2025 TACO PTSD: Why Strategists, Analysts, and Investors Are Still Fighting the Last War — Jordi Visser (22 marzo 2026)

[0:00] All right, another volatile week.
[0:02] Um
[0:05] a lot of stuff to talk about. Obviously,
[0:07] the majority of it is going to be
[0:08] related to
[0:10] the weakness that we've seen in the
[0:11] market and just kind of going through.
[0:13] But, I do want to make one
[0:15] point at the beginning that I've just
[0:17] noticed and part of what I try to do in
[0:19] these videos is hopefully
[0:21] give you a calmer perspective on on on
[0:24] what's happening. Um everything has
[0:26] gotten way too bullish or bearish.
[0:30] Uh
[0:30] there's a lot of nuances with markets.
[0:32] They can go sideways for a long time in
[0:35] a wide range.
[0:37] But, everyone gets caught into the
[0:39] oversold overbought thing.
[0:41] And as I go through this, um
[0:43] you know, I I I want you to start with
[0:46] something I showed for the subscribers
[0:48] this week.
[0:50] Uh just in terms of and just so you guys
[0:52] can see the whole thing, the negative
[0:54] trends
[0:56] that are going on that are
[0:58] I would say uh unexpected relative to
[1:01] what people thought coming into the
[1:03] year. So, credit's obviously been
[1:05] significantly weaker. Uh
[1:08] we have a problem in the credit markets.
[1:10] And as I go through this every week,
[1:12] there's always six, seven, eight, nine,
[1:14] 10 stories to go through about a
[1:17] situation that is worse than what the
[1:20] non-alarmists are talking about. Uh I
[1:23] don't believe it's systemic cuz I don't
[1:24] believe anything is systemic anymore
[1:26] when you have a a central bank that can
[1:28] create liquidity facilities to slow down
[1:31] what used to be uh spirals. But, due to
[1:35] uh
[1:35] the digital economy and news spreading
[1:38] fast in things like what I do each week
[1:41] where people actually get to know more
[1:42] what's going on, you do get SVB-like
[1:46] speed in terms of uh retail wanting
[1:49] their money back or businesses wanting
[1:50] their money back and you get runs that
[1:53] used to take more time because of it was
[1:56] before the information era. So, credit's
[1:58] weakening, the financials are below the
[2:00] 200-day and the 200-day has been pointed
[2:02] down. I've been talking about this now
[2:03] for 6 weeks and I still see people
[2:06] ignoring it and just pretending like
[2:08] nothing has happened this year.
[2:10] You now have inflation expectations
[2:12] rising rapidly. Fed rate cuts
[2:15] are gone and we actually are now
[2:16] starting to price in, believe it or not,
[2:18] a tightening.
[2:20] Uh
[2:21] my first rule of thumb in building
[2:23] models over my career for recession
[2:25] indicator to make sure that I was
[2:27] reducing risk uh while we saw uh
[2:30] recession expectations go higher is the
[2:33] 126-day, that's trading days, so that is
[2:35] half a year, rate of change of the S&P,
[2:38] which broke through zero on Friday.
[2:41] Uh that was always the first warning
[2:42] sign and I'll show you because only
[2:44] about 20% of the time is the 6-month
[2:46] rate of change negative. And the reason
[2:48] that's important, the S&P 500 is a broad
[2:51] measure of the economy.
[2:53] Uh it's probably about 20% of employment
[2:56] in the economy. It's got every sector.
[2:58] So, when the stock market at an overall
[3:01] level is below zero,
[3:03] there should be recession fears going
[3:04] higher and right now uh they're really
[3:06] not going higher and I'll go through
[3:08] some of that.
[3:09] And it comes at a time where there's
[3:10] been zero job creation and again, I hear
[3:13] people debating whether AI has caused
[3:14] this or not caused it, which is insanity
[3:17] to me. We have zero job creation. If you
[3:19] take out the non-cyclical portion of
[3:22] health care, we actually have negative
[3:24] job creation. That is fairly
[3:25] significant. That is not uh you know, a
[3:29] a a prognostication. That is not That is
[3:32] a fact. We have lost jobs over the last
[3:34] year. S&P down 2% for the week, the
[3:37] third week in a row of down about 2%.
[3:39] So, we're not having volatile days at
[3:42] this point. We're not having any signs
[3:43] of capitulation. But the market is now
[3:45] in a uh a steady uh state decline. IWM,
[3:49] same thing, down for the fourth week in
[3:51] a row. Q's, same thing. They're
[3:53] accelerating a little bit, but again, 2%
[3:55] not some horrific thing. But, here's
[3:58] where we stand.
[4:00] We're below the 200-day moving average
[4:02] in the S&P. So, that uh week, the first
[4:05] week of February, when I went through,
[4:06] or I guess it was the second week of
[4:08] February when I did a webinar and I
[4:10] showed that the financials were below
[4:11] the 200-day and
[4:13] we were getting close to the 200-day
[4:14] turning down, which is now turned down,
[4:16] and how we had not had a time this
[4:18] century, going back to 1999,
[4:20] where financials had broken below the
[4:22] 200-day moving average and the S&P did
[4:23] not follow it. Well, the S&P did follow
[4:25] it, and we're now below it. So, we're
[4:27] below the 200-day in the NDX.
[4:30] We're approaching the 200-day on the
[4:33] Russell. Small caps were a theme that to
[4:35] me was part of the broadening out, but
[4:37] now that energy prices have gone up
[4:39] significantly and we're starting to go
[4:41] from easing
[4:43] to tightening, you've got headwinds now
[4:45] for IWM. Here's where we stand
[4:48] year-to-date. Uh and I picked a point,
[4:50] which I'll show in the next slide. So,
[4:51] this is down 5% year-to-date through
[4:54] November 20th. And again, financials are
[4:57] the worst-performing sector, now down
[4:59] 11% year-to-date, 15% off the highs.
[5:03] Never a good thing and never something
[5:05] to minimize, and never something to
[5:07] ignore. So, the fact that this was going
[5:10] on and it was the worst-performing
[5:12] sector and below the 200-day before oil
[5:14] and before the Fed rate hikes were now
[5:17] being built in, before inflation
[5:19] expectations had gone higher, before the
[5:21] credit had gotten to this level,
[5:25] I I I think people again need to start
[5:27] to realize this is not the same year as
[5:29] last year. And you're going to hear that
[5:30] a lot in this video, because I think
[5:32] last year is having an impact on the way
[5:34] people think about the market.
[5:36] So, here is last year through March
[5:38] 29th. Very similar, down 5%
[5:41] year-to-date, but here's the difference.
[5:43] Financials were outperforming the
[5:45] market. They were in the middle of pack.
[5:47] Similar thing with healthcare, energy,
[5:49] utilities, staples, and materials.
[5:52] But remember, this was because of the
[5:54] fears on deep seek that got this side
[5:56] down. Now you've got tech week again.
[6:00] But this time for different reasons. AI
[6:02] is at a completely different stage. As I
[6:04] said coming into this year, this is the
[6:07] year that we have an AI transition. The
[6:09] regime for AI has changed. This is no
[6:11] longer the IQ building stage of of AI.
[6:16] This is the scarcity stage. What are we
[6:18] running out of? What do we need? We have
[6:20] insane amount of demand for AI, but we
[6:23] don't have enough supply for people to
[6:25] use it. We don't have enough compute.
[6:28] So, we're at a different stage and the
[6:30] competitiveness, as we see all long
[6:32] duration assets are under attack.
[6:34] They're under attack because the
[6:35] progress of AI is going
[6:37] insanely fast. And this is going to be a
[6:40] story, which I think is going to look,
[6:42] and if you were on the subscriber
[6:44] webinar this week,
[6:46] I talked a lot about comparing this to
[6:47] the 1970s. For those of you who didn't
[6:49] see it, I'd go through it just because I
[6:52] showed a lot of charts I'm not going to
[6:53] show
[6:54] on my weekly videos because this is
[6:56] really more about what happened, but I
[6:57] do think you should have an idea about
[6:59] why. So, here we are. This is the
[7:02] rolling 126 day. These are the
[7:04] recessions. Basically, it led every
[7:06] recession that happened in terms of
[7:08] breaking below. The S&P to me
[7:10] year-over-year is the the guarantee
[7:13] signal. We are not negative
[7:14] year-over-year,
[7:15] but we're also not that far off. And by
[7:19] the time we get into the summertime, if
[7:20] things haven't improved and the oil
[7:22] prices are still up above 100, there's a
[7:25] chance that you could get into this, but
[7:27] that would mean credit widening, would
[7:28] mean a lot of things have to go on.
[7:31] At the point now, what I want to see
[7:32] from a sentiment basis,
[7:34] when you break here, remember, everyone
[7:36] thought there'd be a recession. When we
[7:38] broke here, as I'm going to go through
[7:39] this, everyone thought there'd be a
[7:40] recession. This year, nobody thinks
[7:43] there's going to be a recession. Scares
[7:44] the hell out of me and it should scare
[7:46] you and the fact that
[7:47] everyone seems to have taco PTSD.
[7:50] Um
[7:52] here we are.
[7:53] Look at that, positive number now on the
[7:55] rates. We've got rate hike fears
[7:57] starting to build in. So, that's changed
[7:59] this week. It's changed every week. Uh
[8:02] two-year rates, here's where they are
[8:04] around the globe. This is the range over
[8:05] the last, let me see if I did 3 months
[8:08] or 6 months. This is 6 months. We are at
[8:10] the high end of the range in every
[8:12] single market except for China in terms
[8:15] of two-year rates. So, we're starting to
[8:17] bring tightenings in around the globe.
[8:19] Rates are moving higher. Just look at
[8:21] these basis point moves for the week.
[8:23] This is 10-year yield, same story.
[8:25] And yet, I still hear people having the
[8:28] bull-bear debate like something is going
[8:29] on that is completely overrated panic
[8:31] just like I heard the same thing about
[8:33] software being a ghost trade or
[8:35] something that is related to it
[8:37] shouldn't be happening. This is all
[8:38] fake. We'll go through the same thing as
[8:40] last year. And that's what I'm saying. I
[8:42] think people are living in this dream
[8:44] world that last year was the same as
[8:46] this year. And I think the reason is
[8:48] because growth expectations are higher.
[8:50] But at the beginning of last year,
[8:52] everyone thought they'd be fine until
[8:53] the tariff thing came in. So, now we're
[8:55] going to minimize everything and just
[8:56] keep saying that everything is going to
[8:57] reverse.
[8:59] And a war is the same as pulling back
[9:01] tariffs. Um I just I I I don't
[9:05] understand the logic. Um I think you
[9:08] have to get it out of your mind and
[9:10] really prepare for more volatility as
[9:12] I've been talking about from the
[9:13] turbulence model. More trading
[9:14] environment, less investment
[9:16] environment. I do think you want to be
[9:17] looking for levels on the AI trade
[9:20] because the AI trade to me is not going
[9:22] to be broken this year. In terms of the
[9:24] the CapEx side, I know it's the number
[9:27] one fear people have. I've gotten asked
[9:29] a lot on that because I'm negative on
[9:31] credit. I still think at the end of the
[9:32] day, the build-out will happen. Even
[9:35] even there's a change in CapEx from 700
[9:37] billion to 680, it's not going to change
[9:39] the fact that nominal GDP will be
[9:41] strong. Here are the inflation
[9:43] expectations in the swap market,
[9:46] different than the TIPS break-evens, but
[9:48] I wanted to just use this because this
[9:50] is usually a better idea of
[9:52] I would say market sentiment from
[9:54] from the basis of the way people are
[9:56] thinking about things. We are up to the
[9:57] same level we were. The move has been
[9:59] sharp, and that is year-over-year CPI.
[10:02] Year-over-year CPI this time is going to
[10:04] go higher, and that's what's a little
[10:06] alarming. Is back here, we didn't get
[10:08] any change in year-over-year CPI when
[10:10] everyone was worried about it on the
[10:12] tariff side. This one is a certainty.
[10:15] Gas prices are going higher as are many,
[10:18] many other parts of the complex for
[10:21] energy.
[10:22] Rates vol, I've talked about this being
[10:24] the factor where I think you're going to
[10:25] get more deleveraging, especially from
[10:27] cross-asset portfolios, risk parity,
[10:30] multi-strats, anything that uses
[10:32] leverage.
[10:34] I think we're going to be in an
[10:35] environment where we should be more up
[10:37] here. I mentioned that for the VIX. This
[10:39] is overlaid with credit spreads. This is
[10:41] the triple B's. I think again, there's
[10:43] no fear of a recession where last year
[10:45] at this point we had gone up to levels
[10:47] over here. I think we should be in a
[10:50] bigger thought process because the labor
[10:52] market is weaker, the disruption from AI
[10:55] is real.
[10:56] And the inflation side
[10:58] at this point is completely uncertain as
[11:00] to where it's going to go, and we've
[11:02] created no jobs for a year where if you
[11:03] looked at the prior year relative to
[11:06] where we were last year, we had created
[11:08] over a million jobs.
[11:10] This is rates vol versus and here we go.
[11:13] This is the TIPS break-evens for two
[11:15] years. We're up basically at the same
[11:17] point again where we were here. We're
[11:19] starting to see rates vol move higher. I
[11:21] think as long as inflation is high,
[11:23] that's going to be the case.
[11:25] We just have not seen any capitulation
[11:27] whatsoever. I'm going to show a bunch of
[11:29] people trying to pick bottoms and things
[11:31] on this. We haven't had a day yet where
[11:33] people have thrown in the recession
[11:35] towel. It just hasn't happened and I
[11:36] always view it as when down vol versus
[11:39] up vol,
[11:40] um, a variety of other things which I
[11:42] showed on the on the the subscriber
[11:44] webinar, but I just want to make sure
[11:45] you guys see this. We This is down
[11:47] volume to up volume ratio. Um,
[11:50] again, we haven't had any signs of panic
[11:52] here. There hasn't been any capitulation
[11:54] yet whatsoever.
[11:56] I think we're only starting to see it
[11:58] now. This is the 6-month rate of change
[12:01] and this would be my answer as to the
[12:02] first data point that I think is going
[12:04] to be important. I'm surprised this
[12:05] hasn't happened yet, but as I go through
[12:07] this again, I think this is because
[12:09] sell-side analysts have PTSD from being
[12:11] so wrong last year. Sell-side
[12:13] strategists have PTSD from being so
[12:15] wrong last year and the fact that they
[12:18] were this wrong this year crying wolf
[12:20] about something that never came through,
[12:22] I think that's putting them in a place.
[12:23] And they matter a lot for the wealth
[12:25] managers. So, for you FAs that are
[12:27] starting to watch this and are starting
[12:29] to subscribe better information and get
[12:31] at least some things
[12:33] that are relevant. This is what they
[12:34] should be talking about. Right now,
[12:36] revisions have not been lowered, but the
[12:38] 6-month rate of change in the S&P is and
[12:40] you can see the relationship. The S&P
[12:43] year-over-year goes down and then
[12:44] eventually the revisions go down. It
[12:46] happens always.
[12:48] That's what we're seeing now. I fully
[12:50] expect revisions to start moving lower.
[12:53] Um, I think Delta Airlines did everyone
[12:56] a disservice cuz there's absolutely no
[12:58] way that the airlines are not going to
[13:00] be impacted in a major way from what's
[13:02] going on. So, for them to come out and
[13:03] say demand is strong, there's nothing
[13:04] going on. If you go through last year,
[13:08] we saw the exact opposite happen. And
[13:11] that's what got people going. If you go
[13:12] back and look in March, all of a sudden
[13:15] everyone was freaking out.
[13:17] This was just after the election, the
[13:19] tariffs were going through, the Trump,
[13:21] uh, fears were through the roof. This
[13:24] year, for some reason, people are not
[13:26] falling for it this time. But, here's
[13:27] the reality. This is the oil price and
[13:29] again
[13:31] I wanted to show this because you have
[13:33] $98 here for WTI, the one that gets
[13:36] quoted in the papers here. Then you've
[13:37] got Brent, more associated with, let's
[13:40] say, the EU.
[13:41] Then you've got here the Dubai oil. And
[13:44] then you have here the Oman oil futures.
[13:48] We've already gone through the 2022
[13:49] level and you can see the spike here.
[13:52] A lot of these are going to Asia. You've
[13:54] got shortages around the globe. So, for
[13:56] people that are trying to predict the
[13:57] end of the war,
[13:59] I want to deal in facts and I want you
[14:00] guys to deal in facts. Maybe oil does
[14:02] come down, but very seldom do you see
[14:04] spikes like this that are going to go
[14:06] all the way back to here. And right now
[14:08] we're up at 167. So, if it comes back
[14:11] here and this goes up to 99 or stays up
[14:13] here, you're still going to have an
[14:15] impact on the inflation side that's
[14:16] going to last. Gas at the pump
[14:19] as of this morning, 3.93. I did this
[14:22] before, it was 3.91. Here are the
[14:25] futures as of the close of Friday,
[14:27] implying a number that is more towards 4
[14:30] and 1/4. You can see the spike here.
[14:33] Um I put this together because I wanted
[14:35] to make sure that you guys were thinking
[14:37] outside of oil because there's been
[14:38] tremendous disruptions. There's lots of
[14:40] stories today uh spreading on the
[14:42] fertilizer side. So, I just created this
[14:45] just you guys can look at it on your
[14:46] own.
[14:47] But, you've got natural gas,
[14:49] petrochemicals, plastic polymers. You've
[14:51] got fertilizer, all things for food.
[14:53] Oil's in everything. So, when oil goes
[14:56] higher,
[14:57] everything tends to go and the inflation
[14:59] numbers tend to surprise on the upside
[15:01] because it's really hard to model when
[15:03] all of these kind of moves happen. Now,
[15:06] that last thing I showed you was built
[15:07] in Nano Banana to give you guys an idea.
[15:10] This I just want to show you cuz I'm
[15:11] using more and more OpenClaw, but I also
[15:13] started to use Perplexity computer and I
[15:16] asked it to build an ecosystem with live
[15:18] pricing for all of the different
[15:20] components in terms of the movement
[15:22] since the war
[15:23] and I created a dashboard with crude,
[15:26] natural gas, petrochemicals,
[15:28] plastics, fertilizer, industrial gases,
[15:30] construction, transport, freight,
[15:32] specialty chemicals, food and
[15:33] consumption. You can go through the
[15:35] numbers, but what you can see here is
[15:36] the massive movements that have happened
[15:39] already. You haven't seen power and
[15:41] electricity go, but those will be going
[15:42] as well once we start to see some
[15:44] movement there. Natural gas prices in
[15:46] the US, as you can see,
[15:49] I didn't give you the heavy hug. That's
[15:50] negative. So, we've seen natural gas not
[15:52] move in the US. This is another page
[15:54] that goes in. This is if I click into
[15:55] petrochemicals, I can see all of the
[15:57] components that go through. This was
[15:59] built by me in Perplexity Computer on
[16:03] Perplexity Max.
[16:05] You'd be shocked at what you can build
[16:07] in a matter of minutes now. It just
[16:09] shows again the agentic world and every
[16:10] every one of you who doesn't believe
[16:12] software is going to be completely
[16:13] disrupted. I've created so many things
[16:15] with Perplexity Computer, not only for
[16:17] this video, but just for me to monitor
[16:19] things in a way that Bloomberg can't do.
[16:22] So, yes, Bloomberg has messaging
[16:24] function. Yes, Bloomberg has some other
[16:25] functions, but if you guys don't think
[16:27] that $30,000 versus $200 a month, I
[16:31] again,
[16:33] $2,400 a year for Perplexity, I pay $600
[16:36] a year for the data providers,
[16:38] for great data providers. I I
[16:41] You have an ability to do everything in
[16:43] AI and it builds it for you without you
[16:44] having to do it while you can do other
[16:46] things. So, yes, the software side
[16:48] continues to roll higher and higher.
[16:50] Here's jet fuel prices in Los Angeles.
[16:55] This is jet fuel in Singapore.
[16:58] Diesel prices have gone up far more than
[17:01] gas. So, diesel has gone from 350 up to
[17:03] 520.
[17:05] This has a huge impact cuz everything
[17:07] that's trucked uses diesel.
[17:10] Here is urea,
[17:12] the orange line, under over the top of
[17:15] uh
[17:16] Sorry, urea's uh down here or right
[17:19] here. Brent crude is right here. This is
[17:22] for fertilizers. This happens exactly at
[17:25] springtime for when a lot of the
[17:26] plantings are going to go on.
[17:28] Major event that occurred this week was
[17:30] in Ras Laffan,
[17:32] which was shot down with drone strikes.
[17:34] This happened on Thursday. In my
[17:36] opinion, this is where things started to
[17:37] spread. You saw semiconductors were weak
[17:39] on Friday. I'm going to go through the
[17:41] reasons why. That was on March 20th.
[17:44] Here's what was mentioned on March 7th
[17:46] when Ras Laffan had an outage. So, now
[17:49] it was hit to the point where most
[17:51] people are saying that it's offline for
[17:53] the next
[17:54] 6 weeks to months.
[17:57] Offline.
[17:58] Um it is a huge part of helium. In fact,
[18:03] the majority of the helium received for
[18:05] semiconductors in Korea and Taiwan
[18:08] comes from Ras Laffan. They Qatar is uh
[18:13] with the US, the two largest in the
[18:15] world, and we're now getting a problem
[18:17] there
[18:18] that has semiconductor implications. One
[18:20] attack in the Middle East just put the
[18:21] entire tech industry on life support by
[18:24] bombing the helium plant. Don't think it
[18:26] wasn't intentional. This is where we're
[18:28] at right now. So, these disruptions that
[18:31] everyone looks at and say, "Well, oil's
[18:33] up, it'll come back down."
[18:35] We're at a different stage now. Um it
[18:37] isn't just about the Strait of Hormuz.
[18:39] It's about the damage that's being done
[18:41] and the fact that the energy facilities
[18:43] this week were hit. So, again, you're in
[18:46] trouble with helium reserve running down
[18:49] for years in the United States. If you
[18:50] guys have tried to buy balloons for
[18:54] birthdays over the course of the last
[18:55] few years, occasionally you'll run into
[18:56] a point where helium isn't there, but
[18:59] the prices of helium-filled balloons
[19:01] have gone up as well before this. So,
[19:03] I'm just saying that we're at a stage
[19:05] right now where you're having a bigger
[19:07] estimate than or bigger implication than
[19:09] people thought. Uh in terms of the oil
[19:13] markets on Friday, the IEA said it could
[19:16] take 6 months to restore oil flows from
[19:18] the Gulf. Politician markets are
[19:20] underestimating the scale of the
[19:21] disruption.
[19:23] Again, and it's because of the damage
[19:25] that's now being done. If Hormuz stays
[19:26] closed another 3 to 4 weeks, it all
[19:28] begins to crumble into an already
[19:29] teetering global sovereign debt problem
[19:31] and consumer credit problem. I didn't
[19:33] show this because Luke Gromen has any
[19:36] sense on here, even though he says based
[19:38] on it's highly likely it will remain
[19:39] closed for at least another 3 or 4
[19:41] weeks. Nobody knows. So, I think if this
[19:44] had been 2 weeks ago and we did
[19:46] something for 1 week and we didn't have
[19:48] all the damage in the energy facilities
[19:50] and we didn't have now 3 4 weeks of 20
[19:54] million barrels per day being impacted
[19:57] and even if 10 million are getting
[19:58] through pipelines or something, you're
[20:00] still dealing with enormous amounts that
[20:02] is not making it.
[20:04] And even if it goes back online, the
[20:05] question is now how much damage has been
[20:07] done. So, everyone was focused on the
[20:10] Strait. That's where most of the focus
[20:11] is. I think at this point the charts are
[20:13] telling you that it's going to take a
[20:15] while to unwind this.
[20:17] I don't think we will uh make a bottom
[20:20] in stocks that is sustainable until
[20:22] headline CPI year-over-year peaks.
[20:25] That's what happened in the 1970s.
[20:27] I think that allows for everybody to be
[20:29] wrong on the strategy side, the
[20:31] recession fears to go higher and
[20:32] everything. So, if we get 2 months of
[20:36] high monthly data, we'll probably make
[20:38] the peak in CPI
[20:40] when we get the data for April, which
[20:42] will be in May. And I'm just saying that
[20:45] based on the fact that I don't think
[20:46] sentiment regarding uh what has happened
[20:49] for inflation is where it should be
[20:51] relative this. The market may have gone
[20:53] down, but I am not seeing people be
[20:55] nearly as bearish and as silly as they
[20:57] were last year during the tariff
[20:58] scenario. This one is real. The tariff
[21:01] situation was all speculation based on
[21:04] something that happened 100 years ago.
[21:07] The fact that people are not as bearish
[21:08] right now on the strategy and analyst
[21:10] side to me is a reflection of PTSD.
[21:16] Uh
[21:17] on Friday, we got this scary situation.
[21:19] Obviously, if this happens, not a good
[21:22] thing uh in terms of for the markets to
[21:24] have ground troops there.
[21:26] Uh there were stories about an oil
[21:27] export ban because of the fact that
[21:29] you've got WTI uh being lower and
[21:32] natural gas in the US is not moving at
[21:34] this point. The The pain that started to
[21:37] filter into the market is the dilemma
[21:40] that now the Fed is as they went through
[21:42] and basically said they have no idea
[21:44] what to do. Uncertainty has gone into
[21:46] the rate picture, whether it's rates
[21:48] going higher, whether it's they should
[21:51] cut. Uh I don't know what else to say.
[21:53] Here is the Cleveland Fed
[21:56] uh inflation forecast for March. Uh just
[22:00] remember that when, you know, the way
[22:02] that the headline CPI, the way that I
[22:04] used to uh
[22:05] forecast it was taking the data from
[22:08] about the prior month on the 15th to the
[22:10] middle of the month on the 15th. That
[22:12] was obviously a long time ago in gas at
[22:14] the pump, but they're forecasting now a
[22:16] headline CPI for the March data of .62,
[22:19] PCE at .47. It's putting all of the
[22:23] quarterly annualized rates well above
[22:25] 3%. Headline CPI only here, but by the
[22:28] time we get through the March data uh
[22:31] for April, you're going to be much
[22:32] higher. Uh I I just think the
[22:34] uncertainty as I'm here that people are
[22:37] writing about has never been greater for
[22:39] the Fed. And as I'll remind you, Kevin
[22:41] Warsh is due to come uh into office in
[22:45] the middle of May, which will be around
[22:47] the time that I believe that the worst
[22:48] of the inflation stuff would be there
[22:50] assuming
[22:52] that the straight is
[22:54] flowing again uh sometime over the next
[22:56] couple weeks.
[22:58] Uh
[22:59] Here's the turbulence model. Uh it was a
[23:01] quiet week for the turbulence model, but
[23:03] that's because a couple things have
[23:05] happened.
[23:06] Uh the vol's already picked up, but the
[23:07] other thing is the correlations are
[23:09] picking up, so it's less about
[23:11] disruption of of gyrations going the
[23:13] other direction.
[23:15] Uh but also, it was not a volatile week
[23:17] when you go through it relative to what
[23:19] you'd expect. I just wanted to highlight
[23:20] this Noah Weissberger
[23:23] who uh I used to follow at Goldman. Uh
[23:26] he put this point out just to show how
[23:28] much the S&P right now relative to other
[23:31] times. I think this is the 1990s. I
[23:33] think this is during 2022. How the S&P
[23:35] responded
[23:36] uh to the VIX moving during those
[23:38] scenarios. And this time, again, as I
[23:40] said it, the VIX has got more volatile.
[23:43] I think that's because people again, and
[23:45] I'll show some data to support it, uh
[23:47] everyone's been talking about how people
[23:48] have been bearish and putting on hedges.
[23:50] In my opinion, based on the data I've
[23:52] seen, they've just increased gross. So,
[23:54] their shorts have gone up, but that was
[23:56] to neutralize a net long position that
[23:58] they hadn't gotten out of gotten gotten
[24:00] out of. And the reason is the bullish
[24:02] side of the trade, whether it's power
[24:04] names, energy names, uh Caterpillar, all
[24:07] the AI trades, semis, they're all still
[24:09] up for the year. There hasn't been that
[24:11] momentum kind of shockwave because we
[24:13] haven't changed our opinion on the
[24:14] economy, and we haven't changed our
[24:16] opinion on earnings.
[24:18] Um
[24:19] and I just again wanted to put this out
[24:21] cuz this was put out in November. I
[24:22] completely uh agree. Trump makes it
[24:25] impossible to hold the short. I think
[24:27] the fact that that was the theme last
[24:28] year and people remained bearish the
[24:30] entire move higher,
[24:31] they're less likely to do it. I posted
[24:33] an X this week one thing, and it was
[24:35] this.
[24:36] Um
[24:37] I'm shocked. I showed this because
[24:40] April 24th is 3 weeks after
[24:43] uh liberation day. And this is what the
[24:45] strategist did last year across the
[24:47] board. Uh I should leave UBS alone and
[24:50] Morgan Stanley alone, Wells Fargo, but
[24:51] everyone else just took their numbers
[24:53] down violently as of April 24th.
[24:58] That was 3 weeks after liberation. We
[25:00] are now 3 weeks after the beginning of
[25:01] the war.
[25:03] Um a year ago in September, UBS gives
[25:05] America recession check up to season 93%
[25:08] probability from the hard data. I think
[25:10] people forget this. I mean, I started
[25:12] writing about PMIs going higher in
[25:13] August. I was attacked by most people
[25:16] that that follow cyclical saying there
[25:19] was nothing showing up. How could you
[25:20] even think this was going to happen? And
[25:22] the recession fears were still alive and
[25:24] well in September. That's how bad the
[25:26] sentiment was last year. In terms of the
[25:28] earnings, you have to go back and see
[25:30] what earnings beat the estimates as we
[25:32] went through. This was from August of
[25:34] last year. Everyone was talking
[25:36] inflation. Everyone was talking
[25:37] recession.
[25:39] This is where we are now. No one's moved
[25:40] a single thing from December price
[25:42] target. Like the S&P is now trading
[25:44] lower and no one's moved a single thing.
[25:46] This is opposite of last year and there
[25:49] has to be a reason. UBS put this out.
[25:51] Confident US stock sees S&P 500 hitting
[25:54] 7700.
[25:56] I Again, I I I don't know what to say to
[26:00] to people at this point other than I
[26:02] think the difference is and the
[26:03] important thing. This is what happened
[26:05] last year with the analyst revisions. If
[26:07] analyst revisions don't change, which I
[26:09] certainly feel will go negative as they
[26:11] have during 2022, during COVID, every
[26:15] time. This was during the energy problem
[26:16] what went down. I fully expect it'll go
[26:18] down, but I don't think the S&P
[26:21] is going to fall in any meaningful way
[26:23] if revisions sit up here. I think people
[26:25] are going to keep trying to bottom cuz
[26:26] they haven't changed their views and why
[26:28] should they change their views if the
[26:29] leaders of telling them what's coming
[26:31] are not changing their views. Um again,
[26:34] this is nothing against Ryan Detrick.
[26:36] It's just everyone's trying to figure a
[26:38] way to measure sentiment and they're
[26:40] just picking the data. I could find 10
[26:41] other things as I showed the other day
[26:44] that I'm looking at. These to me are not
[26:46] a good indication of anything. Yes, they
[26:49] can work at times.
[26:51] We haven't seen capitulation and again,
[26:53] if I'm right about something, the
[26:55] sentiment that is still way off is the
[26:57] analyst and the strategist. They should
[26:59] be more negative than they are, at least
[27:00] be warning about a correction. I'll give
[27:03] Mike Wilson credit at Morgan Stanley,
[27:04] who I used to work with because he did
[27:06] call
[27:07] for a correction of 5 to 10%. Well,
[27:09] we're kind of
[27:12] 2/3 of the way through there at this
[27:13] point. So, give Mike some credit on at
[27:15] least being one of the strategists that
[27:17] said there's going to be a pullback for
[27:18] the things I highlighted at the
[27:19] beginning. We do have a bigger
[27:21] correction that's happening under the
[27:23] hood of the market because you have a
[27:25] lot of names that are down a lot. You
[27:26] obviously have that from what I showed
[27:27] with financials.
[27:29] This got a lot of things. Long-only
[27:31] investors, again,
[27:33] I don't know how to do this. People like
[27:35] to cherry-pick things. I mean, this
[27:36] wasn't any kind of day last year in
[27:38] August of '25 or September of '25 other
[27:41] than I don't know, maybe it was another
[27:44] week of expiration.
[27:48] When you get into this, bearish
[27:49] sentiment just crossed 50% for the first
[27:51] time in over 6 months. You know what AI
[27:53] is really good at, guys? I can do it all
[27:54] day long. Hey, show me a backtest that
[27:56] gives me a positive number for at least
[27:58] 70-80% of the days looking forward. I
[28:00] don't care what you do. Let's just
[28:02] figure it out. Random periods over 6
[28:04] months.
[28:05] Again, I don't I don't know what to say.
[28:08] I do like it when sentiment trades at a
[28:10] negative number for a long period of
[28:12] time. I think that matters as you're
[28:13] going higher. But as you're going lower,
[28:15] when you see something like this, okay,
[28:17] here are the returns. Just look at this.
[28:19] Does this look like panic? This is where
[28:22] we were last year. This is where we were
[28:23] in 2022 and we sat down at these levels.
[28:26] This is not panic levels, guys.
[28:28] We haven't We haven't gotten rid of the
[28:30] extreme readings at this. And again,
[28:32] sees potential for an extreme rally.
[28:34] This was March 11th. I showed a bunch of
[28:37] these things, but hedge fund short
[28:38] exposure is at the highest since 2022.
[28:42] Okay, and gross exposure is near a
[28:45] record. Well, that means that if nets go
[28:47] down, guys, and you add more shorts,
[28:49] which I expected that people did because
[28:51] of the turbulence model,
[28:53] they didn't delever, they just
[28:55] hedged up their net exposure. That
[28:56] doesn't mean they're bearish at this
[28:58] point. It just means they're protecting
[29:00] or trying to protect against further
[29:02] losses.
[29:03] Um just
[29:04] an important thing for the flow under
[29:06] the market because we've been having
[29:07] this long gamma environment and now the
[29:10] move index has moved higher. I would
[29:11] expect that if we don't get a resolution
[29:13] this week, now that we're through
[29:15] expiration, but also the blackout period
[29:18] for corporate buybacks is now coming
[29:20] into play, which will go in through the
[29:22] end of April. That usually leads to more
[29:24] volatility if you're in a down tape
[29:26] because what the buyback desks do is if
[29:29] the market's going to open down, their
[29:31] orders come flying in the door.
[29:33] Um
[29:34] when I worked at Morgan Stanley, unless
[29:36] something has changed, they just tend to
[29:37] be more active on down days,
[29:39] which adds this long gamma component, uh
[29:42] which prevents markets from having high
[29:44] realized vol. Well, we've had very low
[29:46] realized vol.
[29:47] The credit situation has not ended. So,
[29:50] Blackstone went through its first
[29:53] Bcred loss since 2022. They released
[29:56] this late on Friday.
[29:59] Um I just wanted to show that
[30:01] this is their portfolio. So, this is the
[30:03] Bcred portfolio. If you just go to the
[30:05] website and go look at the um the
[30:07] portfolio shaping,
[30:09] 26% is software and 11% is professional
[30:12] services. If you ask me what the two
[30:14] highest things to get be disrupted are
[30:16] by AI, here we go. Well, that's 37% of
[30:19] the portfolio, and then you start
[30:21] getting in into eights and eights and
[30:23] sevens and small amounts. Uh there's no
[30:26] way that these books reflect the risk
[30:29] that's coming in the next few years from
[30:31] the software side, which is why the
[30:32] bonds and the equities, the public
[30:34] equities, are trading down.
[30:36] Uh you had more uh stories this week
[30:38] from another one. Stone Ridge pays 11%
[30:41] of what they were requested to pay out.
[30:43] JP Morgan halted Qualtrics debt on
[30:46] software pain. So, this is
[30:49] supposed to be a very good company
[30:51] uh and they still basically went through
[30:53] as Morgan Stanley saying that default
[30:55] rates to reach 8%. Again, I'm bringing
[30:57] this up because this was not a story a
[30:59] year ago. Um
[31:02] Cliffwater
[31:03] lot more stories, guys. Um I think this
[31:06] is going to remain a big story,
[31:07] especially if you guys get the time to
[31:09] read this. So, this was in X. I highly
[31:12] recommend uh either reading it or going
[31:15] through and uploading it to an LLM. And
[31:18] the reason is because it
[31:20] I mean, this is an attempt to be a
[31:21] Michael Burry situation with mortgages
[31:23] and to go through the issue.
[31:25] This, I think, goes through the issue in
[31:26] a pretty big way. Um everyone who's
[31:29] saying there's not many losses and this
[31:32] is all fine.
[31:33] I think this thing does a good job or uh
[31:36] a job of at least giving you the side of
[31:39] how bad the situation is very likely.
[31:42] But, in particular, with oil raging, I
[31:44] think people should be paying attention
[31:46] more to private credit. Now, Boaz
[31:48] Weinstein did an interview. If you
[31:50] prepare uh prefer to talk to or listen
[31:53] to someone on a podcast, you can do
[31:55] that.
[31:56] I think these were the interesting
[31:57] points. So, the major theme is
[31:58] reflexivity.
[32:00] He gets into the fact that, again, what
[32:02] I said before, which is
[32:04] especially in this day of SVB and
[32:06] digital economy, when more people hear
[32:08] the problem and they're wondering why
[32:09] they got into this in the first thing,
[32:11] even though they are
[32:13] legally, they took the risk. Uh
[32:17] I I've just seen this act before.
[32:19] They're just going to keep asking for
[32:20] their money and he basically said, he
[32:23] ties this stress to what he calls vola-
[32:25] volatility laundering. Um what better
[32:28] way to have your money in something that
[32:29] doesn't have mark-to-market risk. And I
[32:31] think that's the issue that's popping
[32:33] up. So, what I did was I uploaded uh
[32:36] that transcript and the Nick Name Eth
[32:40] uh, open letter to Bessant, and I just
[32:43] went through and you guys can read this
[32:45] for yourself, but
[32:47] the structure can create a series of
[32:48] reflexive market event and a major
[32:50] opportunity once re-priced. The
[32:51] structure itself is fundamentally
[32:53] corrupted and should be reformed. This
[32:54] slowly poisons the financial system.
[32:57] Basically, when you go through it,
[32:58] there's a lot of overlap in the two
[32:59] stories.
[33:01] Um,
[33:03] I think this is something important for
[33:04] people to at least be thinking about as
[33:06] opposed to just thinking, "Well, Trump's
[33:07] going to pull out of the Strait of
[33:08] Hormuz. Everything's going to get
[33:10] flowing again. Oil's going to collapse.
[33:12] And there was no issues before, even
[33:13] though financials were the worst
[33:14] performing." These credit issues never
[33:16] end on their own, but in particular, not
[33:19] at a time that the S&P is breaking down,
[33:21] and definitely not a time when you have
[33:22] structural issues like
[33:25] AI,
[33:27] which is a disruptive force that again
[33:28] has to be dealt with. Life insurers will
[33:31] be part of this. They hold more private
[33:32] credit than ever, and again,
[33:35] most of the private equity companies
[33:36] have captive insurers. I won't go
[33:38] through this again. If you listen to
[33:40] Steve Eisman, you know. The reason I'm
[33:42] bringing this up is this has to be in
[33:44] your mind of could this get much worse?
[33:47] Yes, it could. I think it will get to
[33:50] the point that a liquidity facility
[33:52] needs to deal with this to slow down and
[33:55] give liquidity to people. That's what
[33:56] the liquidity facilities were for.
[33:58] That's what SVB was. Give us your loans,
[34:01] you can work your way out of them. In
[34:03] this case, it was actually mainly
[34:05] treasuries. We never got to the CRE, but
[34:07] that was going to be the next facility,
[34:09] and they could create that one again if
[34:11] they wanted to. Um, I won't go through
[34:14] all the annuity side, but you just have
[34:15] to realize that eventually, they're
[34:18] going to have to deal with something if
[34:19] these loans continue to get marked down,
[34:21] because a lot of annuities are tied to
[34:23] this, and you've had
[34:25] a a an insurance company tied to
[34:27] annuities already be in trouble. PHL
[34:30] closed down in Connecticut. Go read the
[34:32] stories on your own. Here is private
[34:35] equity
[34:36] uh the white line
[34:37] relative to the insurance. The insurance
[34:39] names are just breaking down. That is
[34:42] what I would start watching.
[34:44] Uh we finally went through. And like I
[34:45] said, I've shown this before. Don't get
[34:48] sick of it.
[34:49] This is never getting better. So, AI's
[34:53] progress will continue to disrupt
[34:55] businesses. They will never be able to
[34:58] get away from this. It is the speed of
[35:00] it and it is the structural nature of
[35:02] it. The entire capital structure of
[35:05] debt, if it's in companies that are
[35:06] going to get hit by this in the next 5
[35:08] years,
[35:09] that risk has to be built into not only
[35:11] equity multiples, which has happened in
[35:13] the public markets, it needs to be built
[35:14] into the loan and debt markets of these
[35:16] companies that are private. I would say
[35:19] that those markets should be worse than
[35:20] the public companies, which obviously
[35:22] have the ability of raising capital
[35:24] through multiple channels.
[35:26] Um on the compute side,
[35:29] uh Dylan Patel is a must-listen to, in
[35:31] my opinion, for everyone who's got
[35:33] positions in semis. He had an interview
[35:35] with his roommate, Dwarkesh Patel,
[35:38] uh and it was great. 2 hours, um you
[35:41] know, a lot of good highlights.
[35:43] Uh here's what you got.
[35:46] Ins- He really I think the most
[35:48] important thing as someone who has
[35:50] talked about power for a long time, but
[35:52] you've probably heard me kind of
[35:53] decrease it as a massive problem. That
[35:56] has been because I've seen enough things
[35:58] with behind the meter. I've seen enough
[36:00] things uh with optimization of the grid
[36:03] that could allow us to at least be able
[36:06] to survive through this in kind of a
[36:08] cut-and-paste way. The real bottleneck,
[36:11] I agree with, is semiconductors. And he
[36:12] makes the point, we just don't have
[36:14] enough compute. And because TSMC is
[36:17] really the gatekeeper of this and they
[36:18] refuse to let this go too fast,
[36:21] you're just you have a uh a natural
[36:24] throttle on this to keep it from going
[36:25] crazy. He does talk about ASML and the
[36:27] fact that EUV machines are a dramatic
[36:30] constraint. The memory crisis is
[36:33] worsening.
[36:34] He does, like I said, minimize power.
[36:38] This is all about compute, the scarcest
[36:40] economic resource, which means you need
[36:42] to be long it.
[36:43] I think we're going to be in a period
[36:45] because of the helium situation and
[36:48] because of it being over-owned and
[36:50] having great years, that I do think any
[36:52] move lower in the S&P will probably see
[36:54] the
[36:55] memory names
[36:57] trade down and underperform the market
[36:59] in a beta perspective,
[37:01] but I still think that
[37:03] there won't be cancellation of anything.
[37:05] So, I'm going to be remain a memory
[37:06] bull, and like I've said for the last, I
[37:08] guess it's 4 weeks now, I'm long VIX. I
[37:11] actually have some short positions down
[37:13] in the semis as well, just expecting
[37:14] there to be more of a correction.
[37:17] GPU availability, Warren Buffett does a
[37:19] great job on this. I mean, it's insane,
[37:21] basically. There is just insane demand
[37:24] still. There are no signs of even
[37:26] getting close. Jensen Huang has done a
[37:29] lot of presentations and and interviews.
[37:33] He spoke at GTC for, I believe, 3 hours.
[37:37] And one of the things he brought up,
[37:38] which is this whole thing of tokenomics,
[37:40] and I think more and more
[37:42] people have to get on the framework that
[37:44] that's where we are. The amount of
[37:45] tokens we need is far greater than we
[37:47] realized
[37:48] 3 months ago.
[37:49] And the demand is here faster than we
[37:51] expected, and that's because of this.
[37:53] He goes through the coding agents auto
[37:57] research and the loopy era of AI. This
[37:59] is Andre Karpathy.
[38:01] If you want to go read it, great.
[38:02] Remember, for those of you who are
[38:04] subscribers, I do put in on the website
[38:07] each Sunday
[38:09] all of the podcasts that I reference on
[38:11] here with the link and a summary, that
[38:13] way you don't have to listen to it if
[38:14] you want. If you want to get more
[38:15] details, but I also
[38:17] show you the timestamps and everything.
[38:18] So, if you want to go listen to it on
[38:20] your own, you don't have to waste the
[38:21] time. The main point is here and you
[38:22] have to remember this, it was October
[38:24] that he did the interview with Dwarkesh
[38:26] Patel and as I've mentioned before, I
[38:28] had people calling me up from all over
[38:29] saying, "He just basically said agents
[38:31] won't be impacting us for another
[38:32] decade, that it's all hype."
[38:35] And basically he talks about it, um the
[38:37] December flip sudden capability jump.
[38:41] I had asked, "Does he mention his
[38:42] surprise on the speed in there?"
[38:44] Uh it's not incremental process, it's a
[38:46] step function change.
[38:48] He think most people haven't realized
[38:50] how big the shift is. I can tell you in
[38:52] the conversations I've had where people
[38:53] have tried to minimize Open Claw by
[38:55] asking me what I'm using it for and
[38:57] judging by what I'm using it for at this
[38:59] point is and as I've shown you, I'm
[39:01] using it more and more every day. I'm
[39:02] not only using Open Claw and I'm going
[39:04] to spend all of tomorrow on this because
[39:07] now that my new machine is here that I
[39:08] can upload another one, I really do need
[39:11] to start building more and more on this
[39:13] for my workflows, but now I have Co-work
[39:15] dispatch so I can do that and I have
[39:18] Perplexity computer which as I've shown
[39:20] you is amazing. It's very easy to build
[39:23] things in it. So, if you guys haven't
[39:24] thought of it, pay for Perplexity Max at
[39:27] $200 and just start with that and just
[39:30] build stuff and just literally ask it.
[39:32] You may have to give it an API key, may
[39:34] have to pay for your own data. It will
[39:35] use Yahoo Finance, but like to give you
[39:38] an idea,
[39:40] I have to do things using FMP data which
[39:42] I pay $600 a year that has tremendous
[39:44] amounts of data going back 30 years,
[39:47] but I also have multiple agents and when
[39:49] I build something I say, "Okay, what I
[39:51] want you to do is this. I want an agent
[39:52] to do X, I want an agent to do Y and I
[39:55] want another agent to basically check
[39:57] all of the work on the data to make sure
[39:59] that the data entries are fine."
[40:01] That is the way that you basically, if
[40:03] you've managed people, get to send these
[40:05] things off to do things and they do them
[40:06] while you're working on something else
[40:08] and then you come back and it's just
[40:09] done.
[40:10] So, he talks about this whole thing and
[40:12] about how everything has changed. His
[40:15] auto research side, he talks about that.
[40:17] Again, this is all about getting the
[40:19] models to be improving without needing
[40:22] new improvements, and that's where we're
[40:24] at a scary time. We don't know what it
[40:26] means to have a billion agents running
[40:29] out to solve problems
[40:31] and certainly to create software. So, he
[40:33] gets into that as well.
[40:35] The agentic layer is ahead of where most
[40:37] capital is positioned. And again, I'll
[40:39] say this, this statement is really
[40:42] important.
[40:44] And it's not just important for what
[40:45] he's saying. Markets are still focused
[40:47] on models, GPUs, and data centers. He's
[40:49] pointing to orchestration, memory,
[40:50] multi-agent systems, autonomy loops.
[40:54] This is critical. This whole thing is
[40:56] critical for enterprises and adoption.
[40:58] That's why I spend my time on on talking
[41:00] to people about that. Um OpenClaw at the
[41:04] GTC,
[41:06] this is how important it is. He not only
[41:07] says it's the most important software
[41:10] since Linux and the fastest growing,
[41:13] they've got their own Nemo Claw for
[41:15] OpenClaw. Uh he talked a lot about the
[41:18] different ways they're going to support
[41:19] it. Mark Andreessen OpenClaw and Pi
[41:21] together are in the the top 10 of
[41:23] all-time software breakthroughs. Again,
[41:26] people are minimizing OpenClaw and
[41:27] thinking it's some kind of fad and it'll
[41:29] be gone. Uh it's the gateway to the
[41:31] agentic world, which means going from a
[41:33] uh interactive bot of just giving you
[41:35] answers to a task-oriented situation
[41:38] where it can go out and build things for
[41:39] you,
[41:40] and you can talk to it on your phone.
[41:43] Uh it went from a weekend WhatsApp hack
[41:45] to 316,000.
[41:47] You need to start have started using
[41:49] OpenClaw yesterday.
[41:51] I completely agree. I don't think you
[41:53] can possibly understand what's going on
[41:55] without using that. Even though Claude
[41:56] Code will build things, um there's a
[41:59] very big different feel with OpenClaw
[42:00] and Perplexity Computer and with Code
[42:02] Work. Uh
[42:04] I just wanted to highlight the token
[42:05] usage. OpenClaw 4.5 trillion,
[42:09] Claude 586 billion. Um
[42:13] again, when you go through the numbers
[42:14] on things of all of this and you go
[42:16] through the global rankings,
[42:18] uh
[42:19] you just see that they are bigger than
[42:21] all of the other ones combined, which
[42:23] includes Claude Code in terms of the top
[42:25] 20.
[42:28] Uh
[42:29] remember Peter Seinberger joined Open uh
[42:32] AI
[42:33] well, now it's uh more like 6 weeks ago.
[42:36] And again, they're bringing more stuff
[42:38] out in terms of being able to use.
[42:41] Again, you're going to see more open
[42:42] claw from them. This is dispatch, which
[42:44] came out this week using your phone with
[42:47] it. Alibaba launches an AI platform for
[42:49] enterprises as agent craze sweeps.
[42:52] Perplexity takes its AI computer agent
[42:54] to the enterprise. Again, like I said,
[42:56] Perplexity is a great easy way to do it
[42:59] and it is phenomenal to use. Uh
[43:01] Perplexity computer is better than open
[43:03] claw. I'm not really sure about that,
[43:05] but I am telling you that it is easier
[43:07] to use,
[43:08] uh easier to set up because it's already
[43:10] on your machine if you use Perplexity.
[43:11] So, just go pay the max and computer
[43:13] will be on there. Again, I built this
[43:15] with it. Uh and again, I just wanted to
[43:17] show you one thing. When I first ran it,
[43:20] it gave me uh the wrong thing for
[43:23] natural gas. It said it Henry Hub prices
[43:25] were up significantly. So, I went and
[43:28] said, I looked at it, seems like the
[43:29] Henry Hub LNG price move is wrong. Can
[43:32] you rebuild it this time and have a
[43:33] second agent be in charge of checking
[43:35] the percent moves and another data
[43:37] source to make sure that they are
[43:38] correct. When it came back the next
[43:40] time, everything was correct.
[43:43] Um
[43:45] this was a good article. Again, I think
[43:48] I'm trying to give you guys ones that
[43:49] are not like the Satrini one, which I
[43:51] didn't agree with. I think the ones
[43:53] where are being rational about, and this
[43:55] is from the co-founder at Whisper,
[43:58] about what's happening and how fast it's
[44:00] going and how shocking it's been to
[44:02] leaders at at tech companies. And I just
[44:05] want to show this cuz this is really the
[44:07] way to think about what has happened
[44:09] here. You've got a human organization,
[44:10] so think of this as Morgan Stanley or
[44:12] Goldman Sachs. You've got all these
[44:14] layers here.
[44:16] The AI agent hierarchy. So, I'm above it
[44:20] telling it to do this stuff.
[44:23] I'm giving the orchestrator, and then
[44:25] I'm having them send out and do all
[44:28] these things. So, it's the same as says
[44:31] organizational logic. This is from the
[44:33] paper that he showed.
[44:35] It's just different in terms of who's
[44:36] doing what and what's going on. The
[44:38] reason I wanted to bring this up is
[44:41] how Palantir's approach solves this
[44:43] problem. And the problem we're solving
[44:46] is the other side. So, the question is
[44:49] how can a big organization, like a
[44:51] Morgan Stanley or Goldman Sachs, go to
[44:53] this? It is not easy, and that's the
[44:56] thing I've been saying. Especially when
[44:57] the progress is going so fast that any
[44:59] decision you make today could be
[45:01] obsolete in terms of the direction
[45:03] you've gone in 3 months. That's how fast
[45:05] it's moving. That's why exponential is
[45:07] dramatic. That's why Palantir, which
[45:10] I'll keep saying for people who want to
[45:11] argue with me, not only are their
[45:13] numbers growing at this point, but you
[45:15] have to understand what they're able to
[45:16] do. And I understand that people have
[45:18] very little knowledge of what I just
[45:20] showed you. But the reality is that's
[45:23] the job that we have is to learn and not
[45:25] start with I'd never buy a stock with a
[45:27] 100 multiple.
[45:29] Palantir has an advantage in this, and
[45:31] what I wanted to show was just how their
[45:33] system is somewhat similar to the AI
[45:36] agent hierarchy in the way they've set
[45:38] things up. This alignment ensures that
[45:40] as you move to an economy of trillions
[45:42] of agents, your AI operations are
[45:43] secure, auditable, and fundamentally
[45:45] grounded in the reality of your business
[45:46] data.
[45:48] If there's CEOs of companies going
[45:50] through it,
[45:51] look into Palantir. It may be more money
[45:54] at the startup side of going through it,
[45:56] but if your if your firm is filled with
[45:58] unstructured data across silos, if
[46:01] you're having trouble getting people
[46:02] move into this to where
[46:04] you're going to be doing deploying this
[46:06] one agent at a time,
[46:08] I don't think you have that kind of
[46:09] time. I just don't. I think this is
[46:11] going to be a bigger problem than people
[46:13] realize
[46:14] in terms of the adoption and it's one of
[46:15] my big negatives for this year and one
[46:17] of the things that I'm hyper-focused on
[46:18] to keep monitoring.
[46:20] Andre Karpathy also put this out. You
[46:23] can go to it karpathy.ai/jobs
[46:26] to go through and basically from his
[46:28] perspective
[46:30] which jobs are most likely to be
[46:32] impacted. You can go see it and go
[46:34] through the numbers. Declining jobs,
[46:36] very quickly 34 million. There's 140
[46:39] some odd million people who work, 150
[46:40] million people in the country who work.
[46:43] So again, and this is in the short term.
[46:46] Jobs that can be replaced or should be
[46:48] replaced is probably more it.
[46:50] The friction will prevent it from
[46:52] happening as fast as it otherwise
[46:53] should, but again, this is getting into
[46:55] more the job disruption side.
[46:58] Uh
[46:59] You can tell the different the
[47:01] difficulties they're having on the
[47:02] adoption side by making these
[47:05] arrangements with the PE firms. Again,
[47:08] both OpenAI and Anthropic. Uh they're
[47:11] going to consultants to fight their
[47:12] battle over the enterprise market. So
[47:14] they're not only trying to win it, but
[47:16] they're trying to make sure that they
[47:17] get more adoption. So here's where we
[47:18] are for the rest of 2026. Unlike last
[47:21] year where when tariff
[47:23] the tariff situation was
[47:25] to me silly in the fact that the AI
[47:27] trade is going to happen. AI is going to
[47:28] keep happening. Um but the speed is
[47:31] going to continue to grow rapidly. We're
[47:33] going to start getting more and more
[47:34] things happening that are going to be
[47:35] disruptive. So all long duration assets
[47:38] on businesses that are not going to be
[47:39] able to compete in AI are going to be in
[47:42] trouble. So the asset uncertainty for
[47:44] long duration assets are going to
[47:45] remain. The commodity boom. Take
[47:47] advantage of the weakness you're going
[47:49] to see in silver,
[47:51] um in copper during this period like
[47:53] this. All of these things will be
[47:54] necessary for the massive build out. The
[47:57] adoption gap AI progress is faster than
[47:59] enterprise AI adoption. These are themes
[48:01] to make money on this year. Uh we're in
[48:03] a period right now where it's difficult
[48:04] to make money because everything is
[48:06] going down. But hopefully you got off to
[48:08] a good start in the year with a lot of
[48:09] the names that we've talked about um the
[48:11] key sectors. I still think those things
[48:13] are going to have a great year. But
[48:15] again, in a commodity bull market as I
[48:16] showed with the webs uh the subscriber
[48:19] webinar this week,
[48:21] you're going to have 50% moves in these
[48:23] names and then you're going to have 30%
[48:25] corrections. This is not software stocks
[48:27] of the past where they just continue to
[48:29] grow regardless what happens in terms of
[48:31] uh regime shifts. Regime shifts from
[48:33] investors right now where they start
[48:35] worrying about inflation and recessions,
[48:37] these things are going to get hit and
[48:38] they'll give up percentage of their
[48:40] gains, but their earnings will keep
[48:41] growing and that means they'll get a
[48:42] snapback very quickly. Um Jensen Huang
[48:45] gave a uh
[48:46] a call or an interview with Ben
[48:48] Thompson. The reason I want to bring
[48:50] this up, he gets into something that I
[48:51] wrote about this week with just CPUs and
[48:54] I think it's really critical for those
[48:55] of I mean for the subscribers, I sent
[48:57] this out this week. I sent the expanded
[48:59] version out with the name list. So for
[49:02] those of you who didn't get it um
[49:05] on the institution side, you can go to
[49:06] your sales people.
[49:08] Uh for the subscribers, you can go to
[49:10] the website. Uh it's a list of 18 names
[49:13] that in my opinion uh benefit from the
[49:15] move to the whole rack. Uh and I think
[49:17] you should be looking at those names as
[49:19] we go through corrections.
[49:21] Uh final few slides here. Uh Chamath put
[49:25] out something again on terminal value.
[49:27] This is something I believe in. What
[49:28] happens if AI makes every moat
[49:30] temporary? At a minimum, the uncertainty
[49:33] is growing.
[49:34] And I just wanted to highlight that I've
[49:35] been talking about this
[49:38] for a year and a half. So this was my
[49:40] first thing on the moat erosion engine
[49:43] and you can see I barely had any subs
[49:45] Substack followers uh at that point. You
[49:48] guys should go read my earlier stuff. Um
[49:51] if for no other reason, I did cover a
[49:53] lot of these topics back then, and I
[49:55] think that just shows how slow people
[49:58] have been to understand. This was before
[50:00] Deep Seek.
[50:02] Um, and I did write about how this will
[50:06] question everything due to the rapid and
[50:09] unpredictable changes driven by AI, the
[50:11] disappearance of motes as we know them.
[50:15] I wrote after Deep Seek, when everyone
[50:17] called me up, and I reiterated again
[50:20] that although I think this is a panic,
[50:22] and there's no way that this is going to
[50:25] kill, um, the demand side, and Jevons
[50:27] paradox will happen. I did go through
[50:30] again that anybody, including Nvidia,
[50:33] could be disrupted during this time of
[50:34] AI with novel solutions that are going
[50:37] to come up. This is going to happen for
[50:39] energy stocks down the road. It's going
[50:41] to happen for everything. Not in the
[50:43] near term, but you just have to get used
[50:45] to the disruption and the supersonic
[50:47] tsunami from AI. Uh, a VC, uh, moat
[50:52] situation, and I think this covers
[50:54] really well,
[50:56] uh, the SAS boom produced a few dozen
[50:58] billionaires and a bunch of zero-sum
[51:00] winners. But the SAS, the AI SAS era
[51:02] will produce massive millionaires. This
[51:04] is what I believe.
[51:06] Um, so I'm showing you this just because
[51:08] this is starting to get more credence,
[51:10] especially out of the Y Combinator
[51:11] group. I believe the total value of
[51:14] software goes up, and the number of
[51:15] companies created goes up exponentially.
[51:17] So, this is where everyone can be right.
[51:20] So, everyone that says the value of
[51:22] software is going to go higher, great.
[51:24] I'm fine with that.
[51:26] I don't think public companies will go
[51:28] up because I think the number of
[51:29] companies created will go up
[51:31] exponentially, and I don't think they'll
[51:32] ever be public. And that's what they're
[51:34] talking about here. There will be 50,000
[51:35] companies doing 500,000 to 5 million
[51:37] each, run by one to three people. That's
[51:39] what I believe.
[51:41] It misses that they'll have 10,000
[51:43] people employees. It's just that only
[51:46] three of them will be humans, the rest
[51:47] will be AI agents, which is why you have
[51:49] to understand what open claw does.
[51:51] The number of people who capture the
[51:52] value goes up by a hundred times. This
[51:54] is my belief in crypto. This is my
[51:56] belief in Bitcoin.
[51:58] I believe in the disruption of
[52:02] the concentration. This is the
[52:04] decentralization of ownership of assets
[52:06] and the decentralization of this. So,
[52:09] again, I'm bringing this up to remind
[52:10] people the Fed is confused.
[52:14] We've got a credit prom going on at the
[52:15] same time that we have inflation.
[52:18] We've got the White House calling for an
[52:19] emergency Fed cut
[52:22] during this whole thing. This is this
[52:23] week with gas at the pump soaring higher
[52:25] and inflation expectations going higher.
[52:28] And Kevin Warsh will be um
[52:32] well, it's at 70% now that he gets in
[52:34] before uh
[52:36] before May 15th. But the main point is
[52:38] what is he stepping into and what does
[52:40] that mean for Bitcoin?
[52:43] Uh we've got a buy signal from Ren Mac
[52:45] in terms of the excessive short
[52:47] positions on Bitcoin uh since I've
[52:49] chastised everyone for what's gone on in
[52:52] um in the bear side and bull side.
[52:56] Uh I'm just going to keep saying Bitcoin
[52:58] may fall another 20 30% if the S&P falls
[53:02] violently. I don't expect that to
[53:04] happen. I do think the S&P is going to
[53:06] trade down. If I had to guess, I think
[53:08] we can still get to 6100-ish if the oil
[53:11] price doesn't reverse immediately and we
[53:14] see the inflation data go up and we see
[53:16] the jobs or the recession fear starts to
[53:18] go up. I think we could easily see 6100,
[53:22] uh which would mean, you know, being
[53:24] down on the year 13% or so.
[53:26] Uh
[53:27] Here's the final thing I'll share. I
[53:28] built this in Perplexity computer as
[53:30] well. I didn't use my FMP data, but I
[53:33] told it to do the same check. Uh I just
[53:36] quickly went in and gave it what I
[53:38] wanted, which was at the end of the
[53:39] week, give me the weekly close,
[53:41] where we are in the 50, 100 day, and 200
[53:43] day. Oh, only thing above the 50 day
[53:45] right now is Bitcoin and Ethereum. Get
[53:47] ready.
[53:48] RSI's two year this in my commentary,
[53:51] could go down to 6100.
[53:53] Uh
[53:54] That's it for this week guys. Um
[53:57] again, thank you for everyone who's
[53:58] subscribed. Most importantly,
[54:00] uh hopefully you're learning. Watch the
[54:02] videos for the uh for the advanced
[54:04] subscribers.
[54:06] Uh I will put them up for people who
[54:08] just want to get the videos. Uh it's
[54:10] going to take a little bit of time.
[54:11] We're making sure the subscribers have
[54:13] had the ability. I will do more in the
[54:15] future. That's the benefit I'm going to
[54:17] try to do is I will show people
[54:19] uh hopefully how to do things in Open
[54:21] Claw, and maybe even Perplexity
[54:22] computer, and I'll go through and show
[54:24] people there. And I'll give them the
[54:25] actual things to go build it themselves.
[54:28] Uh meaning the prompts. Because the
[54:29] prompts are what happened. And if I tell
[54:31] you what you have to do on the API keys,
[54:34] which you don't even have to go in and
[54:35] add them to the script, you just tell it
[54:37] what the API key is, and it knows it,
[54:39] and then it goes through it. So, you've
[54:41] reached a point where anyone can do
[54:42] this. Um
[54:44] be safe during this period with your
[54:46] portfolio. Trade the book. Uh don't fall
[54:50] asleep on what's happening from the
[54:51] inflation side. And don't get too
[54:53] bearish. And certainly don't sit there
[54:55] and uh have PTSD uh from the uh taco
[54:59] days of last year. And I'll see you next
[55:01] week.