YT YouTube2Text

YouTube2Text → Transcripts → Jordi Visser: The WINNING investment ideas for 2027

Jordi Visser: The WINNING investment ideas for 2027 — Transcript

by Phil Rosen · 7,685 words · 1,107 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00Before we get to the show, YouTube is
0:02telling us that 70% of you are not
0:04subscribed to the channel. Please double
0:06check that you are subscribed so that we
0:08can keep booking great guests,
0:09publishing good content, and working
0:11hard just for you. Now, let's get to the
0:14episode. Jordan, you've been saying for
0:16a very long time that AI is the macro
0:19story in many ways, and you still got a
0:21lot of people talking about the Fed, oil
0:23prices, the classic economic indicators.
0:26Walk me through your latest thinking on
0:28this.
0:30nothing has changed. The only thing that
0:31matters in the economy is what's
0:33happening in AI and if anything it's
0:35only intensified. Uh I could make the
0:37argument which I do that all the rate
0:41rise this year for what's happened in
0:44treasuries is nominal GDP going higher
0:46which is part of the capex trade
0:48obviously uh but it also includes all
0:50the debt and the capital needs that have
0:52come through this year. So as the
0:55buildout for AI and all the data centers
0:57is happening in very concentrated
0:59companies and they were coming to the
1:01market in a very fast thing because they
1:02needed the money very quickly. It's got
1:04very front-loaded because we need the
1:06buildout because there's a backlog of
1:08demand that needs to be. So these guys
1:11are scrambling to raise capital. So this
1:13whole year, the rise in real yields, all
1:17of it to me has been driven by uh just
1:21AI and what's happened.
1:22>> And my sense is that you think most of
1:26Wall Street is thinking about rates the
1:28wrong way. Can you walk me through that?
1:31>> Not well, let's see. Um,
1:34so the people that I spend time
1:36listening to,
1:39and most of these are very senior credit
1:41people at the banks,
1:44there's a very easy story to talk about
1:46when it comes to AI relative to what you
1:48read on X. So, everyone who wants to be
1:50an a AI bubbleista will tell you that
1:54Core Weave is going to blow up and
1:55Oracle is blowing up and all this stuff.
1:57But when you listen to the people who
1:58run credit at the banks and they go,
2:01"Well, here's the issue. um in the first
2:04first half of this year insurance
2:05companies were willing to buy any paper
2:07we brought to the market that was longer
2:08than 10 years but then they ended up
2:10with heavy concentration. This is a very
2:12unique period of time where the
2:14companies that are raising lots of money
2:17are the biggest equity players as well
2:20which means that if an insurance company
2:23has any risk to Microsoft or to Google
2:25and these guys just keep coming with
2:27more and more money there's a
2:29concentration risk if you're at a fund
2:31that is buying debt on one thing well
2:34and you've got equity in the other as
2:36someone who's been a risk manager at the
2:37banks like this causes an issue. So,
2:40every credit person at the big banks has
2:43talked about that concentration has been
2:44a major issue in the credit markets,
2:46which is why you're seeing CDS widen out
2:48on a lot of these names because they
2:49have to buy protection for the
2:50concentration risk and for the hedging.
2:52That doesn't mean the AI bubble is
2:54blowing up. It just means that like the
2:56equity market, you have insane
2:59concentration with inside the credit
3:00markets. Uh that will subside at some
3:04point. And if I had to pick like a story
3:07that's important for this week, first of
3:09all, tenure yields have started to come
3:11down. Um, because of the needs for the
3:14AI trade, we saw the Treasury respond
3:17with, hey, we're going to increase our
3:19buybacks. Okay, it didn't have an
3:21immediate reaction, but now we've seen
3:22tenure rates come down. The skeptic and
3:24the bubble eastas will tell you this is
3:26because the AI trades unwinding. For me,
3:29watching the earnings grow and watching
3:30the demand numbers keep getting pushed
3:32out in terms of the supply demand
3:34imbalance, I would say this has more to
3:36do with they don't want affordability to
3:38be a major issue. And so don't be
3:41surprised if Bessant and team especially
3:43after he brought David Zervos over in
3:45September.
3:46If maybe now they're going to go, you
3:48know what, the demand side didn't work
3:49as well as we want. We're going to go to
3:52the supply side. And you have a
3:53quarterly refunding announcement coming
3:54up soon. Maybe by the time this is
3:56released, I don't know when you're going
3:58to do it, but on November 4th, we're
3:59going to hear more details. And if they
4:01start to change the supply side of
4:03treasuries, you'll probably see another
4:05reaction in terms of long-term rates
4:07going lower, at least for them. But I
4:08think the big story is for next year
4:10because I think there's a lot of things
4:12that were headwinds this year that'll
4:14become less of a headwind next year.
4:15like what
4:17>> uh what Oracle said
4:20is this is a company that since
4:22September which is really when as I call
4:24it the grinding up in the mud type thing
4:27where you can't get going except for a
4:29very short amount of time uh that is
4:31when Oracle became an issue September of
4:33last year that is when their CDS started
4:35to widen out because they needed to come
4:37to the market and borrow a lot of money
4:39so between equity and debt I think their
4:42number is about 60 billion for them over
4:45over the last 12 months. What they said
4:47this week in an investor day was, "Hey,
4:49we're going to be massively free cash
4:51free cash flow positive in 2029. It'll
4:54start to turn positive in 2028
4:57and we think we're only going to need
4:58about 20 billion more of capital." So,
5:00when you do these large capital
5:02investments and compute prices have gone
5:05higher, the benefit starts to come. And
5:07Gavin Baker has talked about this
5:09extensively that I think people are
5:10missing that compute prices have gone
5:12higher since these guys made the
5:14decision to build all this compute. So
5:16if you make this decision to build all
5:17this compute and the prices have gone
5:19higher and you're doing it at
5:20yesterday's prices, there is a benefit
5:22that comes eventually and when that
5:24monetization comes, you don't need to
5:26borrow as much money. Now that is just
5:27Oracle, but Oracle is the first company
5:30to talk about this. But other companies
5:33are going to run through the same thing
5:34as their data centers come online. And
5:36Gavin has talked about how this could be
5:38the peak of external financing for the
5:41hyperscalers. If that's true, which I
5:44think even if it's not the first half of
5:45next year, the growth rate according to
5:48Goldman Sachs in terms of IG issuance,
5:50it's going to peak. So, we're kind of at
5:52the peak demand for long-term debt. And
5:55I think that's going to be one of the
5:56issues. The other thing is SPVS are
5:58starting to become a bigger way for
6:00these guys to raise money. Um, Elon just
6:03did it with SpaceX for 40 billion. I
6:05think he did it with Apollo. There's a
6:07lot of dry powder in the private credit
6:09markets, believe it or not, because it's
6:11almost been a year since the private
6:12credit issue started to become a major
6:14story. And as time goes on, as we've
6:16learned with commercial real estate, at
6:18some point you start going through the
6:21debt and all these bad things get
6:22written down. You start selling off some
6:24of the assets and things start becoming
6:25a different story. I think next year is
6:27going to be a different story from the
6:29massive impact that we saw from the bond
6:31market. Today's episode is brought to
6:34you by Direction Tesla headlines move
6:36markets on earnings, deliveries, and
6:38robo taxi news alike. The direction
6:40daily Tesla bull 2x and bare 1x ETFs,
6:44tickers TSL and TSLS seek 200% and
6:49negative 100% before fees and expenses
6:52of Tesla's daily performance. Bullish on
6:54Tesla's next catalyst or bracing for a
6:57pullback, trade either side with TSL and
7:00TSLS. These are non-diversified and
7:03concentrated in a single company, and
7:05investing in the funds is not equivalent
7:07to investing directly in Tesla. An
7:09investor should carefully consider a
7:10fund's investment objective, risks,
7:13charges, and expenses before investing.
7:16A fund's perspectus and summary
7:17perspectus contain this and other
7:19information about direction ETFs. To
7:22obtain a fund's prospectus and summary
7:24perspectus, call 866-4767523
7:28or visit the website at direction.com.
7:31Now, let's get back to the episode.
7:35Okay. So, if I'm understanding
7:37correctly, my sense is you are not super
7:40concerned about debt financing in the AI
7:43economy right now. Is that right?
7:45>> Yeah. But I think rates will stay higher
7:46than they otherwise would because the
7:48demand for capital is just going to be
7:50continuous. But it'll be more spread
7:51out. It won't be as concentrated going
7:53forward.
7:54>> Okay, understood. Is there anything in
7:57the earnings we've seen from Amazon or
8:00Alphabet about their negative free cash
8:02flow that worries you?
8:05>> I have been um
8:08negative on a relative basis of the
8:10hyperscalers this entire year and none
8:12of that's going to change. Uh, you will
8:14never hear me unless something changes.
8:19I'm just not positive on public
8:21companies that are based on code. I
8:24think they're going to always have this
8:27competition fear that comes from AI,
8:29which is legitimate. So rather than
8:32focus on these companies, their earnings
8:34are fine. It's their businesses that are
8:36being questioned. Their free cash flow
8:37is not a story to me. It's this is
8:39ridiculous. Like it literally is. is and
8:42if people did their homework, the hyper
8:44scaler's balance sheets are pristine.
8:47So what does free cash flow mean? If
8:49you're pumping in revenues at a very
8:51fast pace, you're spending money on
8:52future
8:54unless people again believe there's a
8:56bubble and the demand for AI will not
8:59happen, which is just not a belief that
9:01I have. So for the people that are
9:02worried about it watching this, if
9:04you're listening to people that keep
9:06telling you there will be no revenues
9:08coming in, I just don't subscribe to
9:10that. It's as someone who uses this all
9:12day long. I find this mainly coming from
9:16people who don't use it. It's my own
9:18impression.
9:20>> I mean, I personally am an owner of
9:23Google stock. I'm still bullish on it. I
9:26was on TV yesterday talking about Google
9:28stock being one of my holdings and why I
9:30still like uh the the stock even though
9:34they were negative9 billion in free cash
9:36flow. I think they're going to make half
9:37a trillion bucks this year. It's pretty
9:40insane those numbers. Um, Jordy, let me
9:43ask you about what we saw with Microsoft
9:45a few days ago. They're offering buyouts
9:48to 60,000 employees. And I'm going to
9:50read you a quote here from one of your
9:52recent notes. In previous economic
9:55cycles, governments often supported
9:57labor because capital was struggling. In
9:59the AI cycle, governments may have to
10:02support labor precisely because capital
10:04is succeeding. Help [snorts] me tie
10:06these things together here.
10:08So um if [clears throat] you go back to
10:11what a recession is and what has
10:13happened
10:15for every recession,
10:17you end up in a scenario where credit
10:19spreads widen and companies that have a
10:22lot of debt, which is what the fiat
10:23system is based on, the credit based
10:25fiat system, uh they need to borrow
10:28money and the markets have shut down.
10:30And so these companies debt starts to
10:31widen out. And eventually after they've
10:33tried to cut rates and it doesn't work,
10:36they have to do something bigger. And
10:37that bigger thing usually means allowing
10:40these companies to survive. And why does
10:43this have to happen? Why do we prevent
10:47recessions? Recessions theoretically
10:49could be cleansing. You could get rid of
10:50the weak and allow the strong to go. So
10:52they can be a rebirth. Um, we all try to
10:55teach our kids life's not fair and
10:57there's going to be bad times and
10:59hopefully you get stronger during the
11:00bad times. If you just bail people out
11:02all the time, then you end up in a
11:05scenario. We have a lot of zombies just
11:06kind of going around that are bad
11:07businesses and they can have good
11:10products, they can have whatever, but if
11:12they have bad businesses where they
11:13can't actually survive without the
11:14government helping them, you run into an
11:16issue. The problem is with the Microsoft
11:18scenario, when people have been fired in
11:20the past from businesses, there's always
11:22more job hiring that's happening.
11:25We're talking about Microsoft. And what
11:28I wrote about is this is a very, very
11:31bad situation for people to think about.
11:34When you offer buyouts to 60,000
11:36employees,
11:3827% of Microsoft's employees, 27%.
11:43Now, not everyone's going to take the
11:44buyout, but for the people that don't
11:46take it, the handwriting's on the wall.
11:48Like, this is what they want. In between
11:51the lines, this means Microsoft
11:53envisions four years from now, 3 years
11:55from now, two years from now, they don't
11:58think they're going to need as many
11:59people and their revenues are growing
12:02rapidly. So, if all 60,000 took the
12:06buyout, you take this back to 2020
12:08levels. So COVID levels before COVID, so
12:12that year,
12:14if you got back to 2020 levels and you
12:16look back and you see the revenue for
12:17the company has gone up 70% off next
12:20year's numbers, then revenue per
12:22employee is going parabolic. That is the
12:24productivity gains that are supposed to
12:25come from AI. That means we're entering
12:28a period right now where the pressure
12:30gets put on the government because
12:32historically they would bail out the
12:35economy if those types of numbers were
12:38happening. But this time it's happening
12:40with all-time highs in stocks with
12:42unemployed, you know, jobless claims at
12:45the lowest level in the history of the
12:46country. Meaning no one is actually
12:49claiming insurance for unemployment. So
12:52what is that saying about where we are
12:53in the economy? AI is h going to start
12:55having a very large impact on the way
12:58companies make decisions. And so far
12:59it's been through a lack of hiring,
13:01which we've seen in the numbers. But at
13:03some point the bigger companies are
13:05saying we have to pay for this capex
13:07buildout in AI and the easiest way for
13:09us to do that is instead of raising
13:12capital if we think we can get rid of
13:1360,000 employees which would be about
13:17$24 billion using about $400,000 of per
13:21person. This is a very unique situation
13:23and it puts the government in problem
13:25because the government needs tax
13:26receipts from the individuals and the
13:28corporate tax rates are much lower than
13:30the individual income tax rates. So you
13:32end up in this very weird dynamic where
13:34the government is now put under pressure
13:36at the same time that they have a huge
13:39deficit.
13:40>> Do you expect more of these large
13:43corporations to make similar moves? Cuz
13:45Microsoft, I think they're a $4 trillion
13:47valuation company and if they are making
13:50the call to cut this many people, they
13:53have almost no financial pressure is is
13:55my guess to do this. It's mostly uh
13:57streamlining productivity uh almost like
14:01positive momentum, but the companies
14:03that aren't doing as well as Microsoft
14:05will probably be under more pressure to
14:07make cuts like this. Is that right?
14:09>> Well, remember the the way they're
14:11handling this is this is buyouts.
14:14They're offering people. What if nobody
14:17takes it? Everyone says, "Nope, I I want
14:20the job. I want to stay here." Then they
14:22have to make it more attractive, which
14:23I'm sure they're making it fairly
14:24attractive for people because if they
14:27could get all 60,000 employees, I mean,
14:29there's a lot of story there in terms of
14:31what it is, but if it's going to save
14:32them 24 billion, what are they willing
14:34to pay people a half years worth of
14:37salary on all of them? So 12 billion to
14:39get rid of them to save 12 I I I don't
14:41know. I'm just saying that I do believe
14:44that the bigger companies that have the
14:46capability of incorporating AI and
14:49that's the critical thing. Microsoft is
14:51a very tech-savvy company. So most of
14:54the large layoffs that we've seen have
14:56been in tech companies. If you think
14:58about like what Block did and a whole
15:00bunch of places where they just got rid
15:01of a bunch of people and said, "Well,
15:02this is just taking them back to 2020."
15:04I think these are the ones that know the
15:06benefits that come from AI and can
15:08incorporate them in their business much
15:09easier. It's much harder for a lot of
15:10the Fortune 500 companies. I mean, if we
15:14got some of these mass buyouts, even
15:17with just tech companies, my sense is
15:20that gets us closer to sort of Elon
15:23Musk's vision of a future where the
15:26government has to give everyone um
15:28universal basic income. Is that
15:31somewhere in your outlook?
15:33>> Uh, it is, but I wouldn't call it
15:36necessarily that. And I think that's
15:37what freaks people out is cuz they have
15:39these preconceived notions of UBI and
15:42what it means.
15:43I think
15:46I would call it a uh
15:50corporate tax rates would go higher for
15:52companies like Microsoft. Let's just
15:54assume that they're successful with
15:57this. And the government goes, okay, if
15:59you've seen your, you know, if if your
16:02revenue per employee is above this
16:04level, we are going to move your tax
16:07rate up to this. And again, this is if
16:09you think about what a wealth tax is and
16:11how hard it is to implement, but with a
16:14corporate tax rate where you're like,
16:16okay, here's what we're going to do.
16:17We're going to cut individual
16:19tax rates and we're going to raise
16:21corporate tax rates only for companies
16:23that make X. Well, then you're
16:25sponsoring an entrepreneurship in the
16:27country, meaning small businesses
16:28thrive, large businesses suffer. Uh, I
16:31think the government's going to get more
16:33involved in large businesses going
16:35forward because of this labor capital
16:37story. I don't know the form, but rather
16:39than call it UBI, I call it a
16:41redistribution of profits from capital
16:43to labor. So how that happens I don't
16:46really know but somewhere within there
16:49along with Elon also talks about the
16:51abundance coming from prices going lower
16:54all of these get related to
16:55affordability and I think for people on
17:00like the buyout side
17:02if they could get a package for a period
17:05of time and then know that the following
17:07year their tax rate was only going to be
17:0920% of what it is there might be a
17:11different viewpoint if they didn't have
17:13to go work in the office. Like there's a
17:15bunch of things that have been happening
17:16in society for a period of time. So I
17:18think to come up with any solution that
17:21is just one word UBI, I don't think it's
17:22going to happen. I think there's going
17:23to be a lot of things that are going to
17:24have to happen because I do think AI is
17:26going to have a big long-term effect on
17:28the labor capital side.
17:30>> I agree with you. I I think that's a
17:32really good framework to use. What do
17:34you make of the notion that AI has not
17:37yet shown up in the productivity data?
17:39because that's something that a lot of
17:42more bearish macro folks have been
17:43pointing to saying, "Hey, why uh does
17:47the macro productivity data not show any
17:50of the supposed gains we're getting from
17:52AI yet?" It has shown up. It just has
17:55shown up in corporate profit margins at
17:57this point. Um,
17:59and again, I think the adoption really
18:01only started this year.
18:03uh the technology and the progress of it
18:06only got to the point where AI agents
18:07which is really the most powerful side
18:09because before AI agents so let's say
18:12before a year ago which is really when
18:14coding agents started before a year ago
18:17I really don't know how companies would
18:19get productivity gains they need the
18:21individuals to use it the companies
18:23weren't allowing them to use it there
18:25was a whole bunch of things so I think
18:26corporations take a little bit more time
18:28to show up what has shown up is
18:31corporate corporate profit margins have
18:33gone up because there's no hiring. And
18:35this is the thing I just don't think
18:36people understand. Like outside of of
18:40nursing, outside of the healthc care
18:42sector, which is not part of the S&P
18:44500, there's been no job hiring for
18:48almost 2 years. I mean, it's like 8 20
18:50months now. So, it's really hard for me
18:53to sit there and listen to people and
18:54go, there's no, you know, no
18:55productivity gains. I'm like, how can
18:57you have profits growing and you're not
18:59hiring anyone? And that's never happened
19:01before in capitalism. You always hire
19:03people as it's growing because you need
19:05more people. So if it's happening
19:07through profit margins, that's one side.
19:08On the flip side, there is zero chance.
19:12Zero. So this is I don't like getting in
19:14anything in probabilities. There is 0%
19:16chance that GDP as a calculation which
19:19was created for the industrial
19:21revolution, which is based on the volume
19:23of goods, can deal with intangible
19:26things like AI. It couldn't handle the
19:28internet. We had to revise higher the
19:30productivity numbers in the '9s. I don't
19:32think people talk about that. So when
19:34you're talking to macro people, the '9s,
19:36we revised higher significantly the
19:39nominal GDP numbers. Then we did it
19:41again because R&D was counted as an
19:44expense. And they're like, well, you
19:45know what? This isn't an expense. This
19:46is actually IP. Like what happens when
19:49the value of a company is on the
19:51technology that they own as well and you
19:54start bringing that in? I just don't
19:56think that people know how to I I know
19:59for a fact that the measure of GDP,
20:02which is the way that you come up with
20:03the productivity numbers, is just not
20:06real. So, if you took nominal GDP and
20:07said, you know what, we're revising it
20:09up 20%. Oh, look, magically the
20:11productivity numbers got better. So, I
20:12think it's it's just a bunch of BS to
20:15sit there and talk about the
20:16productivity numbers. The only way that
20:18I can sit there and say is very simply,
20:19here's a fact. Profits are growing
20:22rapidly in the S&P 500. revenues are
20:24growing rapidly and we're not hiring any
20:25people. Plain and simple.
20:28>> Okay, so here here's something uh maybe
20:31is a contradiction, but I'm sure you
20:32have a good answer for it. If I wanted
20:35to get exposure to the AI productivity
20:37boom and no one's hiring all these big
20:40companies, but their profit margins are
20:42increasing. Is it as easy as saying,
20:45"Let me just buy the S&P 500 and then go
20:49away." If all these companies are
20:50supposed to be more profitable, more
20:52productive, but then you're also not
20:55super bullish on public companies, can
20:57you square those things?
21:01So when the more people you have, so
21:04think about what Microsoft wants.
21:07Microsoft wants 60,000 people to accept
21:09their buy. Okay? like you don't make
21:12this public announcement. They want
21:1460,000 people to basically leave the
21:16company and they'll deal with the
21:17consequence that come in. They have
21:19comfort that AI that so they can get
21:21immediately productivity gains. What
21:24does that say about public companies
21:26that they're overstaffed at this point
21:28relative to what it should be? That's
21:29not a positive thing. It's bad for
21:31morale. It's bad for a lot of it's bad
21:33for the culture, but it's really bad
21:36when you compare it to what's happening
21:37for private companies and how fast
21:39they're growing. So, if you just take
21:41the S1 that we got from Anthropic and
21:43compare it to Google, like he didn't
21:46make a very compelling case for being
21:47long Google. So, um
21:48>> that's fair. I want to hear yours.
21:50>> You just threw out a bunch of revenues
21:52and I was like, "That's fine. I make
21:54money."
21:54>> It's Google. What world do we live in if
21:56there's no Google?
21:58>> Uh I mean, you could have said the same
22:00thing about Ford 50 years ago. So, let's
22:02be careful with with that. Um to to have
22:05a company with a lot of people to
22:07compare it to what Anthropic did. So,
22:09Anthropics S1
22:11um when they when they released their
22:13X1, their revenue was 16.4 billion.
22:18Okay. Google's in 2004 was about 1.5
22:21billion. So, we're dealing with more
22:24than 10 times the size of when Google
22:27did their S1 versus Anthropic. And the
22:30growth to those numbers was about the
22:32same. So, Google was about a 5year
22:34six-year company and Anthropic is a
22:36fiveyear sixyear company. So the time
22:38period was the same. Google right now
22:40has a couple hundred thousand employees
22:43like Microsoft. Anthropic has less than
22:4510. When you start going through AI
22:48native businesses,
22:51they don't have as many people. People
22:53become very very dangerous in a world of
22:56AI for a variety of different reasons.
22:58Um hackings are happening more and more.
23:00And one of the ways that people hacking,
23:02the more people you have, the more
23:04laptops you have, the more exposures you
23:05have, the more holes you have, the more
23:07things are built by humans. An AI native
23:09business starting today is going to use
23:10the most modern technology. They're not
23:12going to have things built from a decade
23:14ago, which allows AI to break in. There
23:17are so many different ways that I can
23:18make an argument as to why and again I
23:21believe earnings for the S&P 500 will
23:23continue to grow. But this should be a
23:26warning sign for everyone that as of
23:29right now S&P earnings are going to be
23:30up about 30% for the year. But the S&P
23:34as of when we sit down here today is up
23:36about 11%.
23:38Now on the surface it's another hey
23:40imagine how bearish everyone is and
23:42we're barely off alltime highs and we're
23:43up 11%. But there's something very
23:45interesting happening. Profit margins
23:47are rising, earnings are growing, but
23:50the multiple of the S&P is going down.
23:52And that's why I keep saying this is a
23:54warning sign. Markets are very, very
23:56good at warning you. And what it says to
23:58me is this, whatever the risk is this
24:00year, which is Iran, straight of horn
24:03moves, bond yields, all these fears.
24:05Will the Fed have to raise rates next
24:07year? I don't think it's going to be
24:08those things, but I also don't think
24:10it's going to be an all clear sound. I
24:11think it could be the labor market. And
24:13I think it could be the politics that
24:14are becoming a major issue. You could
24:16still have more delays and more
24:17bottlenecks between memory and data
24:19centers. You could have much bigger
24:21hackings than we've seen, which are very
24:23isolated at this point, but we could
24:24have significant geopolitical issues,
24:26hackings, things across the board. And I
24:29think it would be foolish to not expect
24:31AI to continue to have some kind of a
24:33negative thing every single year that
24:35prevents us from getting in one of these
24:37easy bull markets where there's nothing
24:40happening. I think it's going to
24:41continually be this question that's
24:42going through the market.
24:44>> So, all these different headwinds, uh,
24:46cyber security, politics,
24:49my my understanding of your thesis on
24:51Eli Liy, which is the one I want to talk
24:53to you about next,
24:54>> this is sort of immune from all these
24:56headwinds you're describing because it's
24:58just a massive health and pharma
25:01juggernaut. Can you walk me through your
25:03thesis on this? I know you've been very
25:04bullish on it for a long time. You're
25:06one of the first people that I spoke to
25:07about Eli Liy that got me in the stock.
25:09So, how are you thinking about this
25:11today?
25:12>> First of all, it it the stock is a
25:15beast. Um, and relative to the amount of
25:18people who want to talk to me about it,
25:21it's a triple beast. You very seldom can
25:25find things that have compounded at over
25:2630% a year, which Eli Liy has for 10
25:29years with 10 straight up years, and not
25:31have every person in the world talking
25:34about it. Um, I get a lot of emails from
25:38subscribers. I get a lot of emails from
25:40institutions.
25:42I I can't remember how many times I get
25:45something on Eli Lilly ever. Like it's
25:47just not a name that people care about
25:50uh at all. And I guess that is partly
25:53the way that people think about I want
25:55to find something that's going to double
25:56this year and Eli Liy just compounds at
25:5930 to 40% a year. But the story behind
26:02it
26:04is so important in the world that we're
26:06in. So everything that I said about AI
26:08being a disruption,
26:10I Eli Liy is in a guaranteed non-yclical
26:14growth business.
26:16They sell medicine
26:19for obesity and diabetes, which is a
26:21huge problem in the United States. The
26:24impact that it has on people's lives are
26:27significant in terms of their ability to
26:29live longer. It extends your life and
26:32it's been extremely expensive. there's
26:34been supply problems, manufacturing,
26:39these numbers, the penetration is still
26:41very small. Even in the US, I think 10%
26:44of the 100 million people that are
26:46determined to be obese actually are on
26:50it. Now, part of it's because of there's
26:52only been injectables, but now we're
26:53going to start having pills. There's a
26:55whole bunch of growth factors that come
26:57in, but it's their numbers that are
26:59staggering. If you didn't know the
27:01company and you didn't hear Eli Liy and
27:03I said to you, okay, this is the ninth
27:05biggest company in the US. So when I've
27:07said publicly, this will be the biggest
27:09stock in the S&P 500 with by 2030, I
27:14absolutely believe that. And this is
27:16someone who believes in the Agentic
27:18Compute side, but there's no reason in
27:19my mind that Nvidia can't trade at a 5p.
27:22Like I literally am that skeptical of
27:24market expan multiple expansion for
27:26companies that have cyclical risk of AI
27:29with terminal terminal value issues. Eli
27:31Liy does not have this. There is nothing
27:34that is going to disrupt the obesity
27:36side in the near term. So that cash flow
27:39business, the the pharmaceutical
27:41companies, not only are the buyers of
27:43them getting older, so they need more,
27:46but they're also financed by Medicare
27:48and Medicaid, and they have it's a
27:50nonyclical 25% of the economy or it's
27:53about to be 25. So from a consumption
27:55perspective, you have it there. It's
27:56like guaranteeing that there's going to
27:58be a new iPhone purchased every year by
28:00a person. But the other thing is it's
28:02not just the non-yclicality.
28:04To compete with it, you have to go
28:05through an FDA approval process. like
28:07it's regulated. So, it's literally like,
28:09so let me get this straight. This is
28:11like a utility company
28:13selling stuff in a world where we're all
28:15worried about when the bubble's going to
28:17end. And if you go look at how
28:19healthcare does and I heard your
28:20interview with Adam Parker, one part you
28:23didn't talk about with healthcare go
28:24actually he may have, but during the
28:26dotcom bubble, you wanted to be long
28:27healthcare. During the great financial
28:29crisis, you wanted to be long
28:30healthcare. Yes, it's on a relative
28:32basis and I think what Adam said which I
28:33agree. So I want to be along the theme
28:35of agent compute and I want to be along
28:38the theme of longevity. I believe AI
28:41needs the compute and I believe the only
28:45benefit that's going to come that can't
28:47be competed away is healthcare. That is
28:49the only one I can argue that as I've
28:52said banks are going to go through a
28:53very big issue verse crypto as the
28:55guardrails go. I can make an argument
28:57that SAS is still going to have the
28:58headwinds. The hyperscalers are still
29:00going to have the issue. Anything built
29:01on code is going to have an issue. But I
29:04think health care is going to benefit
29:06from us entering the sugentic world,
29:08getting solutions, and people will live
29:10longer. And if you believe in people
29:12living longer, I believe you want to
29:13have
29:15an investment in healthcare. And Eli Liy
29:17is just to me the purest play with
29:19company that's revenues are growing 40%
29:21year-over-year.
29:23That is such a well articulated thesis,
29:26way better than I can do. And that is my
29:28largest stock that I own. So, uh, it is
29:31for me now, too, given the fact that
29:32Marll's had a bad week and Nvidia just
29:35doesn't seem to move. Uh, those are my
29:37biggest AI names, although I did buy
29:39some more Micron. I'm getting back into
29:40the uh the Micron game again. So, okay.
29:43So, let let me ask you about that. Uh,
29:45we had a conversation at the start of
29:462025 and you told me Micron's going to
29:49the moon and this was for my best ideas
29:52club. That was when Micron was an $89
29:55stock. So, it was what, a 12 bagger at
29:59one point and went up to $1,200.
30:01Why are you getting back in now? Because
30:03I know at one point you sort of got out
30:05of it. I don't know if you got out
30:06fully, but what what's the thesis now on
30:08Micron?
30:09>> I did get out of it fully back in um
30:12May. I think I think the last sale was
30:14probably early June, but the May June
30:16period is when I got rid of the majority
30:18of it. Um I thought there was a better
30:20place to put my money um at that time,
30:22which ended up being right. uh I got out
30:25a little earlier, but let's assume my
30:26average exit price was somewhere in the,
30:29you know, the 700s and I've averaged
30:32buying back in in the low 800s. Now, the
30:35way that I think about
30:38stocks, especially that one is at that
30:41level, it was too many people talking
30:43about it where when it was 89, nobody
30:46was nobody would listen. Yeah. Nobody
30:48would listen. Um, and again, my
30:50[clears throat] thesis, and I remember
30:51us having this conversation. My thesis
30:53eventually got into humanoids. And I
30:54said, it's the same thing I'm saying
30:56with Eli Liy. In fact, we didn't get
30:57into this, but Eli Liy has the same
31:00story as Micron for me in the fact that
31:04once you get through the cash cow
31:07business that Eli Liy has, they're also
31:10seeing revenue growth in immunology,
31:12oncology, dementia. like there's other
31:15areas that they're benefiting from and
31:17they're spending an enormous amount of
31:19that cash flow on buying M&A for Micron.
31:21I talked about humanoids. We're closer
31:24and closer to humanoids and Elon is very
31:26like open about the amount of RAM we
31:30need for humanoids. We can't have
31:32enough. In fact, I think he said it's
31:34the scarcest asset on Earth.
31:38So, I bought back in because we've now
31:40gone through time and price. So for me
31:43as someone who's managed money, when
31:46something gets extended, I want to see
31:48people exit it, which has clearly
31:50happened. All you got to do is look at
31:51the Cosby and look at all this. We've
31:52we've taken out a lot of retail. We've
31:54taken out a lot of people. Now Micron
31:56has basically gone sideways.
31:59And the reason I've been buying into it
32:00in weakness is because the earnings and
32:03the contracts that they've locked in
32:05now, I don't think there's a bubble. So
32:08what I use the the five months to do is
32:13to really make sure that I'm comfortable
32:15that hey CDS is widening on this. Do I
32:18think there's an issue? Let me read
32:19through the earnings reports. I spend an
32:21enormous amount of time during the
32:23earnings season and especially this one
32:26here. So I had a very small position in
32:27Micron before two weeks ago, but now
32:29it's I mean it's about the same size as
32:32it was in January of last year. Okay.
32:35Now, it got much bigger during the
32:37runup. Obviously, it's much smaller than
32:39when I sold out. But the reason that I
32:41cared is because as I go through all the
32:43earnings reports, SpaceX, okay, they're
32:46talking about 10 gawatt possibly being
32:49done next year. And he's saying it will
32:51happen. Okay, Elon is on hardware. I'm
32:55not going to doubt him in terms of
32:56building stuff. We need a lot of memory
32:58for that. I'm not worried about a bubble
33:00if Elon is saying that he's going to
33:01build this compute. So, I'm less worried
33:04about Micron, even though it would not
33:08shock me in any way if it still went
33:11down to 700 again. But what's happened
33:15is the 200 day moving average has gone
33:17higher. It's gone sideways. The bearish
33:19sentiment and the deleveraging that's
33:21happening at the hedge fund side, it's a
33:23much cleaner position. And they've
33:25signed long-term contracts over the last
33:27two quarters that extend well into 2030.
33:30And you've had SKH say the same thing in
33:32their most recent earning reports. And
33:33when I read the Corning report and I
33:35read the SpaceX report and I read all of
33:37these different ones, I see the exact
33:39same story, which is number one, there's
33:40not only not a bubble. People are
33:43starting to try and lock up supply for 5
33:46years from now and compute prices have
33:48gone higher. So on every level, there is
33:50no bubble.
33:52There's just a bottleneck that's
33:55happening and we've taken out people
33:56that have worried. up.
33:58>> You know, Jordy, one one of the things
34:00that I appreciate about your work and
34:02and always speaking with you, you are
34:04very good at seeing around the corner as
34:06an investor.
34:08>> If let's say myself or another
34:11individual investor is listening to this
34:13wanted to get better at seeing around
34:15corners as an investor,
34:17>> how should they start?
34:19>> Don't don't try to extrapolate what's
34:21happening today. you have to really
34:24think about a lot about um what's
34:27driving it and what things are gonna
34:29like I said I like call options to where
34:32the demand a year from now and two years
34:34from now I can see in this so I spend
34:36time on the macro side as to okay what
34:39are people talking about are do yields
34:41matter does oil matter all that stuff is
34:43important but the reality is to me the
34:45structural trend and this is what I've
34:47always done is I try to figure out the
34:49story that's going to happen for the
34:51next decade
34:53Adam Parker talked about this. I
34:55completely, in fact, his number is spot
34:57on with me. In fact, I'm going to res
34:59I'm I'm going to repeat something he
35:01said to you. We're in about inning three
35:04of the AI trade.
35:07The other thing he said was the stocks
35:10are in inning five. So, let's go through
35:13what that means because this is the
35:15important part for looking ahead. This
35:18is a much tougher time to be in the AI
35:21trade, especially the memory trade than
35:22it was a year ago for the reasons that I
35:25said. I don't expect Micron to 12 bag. I
35:28don't expect them honestly over the next
35:31year. If if Micron doubles from here,
35:35I'd be surprised. This is something
35:37where I'm putting money in because I
35:39think the downside is maybe 10 to 20%. I
35:41think the upside is 60 to 80%.
35:44Okay, so that's a good riskreward to me.
35:46I'm back in that six to, you know, four
35:49to six times scenario. You constantly
35:52have to reassess things, but you have to
35:54find those secular themes that you think
35:56are going to play out. So, I'm telling
35:57people right now, longevity is a major
36:01major theme. Human life is going to
36:04extend living.
36:06So, you don't want to be long McDonald's
36:09anymore.
36:11Partly because OMIC is going to be used
36:13by more people. that trade is still
36:15happening. You do want to be long things
36:18related to travel, things where people
36:20are going to live longer and be
36:21healthier. All of that shift, which is
36:23what's been happening, the recreation
36:25side, positive, the, you know, the
36:28bedding side, the gambling side, the
36:30demographic side, positive. And then on
36:31the negative side, I just think anyone
36:33that thinks that people are going to get
36:35unhealthier and that longevity is not
36:37going to increase significantly, not
36:39just minor, but people are going to have
36:41healthier ends of lives. I just think
36:44that those types of trades are thematic.
36:46You have to believe in them. You have to
36:47spend your time on it. And maybe it's
36:49just the fact that I if I don't hear
36:51people talking about them, crypto is my
36:53favorite theme right now by far. And the
36:55reason is because
36:5795 people out of a hundred that I meet
37:00that have money. So most of the world's
37:02money does not believe in crypto and
37:04they don't know how to play it. So for
37:06me that is a very positive angle for me.
37:08And so I'm very hyperfocused over the
37:10next year as the biggest weight in my
37:12portfolio is crypto by far.
37:14>> And when you say crypto is that anything
37:17beyond Bitcoin or just Bitcoin?
37:20>> So historically it was only Bitcoin. Um
37:23that changed
37:26in the summer.
37:28It changed when uh when Bessant
37:32basically admitted that we have a
37:34problem in long-term yields, which I
37:37never thought there'd be QE again. I
37:39never thought there'd be some bailout
37:41thing again. And the reason is because
37:43the economy is too strong because of AI.
37:45So as long as the government, remember
37:47when we had this new administration, we
37:48were talking about we're going to get
37:49the deficit down. Besson said, "We're
37:51going to get the deficit down to 3%."
37:5233. We heard that all the time. I don't
37:55care who it is, you can't get the
37:57deficit down unless you're going to cut
37:59spending and they can't find a way to
38:01cut spending. So, the government is
38:02going to spend 5 to 6%. The interest
38:06expense has gone higher. So, the overall
38:08situation of the government is getting
38:10worse. They have to do something in
38:11yields for them, but then they also have
38:14a voting block of young people that
38:16affordability is a major issue. That's
38:18why I connected Microsoft to what's
38:19happening in France. Every time that I
38:22listen to anything on a podcast in
38:24France, affordability is a major issue.
38:27If affordability is a major issue over
38:29there, whatever the form that takes and
38:31at the same point we have Microsoft
38:32doing what they're doing, I believe this
38:34is a major issue. So what started for me
38:36after that was I've got the government
38:39which is being pressured by AI but then
38:41I have AI agents coming which means over
38:43the next year a AI agent or aentic
38:46commerce is going to start to tr take
38:48off. I've been a big writer and believer
38:50in stripe. I still believe this is an
38:53enormously important company that enough
38:55people don't spend spend time on but
38:57then on the other side you have
38:58tokenization happening and tokenization
39:00is a big big event. Uh I've written
39:03about this. I don't think I sent you the
39:05work, but I've talked about the
39:07importance of tokenization in terms of
39:09changing M2. It we're going to be doing
39:1324/7 trading. It's going to feel more
39:16like crypto. You're already seeing that
39:18anyway. The AI trade has been incredibly
39:20volatile this year, far more volatile
39:21than Bitcoin. So, what I extended into
39:23was Ethereum and then I extended into
39:26things around privacy. Uh, I also
39:29created and have my own research now for
39:31crypto that has 46 names similar to my
39:33AI 100 and I own a lot of those names
39:37with inside of there and I'll probably
39:38own all of them to some degree because I
39:40really believe in crypto that much. 2027
39:43is the year for me with crypto uh where
39:46I believe the commerce side and the
39:48government pressure from rates is going
39:49to be a major story. [sighs and gasps]
39:52>> We are covering a lot of ground today,
39:53Jordy. Um, where can we send people to
39:56go find all your work cuz you're
39:58publishing a ton of content across all
40:00these different channels. It's all
40:01excellent. What do you say?
40:03>> Uh, they can find me on YouTube for sure
40:05for my weekly which anyone can get for
40:09free and I just kind of cover what's
40:10going on in the week and you get used to
40:11the way my brain works. Um, and I talk
40:14about only the facts that have come out.
40:17Uh, but if people want to go subscribe
40:20and get some help in AI, get some crypto
40:22research, get up to speed on it, but
40:24also a lot of research every week. I do
40:26videos now multiple times a week between
40:29crypto, between AI, technical analysis.
40:32Uh, that's at a.22vressearch.com.
40:36>> Incredible, Jordy. I always appreciate
40:39your time. Always learning from you. So,
40:40thank you very much.
40:41>> Good to be here, Phil. Always.

About this transcript

This page contains the full transcript of Jordi Visser: The WINNING investment ideas for 2027 by Phil Rosen, generated from the public captions YouTube serves with the video. The transcript has 7,685 words across 1,107 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.