel like
[0:53] the 1970s.
[0:55] Uh we're going to have inflation, which
[0:57] is why the there'll be a big part of
[0:58] inflation this week. Uh how you invest
[1:01] in inflation is the opposite of the way
[1:03] uh people have been. This is a scarcity
[1:05] trade. The abundance trade got destroyed
[1:08] again this week. We'll go through that.
[1:10] Uh but, you're going to start hearing me
[1:12] talk a lot more about Bitcoin. I think
[1:14] Bitcoin is entering a period where every
[1:16] investor is this is the beginning of
[1:18] your journey of understanding it. Uh
[1:20] growth as an asset in terms of stocks
[1:22] has just been destroyed. And I think
[1:25] this is going to continue.
[1:27] Uh the terminal
[1:29] value story is only going to worsen.
[1:32] We're going to go through that with
[1:33] Mythos. I wrote a piece on this. Uh I've
[1:36] I've done a lot of writing over the last
[1:38] couple weeks. Thanks to all the
[1:39] subscribers reaching out looking for
[1:41] stuff. I'm going to go through my
[1:42] thematic basket today in terms of
[1:44] just some of the things that I've talked
[1:47] about. Uh there's been some big moves on
[1:49] the upside. There's always going to be
[1:51] some losers in there, too.
[1:53] Uh we'll talk about Brazil, which has
[1:55] been one of the winners. And we'll go
[1:56] through silver, which I did. And also,
[1:58] I'll start to
[2:00] bring in some new things. Uh
[2:02] all right. I I'm just going to go
[2:04] through these first three images cuz
[2:06] I've been showing this since I think the
[2:08] first time I did it was somewhere in
[2:10] October, November.
[2:11] Uh this was all related to Opus 4.5, the
[2:15] acceleration, the destruction of
[2:17] software,
[2:19] uh and the need to have a portfolio
[2:21] which was
[2:22] based on rising PMIs, rising commodity
[2:25] bull market, and anything that was
[2:28] scarce with compute being completely
[2:31] insatiable from the demand side.
[2:34] The physical world upgrade, which was
[2:35] the theme for this year. Again, moving
[2:37] out of anything built on code, moving
[2:40] into anything AI in the real world as we
[2:43] shove artificial intelligence into
[2:45] everything.
[2:46] This is the first inning of this. So,
[2:49] for those of you still trying to short
[2:50] SMH, still trying to follow Michael
[2:53] Burry shorting Nvidia, uh anything on
[2:57] the software side can be shorted. Even
[2:59] Palantir, which is a name that I still
[3:01] like, uh but fully understand that
[3:04] no matter what you like in software
[3:06] right now, uh it doesn't really matter.
[3:09] AI is really accelerating at a pace that
[3:12] it makes any certainty on the future
[3:14] impossible for commodities. And for
[3:16] semiconductors or any kind of hardware,
[3:18] that is not the case.
[3:20] And again, the speed of this is the
[3:22] damage. So, I've shown the supersonic
[3:24] tsunami.
[3:25] Uh these are all photos of what's going
[3:27] on. And that's why when you read stuff
[3:29] like this is a big week for stocks. All
[3:31] right, the S&P 500 finished slightly
[3:33] down for the year as of this week. Uh it
[3:35] was up at the beginning, then it was
[3:37] down. Now, it's back to unchanged. And
[3:39] within the sectors, you've got a lot of
[3:41] things up and down.
[3:43] So again, we were down five weeks in a
[3:45] row.
[3:46] We got a bounce.
[3:48] I don't think at this point uh you
[3:50] should be focused on it being up. I
[3:51] think we're going to see a lot of green
[3:53] and a lot of red. I don't think this is
[3:55] going to be a repeat of last year. Like
[3:57] I said, this was liberation uh week day
[4:00] week. And then most of them the weeks
[4:03] were up. I do not think we're going to
[4:04] see this. I think we're you're going to
[4:06] see a pendulum back and forth, and I'll
[4:07] go through the reasons why. Qs after
[4:09] five weeks in a row down, made back most
[4:12] of it up 8% over two weeks. IWM, same
[4:16] thing.
[4:19] Here we are year-to-date. Uh
[4:23] energy, materials, industrials,
[4:25] utilities.
[4:26] This is basically your
[4:29] scarcity trade.
[4:31] And here's your abundance trade down
[4:34] here.
[4:34] Uh financials are down here because
[4:37] again, they are for the most part
[4:39] something disrupted by code, but also
[4:44] they've got a lot of connection in the
[4:45] debt market to the software side, which
[4:48] obviously has been the private credit
[4:49] side.
[4:50] Uh here's the best way to look at it.
[4:52] Here's the bounce in the S&P this week.
[4:55] Taco time. We get a big bounce. The
[4:58] white line here,
[4:59] IGV making new lows. Not only making new
[5:03] lows, making sharp new lows. And at the
[5:04] same time, private equity continues
[5:06] there. No bounce in the deflationary
[5:09] side of the market. Zero. This is why
[5:11] this is not last year. This was not a
[5:13] rally in getting rid of the problems.
[5:16] And the reason this line is here is this
[5:18] line
[5:20] was when the war started.
[5:22] So, the S&P rallied almost back to where
[5:23] it was.
[5:24] This stuff not only didn't budge, it's
[5:26] lower than it was at the start of the
[5:27] war. It had fallen before the war. Do
[5:30] not forget this cuz this story is still
[5:32] playing out, and it does not have a
[5:35] floor.
[5:36] So, I'm going to say it again. I've said
[5:37] it in software all year. You can try to
[5:39] pick bottoms in something, you are
[5:40] wasting your time. It's like playing
[5:42] trying to pick directions in the S&P
[5:44] this year. You've got better things to
[5:46] do with your money than worry about
[5:47] whether the bearish or bullish sentiment
[5:49] is high.
[5:50] You've got a job to do, and the job to
[5:52] do is to follow inflation on the upside
[5:53] in the shortage theme. Here's the S&P
[5:56] chart. What I want to show you is this
[5:57] is the way people are going to look at
[6:00] This was last year. Now, this day here,
[6:03] we had a gap back above the 200-day
[6:05] moving average. Very similar to what we
[6:07] did here. So, here is Monday
[6:10] in terms of the reversal of Trump uh not
[6:14] obliterating a nation.
[6:16] Uh we get the gap higher. You've got a
[6:18] bunch of short covering. You've got CTAs
[6:21] uh at very large short positions.
[6:23] They're systematic strategies that are
[6:25] going to have to cover. And by the way,
[6:26] the reason these positions are so big is
[6:28] cuz we never saw the vol expansion that
[6:30] we saw down here. So, when you look last
[6:33] year, we didn't really see the buy flow
[6:34] from CTAs. It was spread all out of
[6:36] here.
[6:38] I think this one's going to be much
[6:39] faster just because we never saw the vol
[6:40] jump. So, they're going to have to
[6:41] switch back and forth. I think this is
[6:43] going to be a painful year for
[6:45] systematic strategies uh as opposed to
[6:47] last year where we got this nice smooth
[6:49] uptrend. I think this is going to be a
[6:51] volatile, uncertain year. Now,
[6:54] that gap back after liberation day, that
[6:56] came from, believe it or not, the
[6:59] US-China agreement on
[7:02] in uh in Switzerland last year on May
[7:05] 12th. That's what the gap up was that
[7:07] time. There was this statement on the
[7:08] White House. Guess what happened again?
[7:11] Monday.
[7:12] Another release. Peace through strength.
[7:14] Operation Fury. Again,
[7:17] this is we know the weak point with the
[7:18] administration. At some point with gas
[7:21] at the pump higher, the midterms in
[7:22] check, stocks down,
[7:24] you got the flip.
[7:26] Everyone uh had to cover shorts. And now
[7:29] we're back to a more neutral stance. The
[7:30] difference is, and this is what we're
[7:32] going to see. Last year, after we got
[7:35] the bounce out of liberation day, here
[7:37] are the estimate revisions for the end
[7:39] of the year. This week, we have a big
[7:42] down move. Before then, we had nothing.
[7:44] Analysts did not turn negative. Why did
[7:47] they turn negative here on tariffs?
[7:49] I don't really know in hindsight. They
[7:51] had to revise everything back up. This
[7:53] time, the bar
[7:54] is far different. And that's what you
[7:56] have to remember. We kept getting
[7:57] earnings revisions last year because
[7:59] people expected inflation to be higher
[8:02] and expected there to be a recession.
[8:04] Now, with inflation certainly to be
[8:06] higher, it never budged here. We've
[8:08] already seen it go higher. Analysts are
[8:10] not turning negative. I think there is
[8:11] negative news to come both from AI and
[8:13] the software side. Uh everyone keeps
[8:16] hoping that there'll be good news on
[8:17] software. At this point, with good news,
[8:20] software has gone down. Can you imagine
[8:22] what'll happen if we start getting bad
[8:23] news in software where you can
[8:24] extrapolate it out to three years? Uh
[8:26] that will come at some point, mark my
[8:28] words, because this is a true
[8:30] disruption.
[8:31] Last year,
[8:32] we still had rate cuts going on around
[8:35] the globe.
[8:36] So, during this whole time period last
[8:38] year in 2025, while we still had into
[8:42] September, we still had rate cuts. Well,
[8:43] now we've got rate hikes
[8:46] around the globe that are going to
[8:47] start.
[8:49] And last year, we never got a budge in
[8:50] inflation. In fact, here were the
[8:52] inflation numbers. And today,
[8:55] we got a bump up of 0.9 on headline CPI,
[8:58] the highest monthly number since
[9:01] the peak in year-over-year CPI back in
[9:03] June of '22. So, we already are seeing
[9:05] inflation go higher. Now, core was
[9:07] lower. I'll get into that later. Uh this
[9:09] will be the year that uh you'll learn
[9:12] the difference between core and
[9:13] headline. But, it's going to come at a
[9:15] time when affordability is an issue and
[9:17] when there's a distribution of wealth
[9:18] problem. Uh higher gas prices, higher
[9:20] plastic prices, higher diesel prices,
[9:22] this is not going to do well for either
[9:24] the administration or it's not going to
[9:26] do well for the market in terms of a lot
[9:28] of things in the past through. Here's
[9:30] gas prices. This is the 40-day rate of
[9:33] change. The reason I want to bring this
[9:35] up, this is the biggest move in the last
[9:38] 22 years in terms of gas over a 40-day
[9:41] period. That's we have a lot of
[9:43] inflation still to come.
[9:45] Diesel, even more dramatic. Diesel,
[9:48] every truck in the country, everything
[9:50] related to refined products is going to
[9:52] continue to see. And right now, we
[9:54] haven't seen any pullback in this. So,
[9:56] even though we have a ceasefire, we've
[9:58] barely had a budge in even crude
[10:00] futures. We had one up day on Friday
[10:02] before the weekend of 11 bucks. Now,
[10:05] we're back down and we basically settled
[10:07] where we were before the Friday. So, as
[10:08] of the Thursday, we're down about four
[10:10] or five bucks.
[10:12] But, diesel prices and all these refined
[10:13] products, we got a lot of damage there.
[10:15] So, I think the big difference is how
[10:16] long is it going to take for this stuff
[10:18] to normalize? Took a long time in 2022.
[10:21] Uh we'll see how long this takes. This
[10:23] is uh basically plastic uh again,
[10:26] straight up. So, this is the plastic
[10:28] price, which is going to flow through to
[10:29] everything.
[10:30] Uh this is for fertilizer. We haven't
[10:33] seen food prices up yet. We haven't seen
[10:35] the the plastic prices up yet. These are
[10:38] all prices that are going to flow
[10:39] through. So, that CPI number you saw
[10:41] 0.9, you can expect at least another two
[10:44] months of numbers that are going to be
[10:45] somewhat similar uh unless we see a
[10:48] quick reversal, which just seems
[10:49] unlikely. Today, China moved to ban
[10:52] sulfuric acid. That's a key component in
[10:54] a lot of different things.
[10:56] But, in particular, it's part of the
[10:58] fertilizer uh side. It's also part of
[11:00] the copper side. So, them banning that,
[11:03] again, when the Chinese ban something,
[11:04] that highlights that the shortage is
[11:06] there. It means they're focused on that.
[11:08] It means someone is suffering from that
[11:09] cuz China is a a big exporter. So,
[11:12] they're banning that. You still have
[11:14] helium prices that have spiked. That
[11:15] goes into a lot of different things, but
[11:17] obviously semiconductors.
[11:20] You think about the breadth of the
[11:21] problems that are coming from the
[11:22] commodity shortage side, which is not
[11:24] going away. Last year,
[11:26] we had rate hikes or rate cuts still
[11:29] built into the market the entire year,
[11:32] and we actually ended up getting rate
[11:33] cuts. This time, we had rate cuts before
[11:36] the war as of here. We had 2 and 1/2 in
[11:38] the market. This is for the December or
[11:40] the January Fed funds.
[11:42] We're zero now. We've taken out the
[11:45] cuts.
[11:47] Here's the services PMI. We got the PMI
[11:49] this week. This is prices paid overlaid
[11:51] with the CPI. Now, remember, this is the
[11:54] service
[11:55] PMI. I just want to highlight that the
[11:58] service PMI matters a lot from prices
[12:00] paid. And aside from this period in
[12:03] 2022, up at 71, you're talking about the
[12:07] highest level since 2010.
[12:10] Now, this CPI print, we're now up at
[12:12] 3.2. So, we're up here.
[12:14] We have two very easy comps coming over
[12:17] uh uh the next two months. It is highly
[12:19] likely, based on what's going on, that
[12:21] we will be well above 4%, maybe even
[12:23] let's assume 5%.
[12:25] At 5% CPI, if we get there, we're
[12:28] talking about the highest number since
[12:31] just before the great financial crisis.
[12:33] If somehow or another we get above 5.8
[12:36] and up to six, you're talking about the
[12:38] highest number since the 1990s.
[12:40] So, again, I think it's very hard to
[12:42] look through what's happening on CPI.
[12:44] Now, for core, even though the core came
[12:46] in at 0.2 today,
[12:48] the PCE core, which is the Fed's chosen
[12:51] part, we just had our third month in a
[12:53] row of greater than 0.3.
[12:56] The only time over the course of the
[12:59] last 25 years that we had three months
[13:02] in a row of greater than 0.3
[13:04] was during the inflation boom here. It's
[13:07] the only time, guys. We haven't even had
[13:09] back-to-back months, except for this one
[13:11] little point at the end of '24. So,
[13:13] we've had three months in a row of core
[13:15] PCE above 0.3. A lot of this, again, has
[13:19] to do
[13:20] with technology.
[13:23] Core PCE was firm in February because of
[13:25] goods. Goods inflation was 0.84.
[13:29] Okay.
[13:29] Again, if you exclude '20, it's the
[13:31] highest run rate since 1991.
[13:34] You cannot exorb ignore this. You just
[13:36] can't. Uh so, here we were, 0.2, 0.9,
[13:40] 3.3 on the headline, 2.6 there. Market
[13:43] initially liked it, and then I think we
[13:45] finally have reached a point. The wedge
[13:46] between core goods in the CPI
[13:49] at 0.2 and PCE very high is due to the
[13:52] computer software and accessories
[13:53] category.
[13:57] This is a problem, guys.
[14:00] And look here, the software category was
[14:02] 6.6 five in February, has 40 times more
[14:05] weight in the overall PCE price. As I go
[14:07] through this, I'm just going to tell you
[14:09] that compute prices are only going to
[14:11] get worse going forward.
[14:14] Uh cell phones, computers, we've got
[14:16] shortages in Mac minis and Mac studios
[14:19] that are extending further, and we're
[14:21] going to have a server problem, CPU
[14:23] problem. All of these things are coming.
[14:25] Uh
[14:26] the Fed expects little change to
[14:28] underlying inflation. It's great. We've
[14:30] got a 0.9 versus a 0.2. We put it post a
[14:33] couple more months of this, and the
[14:34] Fed's going to leave that alone, which
[14:37] means rates are going to stay below
[14:38] inflation for a period of time. Look at
[14:41] the ISM PMI report, and just go to the
[14:44] little section in the report.
[14:45] Commodities reported up or down in price
[14:47] and short supply.
[14:49] So, if you think there's a bull market
[14:50] in commodities, look, up and down in
[14:53] price and in short supply.
[14:57] Aluminum, every single commodity here.
[14:59] This is not just oil. This is everything
[15:02] in here. Nothing is down in price.
[15:05] Commodities in short supply, electrical
[15:07] components, bearing components,
[15:09] electrical components, memory, rare
[15:11] earth. All of this is AI related.
[15:14] This is only going to get worse, guys.
[15:18] This is going to be a problem. This is
[15:20] one of the positive things for Bitcoin.
[15:22] For all of you uh who follow me with
[15:24] pump and your your Bitcoin lovers,
[15:28] uh and all you institutional people that
[15:30] don't embrace Bitcoin, start to embrace
[15:32] it. Uh if you've seen some of my calls,
[15:34] and if you've gone through this, and you
[15:36] think I've had a decent uh a decent look
[15:38] at AI, especially with some of the
[15:40] names, I'm telling you this is the thing
[15:42] Bitcoin is made for. Um you're going to
[15:44] have negative real yields very soon on
[15:46] headline inflation, and I'll show you
[15:47] why that matters. But, you're also going
[15:49] to have problem in a lot of countries
[15:51] based on what's happening with the
[15:52] Middle East in terms of EM inflation and
[15:54] PPI prints
[15:55] saying we're going back to double digits
[15:57] globally. You're going to start seeing
[15:58] some devaluations in some of these
[16:01] currencies over in emerging market land.
[16:04] Um
[16:04] and we'll go through the ones that I
[16:05] still like, like Brazil. Uh again,
[16:08] abundance versus scarcity. That's what
[16:09] that whole thing is for. I can't say it
[16:11] any louder. We've got inflation over
[16:13] here. We have deflation over here in
[16:15] terms of software problems.
[16:17] Financials are fitting into it. The
[16:19] problem is you can't print your way out
[16:21] of this one. So, for the scarcity side,
[16:24] it's not getting any better because
[16:26] where you going to find the the supply?
[16:27] It either comes through demand
[16:28] destruction,
[16:30] uh which I don't think is going to
[16:31] happen cuz I think they're going to deal
[16:32] with the energy side. But, just
[16:33] remember, this whole physical upgrade
[16:35] trade,
[16:36] first inning.
[16:39] Jeff Currie on the structural trade, and
[16:42] I mean, he's getting in it. I I could
[16:43] not agree more. The bargain that trades
[16:46] US security for dollar dominance is
[16:48] breaking down. So, think about what
[16:50] happened with NATO in terms of the
[16:52] arguments. Think about what Trump posted
[16:53] on Saturday and how it just seemed to be
[16:56] this flailing thing uh about what was
[16:58] going to happen. So, we've backed off.
[17:00] That's all well and good, but the
[17:02] question is, does the world trust the US
[17:05] to be there when it needs it? The
[17:07] nations didn't come in. This There's a
[17:08] much bigger story here that impacts the
[17:11] odds. Regardless of what your political
[17:13] beliefs are, you have to view this as
[17:15] where does the dollar sit in this whole
[17:17] thing? So, again, the dollar rallied
[17:19] during a crisis mode. It fell right back
[17:21] off during a period here.
[17:23] I agree with Jeff on this thing. And one
[17:25] of the things he adds,
[17:27] if I if Currie could put all his money
[17:29] into one technology, it's batteries. The
[17:31] military case, the grids case, the
[17:33] security case, all converge on storage.
[17:36] I've been talking to a lot of you about
[17:37] batteries. It's a hard thing to play.
[17:39] Lithium is on the critical minerals
[17:40] list. I have a bunch of uh battery
[17:43] stocks. A lot of these have fallen, like
[17:45] EOS have fallen sharply with the
[17:46] software unwind and the growth unwind.
[17:50] I still think the battery trade, and
[17:52] Tesla's fallen off, one of my favorites,
[17:54] too. Uh all the way back down to where
[17:56] it was in the summertime. The battery
[17:58] and energy trade for these names is
[17:59] going to happen again. Uh just got to
[18:01] wait for a little momentum to start
[18:02] going, but we saw some change this week.
[18:05] I think this is going to be a bigger
[18:06] theme coming out of this. Computer
[18:08] equipment investment just hit 1% of GDP.
[18:10] Just look at these numbers, guys, and
[18:12] just look back in history. Business
[18:14] computer equipment investment as a
[18:16] percentage of GDP.
[18:18] This is why being long consumer or being
[18:20] long software or being long any of this
[18:23] stuff about the way GDP was in the past
[18:26] is completely irrelevant. We are now in
[18:29] a commodity hardware bull market,
[18:31] software, all of that. Last week, I put
[18:34] something out on transports. Just like I
[18:37] did when I put a Marvell piece out, and
[18:38] I got pushback, and I got a pushback on
[18:41] a whole bunch of things. I love pushback
[18:44] right now.
[18:45] Transports, there was pushback. You guys
[18:47] are missing the boat on this. For people
[18:49] who are permabears on AI, suffer the
[18:52] consequences. For people that are
[18:54] bearish on the economy right now,
[18:57] stop looking at housing. Stop looking at
[18:59] autos. Stop looking at all of the things
[19:01] related to the consumer. The consumer is
[19:03] fine. The wealthy people have more money
[19:05] to spend than you can believe. Transfer
[19:07] payment people are getting 5-plus
[19:09] trillion a year. We have ex-
[19:11] expenditures almost up to 20% of
[19:13] healthcare, which is non-cyclical. This
[19:15] is not 2006 before the baby boomers
[19:19] retired. What we have right now is a
[19:21] boom in this white line, which is
[19:23] capital goods X
[19:26] air and defense.
[19:28] This is a straight line up, and this is
[19:30] overlaid with the transport sector. So,
[19:32] everyone who sent me something when I
[19:34] showed
[19:35] flatbed rates last week and said the
[19:38] economy's not going, you're missing it,
[19:40] blah blah blah. I think you guys have to
[19:42] do some homework right now and start
[19:43] looking at more than some people that
[19:45] you like to read who are bearish.
[19:47] Uh Goldman Brack Goldman Sachs embraces
[19:49] the picks and shovels of AI with more
[19:51] capex ahead.
[19:53] AI trade in full full full
[19:56] full-blown bubble right now.
[19:58] Uh
[19:59] all right, I'm going to write a piece
[20:00] about Nvidia this week. Uh it's it's in
[20:03] at least one, I think two of the
[20:04] baskets, but definitely one of them. So,
[20:06] it's on my thematic list. As we've had
[20:08] multiple compression as a big story this
[20:10] year, we've now taken Nvidia's
[20:13] 2027 PE
[20:16] down to the lowest level in the last
[20:17] decade.
[20:18] Stock is having this nice consolidation
[20:21] here, in my opinion.
[20:23] Uh
[20:23] in my opinion, it's going higher.
[20:26] Earnings just continue to grow while the
[20:29] PE has sunk down to 20.
[20:31] We are at an agentic side, and if you
[20:33] don't believe me,
[20:35] go listen to this podcast.
[20:37] Uh part of the reason I'm writing this
[20:39] paper on Nvidia is Jensen Huang has been
[20:42] a gift that's kept on giving. I run
[20:44] almost all of my AI models on things
[20:47] that he says during the week, and the
[20:48] names that I get, like a Marvell, like
[20:52] Coherent, any of these ones that he's
[20:54] made investments in,
[20:55] he tells you way ahead of time to put
[20:57] money in. Not literally, but if you use
[21:00] AI and you create a knowledge base, go
[21:02] look up Andrej Karpathy and figure out
[21:03] what a knowledge base is,
[21:05] you can go out and you can start getting
[21:07] data, but you have to go through this to
[21:09] figure out where we are in AI right now,
[21:11] because everything he said,
[21:15] the market got inference wrong. And when
[21:17] he says the market, it's pretty much
[21:20] everyone, guys.
[21:21] Inference demand is exploding because
[21:23] models are no longer just answering
[21:25] prompts. They're increasingly reasoning,
[21:26] planning, searching. This was not
[21:28] supposed to happen this quickly. So,
[21:30] it's not just that people were wrong,
[21:33] they were wrong this soon. So, Open
[21:36] Claw, Cohere, Perplexity, compute, every
[21:40] single one of them is now agents. Agents
[21:42] are running everywhere.
[21:44] We don't have enough compute to deal
[21:47] with all these agents, and now
[21:48] enterprises are adopting at a pace far
[21:51] faster. If you've seen Anthropic's
[21:53] number, go back to what people were
[21:54] saying. There was no way to monetize AI.
[21:57] That was 6 months ago, guys. And now we
[22:00] were at a run rate of 30 billion, up
[22:02] from
[22:03] 14 billion, up from 9 billion. The
[22:05] numbers keep growing now in terms of
[22:07] this because they're taking in tons of
[22:09] enterprises. So, when you go through
[22:11] this, we need more CPU, we need more
[22:13] memory, we need more switching, we need
[22:15] more networking, we need more power, we
[22:16] need more cooling, we need more storage,
[22:18] we need more software.
[22:21] All of this stuff is what you can invest
[22:23] in outside of software. He only says
[22:25] that because a lot of his customers are
[22:27] in there in software, but we do need
[22:28] more software. It's just going to come
[22:30] from smaller businesses.
[22:32] Everything is about performance per watt
[22:34] and heat, everything.
[22:36] Computers are turning into AI factories.
[22:38] It's becoming the center of the system
[22:40] and the center of the business model.
[22:43] I'm about to start
[22:45] thematic ideas on the edge device side,
[22:48] because we are now at the point,
[22:50] especially with Mythos, as I'll get
[22:52] through that. If you haven't read the
[22:53] paper I wrote on Mythos last week, uh
[22:56] guys, you really need to subscribe and
[22:58] be a part of this just because I don't
[23:00] see the uh street as being on top of
[23:02] this in terms of what how to go through
[23:04] it. This is really a macro theme that
[23:06] goes across technology.
[23:08] And when you go through the edge device
[23:09] side, there are a lot of winners that
[23:11] are not part of the current trades. This
[23:15] is why I spent time on the rack. The
[23:16] rack is part of it, but there's other
[23:18] places to go through cuz remember, as
[23:20] much as right now we've got AI
[23:22] disrupting software, it's going to get
[23:24] into the physical world through
[23:25] networks, vehicles, robots, satellites.
[23:27] That is going to disrupt other parts.
[23:30] So, it's not just a long side, you have
[23:32] to be on top of this for the short side
[23:33] as well, because a lot of the alpha this
[23:35] year has been traded by using
[23:37] hyperscalers on the short side and using
[23:39] software on the short side versus semis,
[23:42] but other longs.
[23:43] I want to bring this up, too, because he
[23:45] also the only non-tech company he
[23:48] mentions
[23:49] is Eli Lilly. It's a gift for you
[23:51] non-subscribers. Eli Lilly's in my
[23:53] basket. The reason it's in my basket is
[23:56] because Nvidia and Jensen Huang talks
[23:58] about him more than anyone on this.
[24:00] Uh uh
[24:01] The chart looks like it just had a
[24:03] reset. They have earnings growing
[24:05] because of uh
[24:07] Ozempic or whatever their version is. Uh
[24:10] but you also have AI discrub drug
[24:12] discovery, and I wrote this paper in
[24:13] November. This is still a theme.
[24:16] It's still real.
[24:19] Here's the Anthropic chart. So, again,
[24:21] everyone that was fading AI adoption
[24:24] could not be more wrong. It's gone
[24:25] parabolic. This is the Open AI chart.
[24:28] Don't be surprised if Open AI starts to
[24:30] pick up, but regardless,
[24:32] uh
[24:33] this number has gone parabolic. I will
[24:34] say, remember the last time Anthropic
[24:36] said what their earnings were, they had
[24:38] a problem because the margins were still
[24:39] coming down.
[24:41] They're going to have to raise prices,
[24:42] but we have a bigger issue.
[24:44] And the issue is,
[24:46] as Warren Pies shows in probably the
[24:48] best thing of
[24:51] measuring supply and demand, there is no
[24:53] compute available, guys.
[24:55] Uh literally, there is no compute
[24:57] available. So, look at how far down this
[24:59] stuff has gone this year as the agentic
[25:01] side has kicked in. We can't keep up
[25:04] supply versus demand. That's why, as
[25:06] Axios put out uh this week, AI compute
[25:09] wars. Now, again,
[25:12] Anthropic's server capacity isn't
[25:14] keeping pace with demand, leaving paying
[25:16] customers stuck on usage limits and
[25:18] outages.
[25:19] I can confirm that that has happened to
[25:21] me for the very first time. Yes, I had
[25:24] issues with Anthropic a long time ago,
[25:26] but once I upgraded to the highest
[25:28] package, I never had a problem. I did
[25:30] this week. Dylan Patel, I highlighted
[25:33] this video that he did in uh in terms of
[25:35] uh interview with Dwarkesh Patel, his
[25:38] roommate,
[25:40] highlighted that Anthropic may be pushed
[25:41] towards lower quality computers. Open AI
[25:43] locks up premium supply. Commute compute
[25:46] costs are plummeting as efficiency in
[25:49] chips and uh uh
[25:50] increase. However, usage is skyrocketing
[25:53] faster, so total spending keeps
[25:55] climbing. It's Jevons paradox. Even at
[25:58] rack record capex levels, the industry
[26:00] isn't buying enough compute to meet full
[26:02] demand.
[26:04] On the 30 billion, this is what they had
[26:05] to do this week. Anthropic expands
[26:07] partnership with Google and Broadcom for
[26:09] multiple gigawatts. They're basically
[26:11] going to start using TPUs. This is what
[26:13] Dylan had talked about. They don't have
[26:15] a choice. This is what they put in there
[26:17] 2 years ago, a dozen customers
[26:19] paid Anthropic over 1 million uh on an
[26:22] annualized basis. Today, that number
[26:24] exceeds 500.
[26:25] So, 2 years ago, they've gone from a
[26:27] dozen customers to 500. These are uh
[26:30] Palantir-type numbers. Eight of the
[26:32] Fortune 10 are now Claude customers.
[26:34] Eight of the Fortune 10 are Claude
[26:36] customers. That is not good for
[26:38] Microsoft. That is not good for the
[26:40] software companies. Business
[26:41] subscriptions to Claude code have
[26:43] quadrupled. Uh you guys can read this.
[26:46] They're getting into financial and data
[26:48] analysis, sales, cybersecurity,
[26:49] scientific discovery and beyond.
[26:52] In January alone, we launched more than
[26:53] 30 products and features. That's how
[26:55] quickly this stuff can come out. Think
[26:57] about Copilot. Think about how useless
[26:59] it is.
[27:01] 3 years plus since ChatGPT, and that's
[27:05] what we have. They put out 30 new
[27:06] products, including
[27:08] PowerPoint,
[27:09] including Excel.
[27:11] Uh
[27:12] software companies cannot move at the
[27:14] speed. It's just not possible. Anthropic
[27:16] admits Claude code's
[27:18] users used limit way faster than
[27:20] expected. Um again, nobody was ready for
[27:23] this. So, this was the $200 one. They
[27:26] just unveiled manage agents. So, this is
[27:28] their competition for Open Claw.
[27:31] It's getting easier and easier to use
[27:33] agents, guys. Uh if you doubted the
[27:35] token use from them, here's the token
[27:37] use going up through the roof. And
[27:38] remember,
[27:40] this is basically mainly
[27:44] models that are open source.
[27:46] So, the majority of this
[27:49] is open source. So,
[27:51] the frontier most expensive models have
[27:53] gone from 22% to just 4% today.
[27:59] Alex Finn, um a must must watch, but
[28:02] also a must listen to in terms of his
[28:04] ex-post and going through it.
[28:06] Uh something colossal is going to happen
[28:08] in the next 6 months. He's writing this
[28:11] all related to AI agents, which again
[28:12] goes back to the point. All of my 100
[28:15] names in the thematic baskets, one way
[28:17] or the other, are directly connected to
[28:19] AI agents or at least AI adoption. They
[28:21] are all on the physical side, but they
[28:23] all fit in with the fact that we don't
[28:25] have enough compute.
[28:27] People keep fading this, but I'm just
[28:29] telling you that when you go through,
[28:32] he's also highlighting something I've
[28:33] said to parents, professors, kids at
[28:36] school, my son, as I've talked about,
[28:39] it has never been more critical that you
[28:41] were up to date on all the latest AI
[28:43] tools. The future is entrepreneurship.
[28:46] Jobs are about to be impacted. So,
[28:48] whatever you thought about the job
[28:49] situation,
[28:50] be prepared now cuz we're getting into a
[28:52] very different situation with Mythos,
[28:54] which again, I'll cover again. But
[28:56] Mythos is something you need to
[28:57] understand.
[28:59] Uh Michael Dell spoke.
[29:01] He talked about how much memory demand
[29:03] is surging, and again, he did this at a
[29:06] Bank of America event this week.
[29:08] Uh I just wanted to highlight how big it
[29:10] is and just
[29:12] that it's all happening at the same
[29:14] time. The demand is just picking up. He
[29:16] noted that investment pullbacks by
[29:18] memory companies amid the 2023 downturn
[29:20] have resulted in supply shortages.
[29:22] Again, this is a very similar story to
[29:24] what's happened commodities. There's
[29:26] just been underinvestment in the space
[29:27] because no one wanted to get hurt again
[29:29] hurt again in terms of overinvesting.
[29:32] Everyone, investors,
[29:35] uh
[29:35] semiconductor companies, people
[29:37] underestimated how quickly inference
[29:39] would get here
[29:40] and go to the demand. The reason is
[29:43] because of this. Not just because of
[29:46] Mythos, but once Opus 4.5 came out, the
[29:48] world changed. I'm going to show you
[29:49] more charts on that. Make sure that from
[29:51] an investment standpoint, you understand
[29:53] the reason why it's important. The
[29:55] Mythos signal. I wrote this for 22V last
[29:58] week. Again, for the subscribers who saw
[30:00] it.
[30:02] Mythos matters not just for what it may
[30:04] be,
[30:05] but for how it surfaced that's evidence
[30:06] that frontier may be moving faster in
[30:08] private than the public fully
[30:11] Recursive self-improvement basically
[30:12] means that the models are able to make
[30:15] themselves better,
[30:16] develop things on their own.
[30:19] Here are the podcasts you should go
[30:21] listen to to learn more about Mythos.
[30:23] Uh I listen to this one regularly, the
[30:25] AI daily brief. I haven't really
[30:27] mentioned it too many times on here.
[30:29] Um
[30:29] no real reason. It's a short one, but
[30:31] it's every day. But they went through
[30:34] should we be scared of Anthropic's
[30:35] Mythos? They also put one out the calm
[30:37] before the AGI storm. It's here, guys.
[30:40] We are at that moment. Uh that is the
[30:42] reason why software dumped this week is
[30:44] because now people are starting to come
[30:46] to terms with Mythos, which again came
[30:48] out last week or at least was released
[30:50] leaked last week. So, speaker says
[30:53] Anthropic has officially confirmed
[30:54] Mythos, its most powerful model yet, but
[30:57] it is not releasing it publicly. It is
[30:59] flaming It is framing it as powerful
[31:02] enough to require a controlled rollout
[31:03] because of the cyber risk. The reported
[31:05] benchmark jump is unusually large.
[31:08] The biggest gains are in coding and
[31:09] agentic tasks. I'm not going to read all
[31:11] this other part. It is capable of highly
[31:14] autonomous behavior, which gets back
[31:16] into the recursive self-improvement
[31:18] side.
[31:19] This is the calm before the AGI storm,
[31:21] guys. Um
[31:23] It's faced backla-
[31:25] Anthropic faced backlash over user
[31:26] limits. Uh Gemini 4 is presented as a
[31:29] major open source model. The next
[31:31] generation of models may represent a
[31:32] real step change with speculation around
[31:35] new Open A multi- multimodal systems and
[31:38] Anthropic's Mythos
[31:39] level advances. The biggest conclusion
[31:42] is that the industry may be approaching
[31:43] a phase where incremental updates are no
[31:45] longer enough and a much bigger
[31:47] transition towards superintelligence is
[31:48] starting to come into view.
[31:50] Something that people didn't believe in
[31:52] last year, but didn't think was anywhere
[31:54] near where there.
[31:56] Why Anthropic's most powerful AI model
[31:57] Mythos preview was too dangerous for
[31:59] public release?
[32:01] Anthropic model scare sparks urgent
[32:03] Bassant Powell warning to bank CEOs.
[32:06] He called for an emergency meeting. When
[32:08] we have emergency meetings,
[32:11] and this was so sudden that Jamie Dimon
[32:13] couldn't attend.
[32:14] So, this is when they saw the model
[32:18] and the Treasury Department wanted to
[32:20] have a meeting on it. If you guys are
[32:24] not alarmed at this or not paying
[32:27] attention to how important this is, then
[32:28] I'll keep going. Anthropic teams up with
[32:31] its rivals to keep it from hacking
[32:33] everything.
[32:35] Project Glasswing was created.
[32:38] We're announcing Project Glasswing, a
[32:40] new initiative that brings together
[32:41] Amazon Web Services, Anthropic, Apple,
[32:44] Broadcom, Cisco, CrowdStrike, Google, JP
[32:46] Morgan Chase, blah blah blah.
[32:49] It's all in an effort to secure the
[32:50] world's most critical software. For
[32:52] people that doubt how strong the model
[32:54] is, you don't do these types of things.
[32:56] They showed it to the government. The
[32:58] Remember the government's at war with
[32:59] Anthropic.
[33:01] And Bassant still calls a meeting.
[33:04] AI models have reached a level of coding
[33:06] capability where they can surpass all
[33:08] but the most skilled humans at finding
[33:10] and exploiting software vulnerabilities.
[33:12] I'm not going to give you the details on
[33:14] what it was able to do, but let's just
[33:15] say it was able to break into things
[33:17] that
[33:18] were thought safe for 27 years since the
[33:20] time they went up.
[33:22] OpenAI is working on a cybersecurity
[33:24] model to rival Anthropic's Mythos.
[33:28] Google interview with Sundai Sundar
[33:31] Pichai on Cheeky Pint. I recommend this
[33:34] one.
[33:36] I'll just read this one. These models
[33:37] are definitely really going to break
[33:38] pretty much all software out there. You
[33:40] guys can go listen to the rest of it.
[33:43] This all this led to software stock
[33:45] sell-off. Palantir, Microsoft drop on
[33:47] Anthropic's Mythos. I've been pitching
[33:49] Palantir relative to Microsoft. It was
[33:51] up 25% over a month during the IGV
[33:55] unwind. It gave it all back this week on
[33:57] the fears that even Palantir would be
[33:59] disrupted. I think when we get the
[34:01] earnings reports, uh they will still
[34:03] have massive growth on the enterprise
[34:04] adoption side, so I don't think this is
[34:06] going to hold, but
[34:08] I am not one to sit there and bet that
[34:10] software isn't going to continue to be
[34:12] under pressure because this is about
[34:13] model compression. And the one thing
[34:15] Palantir has anyway is high multiples.
[34:19] And when Nvidia is sitting at the lowest
[34:20] multiple of the last 10 years and
[34:22] they're in the hardware side, there's
[34:23] just better places to put your money
[34:25] right now than hope that Palantir is
[34:26] going to bounce back. Uh relative to
[34:28] Microsoft is still the way I'd play it,
[34:30] but it's giving back everything and here
[34:32] it is right there.
[34:35] Uh Meta finally coming with a new AI
[34:38] model. Again, this is all for the
[34:40] deflationary component of moving the
[34:41] prices lower, which is screwing the
[34:43] margin side. It makes the hyperscalers
[34:44] in a
[34:45] fight
[34:46] to get prices out there and deliver this
[34:48] stuff. They've got compute shortages.
[34:50] They're in competitive thing. Google
[34:52] released their Bloomberg competitive
[34:54] version. Uh Perplexity has Perplexity
[34:57] Finance. Now you have Google Finance.
[34:58] This is just a snapshot of the page.
[35:00] It's good. Uh and again, as a reminder,
[35:03] this is private equity and IGV versus uh
[35:06] SPX.
[35:10] We had more news in private credit this
[35:11] week. Wow, Carlyle, the new one with
[35:14] requests totaling 15.7%.
[35:18] They will limit shareholder redemptions.
[35:20] Uh this won't stop and remember, these
[35:22] are not a minor deal, guys. This is a
[35:24] big deal.
[35:25] Uh
[35:27] They have to Next time they come up, I
[35:29] mean, they have to come up with more
[35:30] money. So, the problem is fundraising
[35:34] for private credit is done and
[35:36] fundraising for private equity falls to
[35:38] the slowest pace in a decade. Remember,
[35:41] if you've got redemptions leaving, think
[35:42] about this like a Ponzi scheme.
[35:45] Private credit's not a Ponzi scheme, but
[35:46] they have a certain amount of money.
[35:48] They leverage it by a certain amount to
[35:50] get the returns that are up near equity
[35:52] levels. Let's assume they leverage it by
[35:53] about 1 and 1/2 to 1 and 3/4, which
[35:56] seems about industry standard. And they
[35:58] have about 25 to 30% of the money that
[36:02] they have, not the leveraged money, of
[36:04] the money they have from investors in
[36:06] software.
[36:07] Then you multiply that number times 1.7
[36:09] and you end up with about 50% of the
[36:13] cash
[36:15] is in something related to software for
[36:17] a lot of these funds. And if money keeps
[36:19] leaving without money coming in,
[36:22] they have to sell those bonds if they
[36:24] can. They have to sell the loans. This
[36:27] is where the issue starts to become a
[36:28] bigger issue. So far,
[36:31] they seem to have been trying to plug
[36:32] the hole.
[36:33] Now we're going to start getting into
[36:35] this every quarter. This doesn't go
[36:36] away. I will keep saying it.
[36:39] Pro- credit situations and credit cycles
[36:41] do not stop on their own.
[36:43] This isn't a private credit issue that
[36:46] will continue.
[36:50] It's starting to come into part of the
[36:52] muni market.
[36:57] Software pain is about to get a lot
[36:59] worse as the public software companies
[37:02] go down and the private equity markdowns
[37:05] happen, obviously the debt has to get
[37:08] hit, too. And who wants to buy anything
[37:11] related to software right now? So,
[37:13] that's your issue.
[37:15] Uh I mentioned this before. I will keep
[37:17] showing it until it's no longer true.
[37:19] When financials are below the 200-day
[37:21] moving average and they are the
[37:22] worst-performing sector, which is where
[37:23] we are right now,
[37:25] bad things happen.
[37:26] Uh
[37:28] Even with the bounce we've seen,
[37:30] financials are still there. And as I
[37:31] showed you, this happened this week. So,
[37:34] software went down, but we didn't get a
[37:36] bounce in private equity. What we did
[37:39] get a bounce in
[37:40] is Bitcoin.
[37:42] So, for everyone there who loves Bitcoin
[37:44] and
[37:45] watches me and reads my stuff on
[37:47] Bitcoin,
[37:48] uh this will be your favorite five or
[37:50] six slides. For the institutional side,
[37:52] you may not like this.
[37:54] Uh by the end of this year, guys, you're
[37:56] going to be focused on it. Um it was a
[37:57] big week for Bitcoin relative to
[38:00] software. It was a big week for the
[38:01] miners as well.
[38:03] It was a big week for Ethereum. We are
[38:05] starting to get to a point where we're
[38:07] getting a break in correlation. So, last
[38:09] Friday was the break in correlation
[38:11] between oil
[38:13] and stocks.
[38:14] We have a break in correlation right now
[38:16] between Bitcoin and software. It is
[38:18] incredibly noticeable. Straight line
[38:20] down,
[38:21] a move higher.
[38:23] This is
[38:26] two years of overlay between software
[38:28] and Bitcoin.
[38:31] Here's the relative trade. Almost back
[38:34] up to the all-time highs. I showed this
[38:36] chart many, many times and said this
[38:38] thing had done well even when it was
[38:39] down here relative to software.
[38:42] Once we get to the point Bitcoin has no
[38:45] disruption to AI.
[38:48] Every bank has a disruption as we're
[38:50] seeing to Mythos.
[38:51] Get ready for hackings. Your money
[38:54] inside the bank, as much [snorts] as you
[38:56] think it is safe, is now at risk of
[38:58] hack.
[39:00] Cryptography is far safer. Quantum is
[39:02] down the road. So, for all of the ironic
[39:04] nature of this,
[39:06] one hacking in a bank, I think a lot of
[39:09] money's going to move into Bitcoin.
[39:11] Get involved with the theory of getting
[39:14] things on chain and getting them safer.
[39:17] This is Bitcoin relative to software.
[39:19] Here is the
[39:22] weekly move.
[39:25] Whoop.
[39:26] Sorry, guys.
[39:28] Here's the weekly move. Largest in the
[39:30] last two years by far, 22%
[39:33] outperformance this week.
[39:36] Here's Bitcoin with MACD crossover,
[39:38] guys. The weekly MACD crossover. I would
[39:41] draw a trend line right down here. We
[39:42] are right on that right now.
[39:44] Uh I think once we get above this high
[39:46] right there with the MACD's now crossed
[39:48] over in there and a break in correlation
[39:51] and the fact that I think we're about to
[39:52] get some really good news,
[39:54] which I'm going to show you now,
[39:56] I'd be ready. So, this is Fed funds
[39:58] versus CPI.
[40:00] Uh
[40:01] right now we've got the Fed funds
[40:03] contract here. CPI actually bumped up
[40:05] today. It's up to 3.3. We're about to
[40:07] cross over.
[40:09] That's
[40:11] See this period here? This is when we
[40:13] had CPI above there. CPI above
[40:17] Fed funds rate doesn't happen too often.
[40:20] Here's three-month bills, real
[40:21] three-month bills. So, this is using
[40:23] year-over-year CPI and three-month
[40:25] bills, which are currently around 364
[40:28] and you've got CPI at 3.2. So, that's
[40:31] where you get this point four number.
[40:33] Uh the the red line here is zero.
[40:37] The reason this matters,
[40:38] here are the returns for Bitcoin using
[40:41] the matrix of
[40:44] are real rates positive
[40:47] and are the Fed funds rates rising. So,
[40:49] think of this as
[40:52] positive real yields. So, we've got the
[40:54] Fed funds rate above
[40:57] uh there and
[40:59] we have Fed funds rate rising.
[41:02] You don't want to be in there when the
[41:03] Fed funds rate is rising. You don't want
[41:05] to be in there when you've got positive
[41:06] real yields. Now, if you've got negative
[41:08] real yields
[41:10] and the Fed fund is on hold or easing,
[41:13] here's your return.
[41:15] All of the return is basically in that
[41:17] period. We are about to enter this
[41:19] quadrant very, very soon. The next CPI
[41:22] print will put us there.
[41:25] That's the time you want to own Bitcoin.
[41:28] Uh
[41:29] just showed you that one. No need to
[41:31] show you that. All right. So, I finally
[41:33] got my thematic list out, 100 names.
[41:35] This is five themes. For those of you
[41:37] who are new subscribers, this is going
[41:40] up on the website. You also have all the
[41:42] thematic list that make up the 100
[41:44] names.
[41:46] Uh that is a bunch of them, which I'll
[41:47] go in and show you.
[41:49] Here is the
[41:51] uh Bloomberg chart of that basket. So,
[41:53] the 100 names uh made new all-time
[41:56] highs. You can see how it's gone
[41:57] straight up.
[41:59] It's up big for the year while the
[42:01] market is down to unchanged.
[42:04] Uh
[42:05] there is a technical scoring thing
[42:07] thanks to John Roehke who did all 100
[42:10] names as of uh Thursday. So, you can go
[42:14] through, pick your names. This is just a
[42:16] small snapshot. These are all Japanese
[42:18] names, but there's a lot I mean, they're
[42:20] mainly US names. Uh but you can go
[42:22] through a lot of these are chemical
[42:23] names, by the way, and and uh memory
[42:26] names.
[42:27] Here are the individual themes. You have
[42:29] the rack. You have chemicals. You have
[42:31] optical fibers. You have advanced
[42:33] packaging and you have power. Uh
[42:36] new highs for
[42:38] four of them. The only one that didn't
[42:40] make a new high, ironically enough, is
[42:42] the energy power one.
[42:44] Uh
[42:44] but I would expect that one, too. So,
[42:46] all of these are different themes and
[42:48] this is what I want to show you with the
[42:49] AI trade. What I'm trying to find you
[42:51] guys are themes that are durable, that
[42:53] are all going to benefit from AI agents.
[42:55] Every one of these was put together from
[42:57] November on. The first one being the
[42:58] optical fiber one when I pitched Corning
[43:01] and included all of the other names that
[43:03] are in there. All of these lists, which
[43:04] total about 95 of the themes, uh are
[43:08] there. And then there's five names
[43:10] including Lilly, including EWZ,
[43:12] including Palantir, including uh
[43:15] silver and then one other one that I
[43:18] included in there, which I can't think
[43:19] of right now. But you guys can see all
[43:21] 100 names. This is the way I think you
[43:24] guys should be invested at this point in
[43:25] terms of an AI trade. And what I want to
[43:27] show you is what's happened since right
[43:30] around the time of Opus 4.5. This was
[43:32] the turning point in everything. So, the
[43:33] green line here is the S&P software
[43:36] index.
[43:37] So, again, Opus 4.5, the rumor start
[43:39] coming out, it gets released couple
[43:41] weeks later, this breaks down.
[43:44] Um this here is my thematic side. This
[43:47] is when the transfer goes into the
[43:49] hardware side. The PMI's break out in
[43:51] December. All of these trades, all of
[43:54] these themes are what I consider the AI
[43:57] theme. Now,
[43:58] you also have this uh orange line here,
[44:00] which is the hyperscalers. The
[44:02] hyperscalers to me, even though they're
[44:04] making numbers, are going to continually
[44:05] be under multiple compression, which
[44:06] means best-case scenario, they kind of
[44:08] go sideways. They are great short
[44:09] relative to this basket. The white line
[44:12] here, not to pick on Dan Ives cuz I like
[44:15] Dan Ives and we got to know each other
[44:17] and I think he does good work. The
[44:19] problem is the where Dan and I disagree,
[44:21] which is the important thing for this,
[44:23] is in software.
[44:24] Dan has a high component of this related
[44:27] to software. So, his software names have
[44:29] dragged him down. We have some
[44:31] similarities in names up here. They're
[44:32] mainly on the semi side. This is the
[44:35] difference, guys. Dan
[44:37] is a technology person. He has viewed AI
[44:40] as a software trade. This is not Web
[44:43] 2.0. I repeat, this is not Web 2.0. AI
[44:46] is not a technology. It is electricity.
[44:49] It needs power. It needs copper. It
[44:52] needs optical fibers. Those are not his
[44:54] strength. This is what's going on and
[44:56] this is why if you have money in these
[44:58] types of things, you want to start
[45:00] figuring out a different way. If you go
[45:01] to my list, you'll have a better way to
[45:03] go through this. You can run it on your
[45:05] own. EWZ's included in there. Made new
[45:08] highs. We are getting close to five-year
[45:10] highs on EWZ. EWZ
[45:13] Brazil
[45:14] benefits significantly from what's going
[45:16] on. I wrote a paper on them uh a month
[45:19] and a half ago. Subscribers can go see
[45:21] it. Brazil in the green compute
[45:22] arbitrage of the AI decade. Highly
[45:25] recommend people understand why Brazil
[45:27] is such a key component of AI, but it's
[45:28] only gotten more important since I wrote
[45:31] that paper because
[45:33] the data centers that are being built in
[45:35] the Middle East are at risk. In the
[45:37] early days of this conflict in the Gulf,
[45:38] Iran made a strategic decision. It
[45:40] struck two AWS data centers in the UAE.
[45:45] Middle East security and data centers,
[45:47] the Latin America factor. Everyone's
[45:48] starting to focus again on how safe
[45:50] Latin America is for data centers, but
[45:52] more importantly, how Brazil already has
[45:55] clean energy and is abundant in it.
[45:57] Mexico and Brazil lead data centers boom
[46:00] amid US policy shakeups. Brazil energy
[46:03] demands and data center growth.
[46:05] Finally,
[46:06] silver.
[46:08] Wrote a piece on silver this week.
[46:11] My favorite commodity in this. You can
[46:13] go learn about why it is a critical
[46:15] component of the AI trade.
[46:17] That's it for me for May this week,
[46:19] guys. Uh I appreciate all the
[46:20] shout-outs. Keep reaching out. Send this
[46:23] to your friends. Subscribe. Subscribe at
[46:25] the at the site.
[46:28] I'll see you guys next week from
[46:30] Brooklyn.