0:00

Forward Guidance is brought to you by VanEck,

0:02

a global leader in asset management since 1955.

0:05

You'll be hearing more about a VanEck ETF

0:07

later on, but for now, let's get into

0:09

today's interview. Very

0:15

pleased to welcome back to Forward

0:18

Guidance, Sotrini, a stock picker

0:20

and trader who's been on this program

0:22

two times before and over

0:24

the past year has a quite an

0:26

impressive track record. When Sotrini

0:28

first came on June

0:31

2nd of 2023, that was when he filmed

0:33

it, it aired later on in June, he

0:36

had just set up his AI baskets,

0:39

Investing in Artificial Intelligence, and

0:41

that basket as we stand here

0:44

today on June 15th, over a year later,

0:46

is up somewhere around 60 to 70%, and

0:49

the entire Sotrini

0:52

index or Sotrindex from

0:54

May 30th, 2023 to June 2nd, 2024 is up 100%. So congratulations to

1:03

Sotrini. Welcome. Congrats on the great

1:05

performance. Thanks a lot, Jack. I

1:07

appreciate it. Let's dig into this performance. So

1:09

some of that performance has come from the

1:11

AI beneficiaries, Nvidia, SMCI, that

1:14

kind of stuff. Some of it

1:16

came from GLP ones or, you

1:18

know, Ozempic type investments, which

1:20

we've not discussed really on either podcast, but

1:22

you did discuss on the the Oddlaws podcast

1:25

last year, which was excellent. Some

1:27

of them came from fiscal beneficiaries, which we

1:29

will be talking about today, and particularly how

1:31

it comes to the upcoming election. And a

1:33

lot was from interest rates, interestingly.

1:36

But Sotrini, I just want to say

1:38

when it comes to the overall

1:41

market right now, how are you

1:44

feeling you have been a into

1:46

fattagal bull, which has worked out

1:48

quite well. But you wrote on

1:50

Twitter recently, actually last

1:52

night, you wrote that the vibes are

1:54

starting to shift. What do you mean by that? The

1:57

thing about the environment right now

2:00

is we've kind

2:02

of hit this point

2:04

where we're going to find

2:06

out, right? We've gone back and forth. Is it going

2:08

to be a soft landing? Is it going to be

2:10

a re-acceleration of inflation? Is it going to be a

2:12

recession? If anything, you

2:14

know, that constant wall of worry

2:17

has driven returns,

2:19

right? And now we're in

2:21

a situation where I think over

2:24

the next six to 12 months, we're

2:26

going to have our answer,

2:28

right? Are we going

2:30

to be able to pull off this soft landing? Or

2:33

are we going to overshoot to the

2:35

downside and end up with an economic slowdown? I

2:37

think the reason why it's tricky now is because,

2:39

you know, you're starting to

2:42

get some on earnings calls,

2:44

you know, they'll say, you know, like

2:46

Toyota Motor, for example, had

2:48

some cautionary stuff about auto loans and delinquencies and,

2:50

you know, the unemployment rate is taking up a

2:52

little bit. The household savings rate isn't, you know,

2:55

I think when it comes to

2:57

macro, I try

2:59

my best to not

3:02

make decisions when the data is unclear.

3:05

And right now, is

3:08

this just a classic summer slowdown? Is it,

3:11

you know, something where this is kind

3:13

of the necessary, quote unquote,

3:15

Goldilocks situation where we get like a

3:17

little bit of economic weakness, but ultimately

3:21

that's what's necessary to bring inflation back

3:23

within a reasonable level? You

3:26

know, I don't know right now. But what's served me

3:28

pretty well over the past three

3:30

years is I don't

3:32

make decisions on macro until

3:34

it's painfully obvious.

3:38

For example, in 2022, you know, inflation in

3:40

October 2021 was going higher. And

3:45

there was all this debate about, you know, is it base

3:47

effects? Is it transient this, that,

3:49

and the other thing? It wasn't until like

3:52

the Fed really was, yeah, we're

3:54

going to have to hike rates off the zero lower

3:56

bound. That's when I said, okay, you know, let me

3:58

start really getting concerned about I

4:01

think the same thing applies here with the

4:03

recession stuff. When the Fed starts cutting, yeah,

4:06

then that'll be something where I'm going to spend

4:09

a lot of time saying, is

4:12

this going to become something that I

4:15

have to really be concerned about? But

4:17

for now, I don't want to get thrown

4:19

off by summer slowdown.

4:22

I don't think there's much to do until

4:24

November, if we're being honest. It's an

4:26

election year. So the

4:29

environment right now, I would say, is not something

4:31

that I'm too concerned about, but I am starting

4:33

to get a little bit attentive to the risks.

4:35

I want to do my best Jerome Powell impression.

4:39

Attentive to the risk, yeah. That's good. So

4:41

you don't make trades primarily based on macro.

4:43

I would say you are an investor who

4:46

looks at the fundamentals of a

4:48

theme. You dive very deep into the

4:50

weeds. I guess compared to

4:52

analysis you'll see on TV or in some bank

4:55

sales types reports, you get in the weeds. Not

4:57

the most in the weeds first enough, by your

4:59

own ambition. And

5:01

then you construct a basket of it. So right

5:03

now your baskets are your AI basket, your

5:06

fiscal beneficiaries basket, your

5:08

GLP1 bazented basket, let's just call it that,

5:11

and a handful of other baskets that are

5:13

smaller. So it sounds like

5:16

you are not the

5:18

most bullish you've ever been, certainly not as bullish as you

5:21

were a year ago. You're attended to

5:23

a macro slowdown, and macro can always be a threat

5:25

to your core longs, but you're not

5:27

getting wildly bearish. Is that correct? Yeah,

5:29

that would be an accurate way to put it. I

5:31

think what I'm more focused on right now

5:34

rather than macro would be how these

5:37

kind of mega

5:39

trends are progressing and kind of

5:42

focusing on narrowing

5:44

down for their

5:47

evolution. If you look

5:49

at something like AI, when chat GPT

5:51

first came out, you could

5:53

look at AI and everyone

5:56

did this, right? You look at AI and you

5:58

say, well, I know what it looks like. like

6:00

now, but if I kind of

6:02

blue sky and say what it's gonna look like in five

6:04

years, well, oh my God, maybe this

6:06

is gonna put everyone out of work, you know, or,

6:08

and, you know, the thing is,

6:10

I think the best way to have played that

6:13

was, like I said, you know,

6:15

with basically AI is going to be a thing. And

6:18

in order for that to happen, you are going

6:20

to need a data center build out, and you're

6:22

going to need a lot more GPUs. That was

6:24

the quote unquote, picks and shovels thesis. And

6:27

that has played out, right? I don't

6:29

think that there's anyone that denies that

6:31

that's played out. What you needed there

6:33

was for hyperscalar capex to kind of

6:35

skyrocket so that they could compete in

6:37

AI. And that's what happened. And you

6:40

look at some of the areas

6:42

where fundamentals got a little

6:44

too disconnected from reality, like super micro,

6:46

for example, that thesis played out, and

6:49

then it was time to get out

6:51

because the theme had progressed. And I

6:53

think it's just so interesting, right now,

6:55

when you're looking at what is really

6:57

the next phase of this going to

6:59

look like, you know, I would say

7:01

we're in like phase one B right

7:03

now, where there's obviously still some upside

7:06

on some of these

7:08

picks and shovels, so to

7:10

speak plays. But we're also getting

7:12

into the area where, you

7:15

know, what is this actually going to be used for? I think the

7:17

apples, WWDC was, I

7:20

mean, I was blown away. And I think

7:22

that that really kind of informs what the next

7:25

phase is going to look like. It

7:27

appears to me that over

7:30

this AI bull market

7:32

that let's say began with

7:34

the Nvidia reveal of 2023,

7:38

that there's the real

7:40

beneficiary list of beneficiaries

7:42

into whose earnings have, you

7:44

know, gone up a tremendous amount. There's

7:47

a handful of names. And I really don't know

7:49

how many I could come up with other than

7:52

Nvidia. I mean, maybe see MCI, I don't

7:54

know their earnings, but I mean, Nvidia, their

7:56

earnings, and this is what people

7:59

say and be is a bubble, it

8:01

may be, but people cannot

8:03

deny, and if they do deny, I

8:05

think they're wrong. Just the tremendous revenue

8:07

increase from $7.2 billion in Q1 2024, which really means 2023, to

8:14

26 billion of the most recent quarter in

8:16

Q1 2025, fiscal

8:18

year 2025. So that's an increase of 3.6

8:20

times or 260%. And

8:25

the stock is up more than that,

8:27

but the stock is priced again, that's

8:29

it can, and maybe it can continue

8:31

to grow. So, and Nvidia's forward earnings,

8:34

I'm just gonna say it's something like 40 or 50. Interestingly,

8:38

last year, people were saying,

8:40

Nvidia is trading at 300 times earnings, and

8:42

it was based on the trailing earnings from the cyclical

8:44

downturn, when we had the slowdown of 2022.

8:49

And like relative to those forward

8:51

earnings, Nvidia at the time must have been trading at

8:53

a PE of 15, which is

8:55

probably cheaper than the value stocks that they like, I don't

8:57

know. Well, I

8:59

don't wanna be too antagonistic, right? Because

9:04

the first time that I was

9:06

on, I had a big put, those YouTube

9:08

comments, definitely kept me in the

9:10

trade for longer than I would have otherwise

9:13

been in, because it was, you

9:15

couldn't imagine a better environment to generate kind

9:17

of eye popping returns in the clear winners,

9:19

right? Most of the gains were accrued to

9:21

Nvidia, and Nvidia suppliers

9:24

and hyperscalers. And when

9:26

you get into a theme, you

9:29

have like a fundamental narrative, and then

9:31

you have a sentiment tailwind, and the

9:34

first thing you focus on is kind of that fundamental story.

9:38

But as far as Nvidia

9:41

goes, I mean, I think

9:44

that using the term bubble,

9:48

it's much more like a mania, right?

9:50

Because you look at railroad

9:52

stocks in the 1800s, there's

9:55

a great chart out there from an academic

9:57

paper the dividend yield

9:59

basically. of these stocks

10:01

because most of them were preferred shares back

10:03

then. And it basically

10:05

represents earnings growth and that went

10:08

up commensurate with the price of

10:10

the stock. And so

10:12

I think there's a difference between a bubble

10:14

where things just get completely fundamentally disconnected and

10:16

a mania. And obviously a mania can progress

10:18

into a bubble. I don't think

10:21

that we're going to see something where Nvidia's

10:23

earnings kind of level out and

10:25

the stock continues to go up.

10:27

So I think that what's

10:30

the most important thing you can do here

10:32

in AI looking forward is basically, like I

10:34

said, when I started, I had 130 securities

10:36

that I was putting

10:39

on and saying, this

10:42

could benefit from AI and then it progresses

10:44

and you end up deleting some

10:46

or adding some and kind of

10:48

narrowing it down as you learn more

10:50

about what's going on. And like

10:54

a good example for

10:56

the first phase, kind of like build

10:58

out, you know, TSMC

11:00

has had this kind of

11:02

overhang or I

11:04

guess like a litany of overhangs. You

11:06

have, you know, geopolitics, you

11:09

have the fact that

11:11

it's kind of lower margin, you have

11:14

their diversification and how exposed they are

11:16

to the automotive cycle. But

11:19

as everything progresses and as Nvidia is

11:21

more successful, the TSMC AGM was

11:23

a couple of weeks ago and

11:25

the CEO basically said, you

11:27

know, given Nvidia's success, we can probably

11:30

be charging a little bit more. And

11:33

yeah, and the surprising thing

11:35

was, you know, I

11:38

mean, however you would expect Nvidia

11:40

to react to that. Now

11:43

Jensen came out, I think

11:45

the next day and said, yeah, no, they

11:47

should be. And, you know, so I

11:49

think you kind of look at some of these

11:51

names that have had, they have AI tails, but

11:53

they have kind of been depressed a little bit,

11:55

like by external factors, and you wait for those

11:57

external factors to kind of get a little better.

11:59

This is the same thing. The last time I

12:01

was on, we spoke about

12:04

Micron and

12:06

the high bandwidth memory requirements

12:08

for artificial intelligence.

12:10

That was the same

12:12

story where you had the memory cycle was

12:15

not doing great and you had a big

12:18

overhang from that. But the second that that

12:20

started to inflect a little higher, you got

12:22

this huge outsize move because the

12:24

AI tail was front and

12:26

center. So I think that you can

12:29

see that on TSMC and maybe to

12:31

be a little bit more controversial, I

12:33

think that it might be time

12:35

on Intel. Just saying that, I feel

12:38

like, oh, God, this is going to risky

12:41

things to say. But I look

12:43

at Intel's pipeline and one of the big

12:45

things that has kind of become a central

12:48

narrative here is the power

12:50

requirements for AI. And

12:54

as like a student of past

12:57

financial manias and popular

12:59

delusions, you can

13:01

go back and see sell side research

13:04

from the year 1999

13:08

during kind of the personal computing revolution.

13:10

And what you see there is they

13:13

were saying, well, if everyone in the

13:15

US has a computer, then

13:18

the power that's necessary for that is going to

13:20

be immense. And we're going to see electricity demand

13:22

grow at a 13%

13:24

kegger over the next 10 years. And electricity

13:27

demand from 2000 to today is pretty

13:29

much flat. And I think that

13:31

that's because that kind of analysis discounts

13:34

the innovation, right? Where

13:38

if a technology becomes intrinsic to day

13:40

to day life, what you'll see is

13:42

innovations that make it possible to have

13:45

the capacity to use that technology. And I

13:47

think that's probably what will happen on the

13:50

power side here. You have

13:52

something like this, which, you

13:55

know, I played on the way up, because I'm

13:57

definitely not someone who shies away from just

13:59

like solely. narrative driven price. But

14:01

the idea that something like Intel,

14:03

they're the only foundry that's

14:08

going to be doing, if they deliver

14:10

on their promises, which they haven't historically been great

14:12

at, but they're the only one that's going

14:15

to be doing what's called ribbon-fed and

14:17

backside power delivery on the same chip. You

14:19

get into this calculating what's

14:21

called flops to watts, basically

14:23

how much compute can you do for the power you're using.

14:26

And I think Intel is pretty

14:28

well positioned there. And you look at the stock and it's

14:30

all a margin of safety thing. You have

14:32

a narrative that could materialize and then you have

14:34

a reasonable margin of safety.

14:36

And I think that those

14:39

two, that's what I'm trying to

14:41

do. I'm trying to narrow this down to places

14:43

where I don't have kind of this

14:46

asymmetric risk like I would have if

14:48

I was still holding Super Micro. Got

14:50

it. So you said TSM, TSMC, that

14:53

is a Taiwan semiconductor, a world's

14:55

leading producer of semiconductors and they actually

14:57

are foundry. They make the chips unlike

14:59

Nvidia, which is a fabulous thing.

15:02

They designed them. They do not make

15:04

them. Is Intel kind of just like

15:07

an American Taiwan semi that is perhaps

15:09

slightly less competitive? I mean,

15:11

I'm sure if Pat Gelsinger heard you call Intel an

15:13

American Taiwan semi, he would be real happy. But

15:17

essentially, that's one

15:19

way of putting it where even though it's an

15:21

election year, I've been focused on

15:24

what are the priorities going to be regardless

15:26

of who wins the House, who wins

15:29

the Senate, who's the president. And

15:31

something that is pretty

15:34

obvious from both sides is basically

15:36

chips. You

15:38

could make a lot of heated debate

15:40

over who has

15:42

better technology, who is better in this

15:45

area or that area with the US

15:47

and China. One area where

15:49

it's kind of unequivocal is AI and machine

15:51

learning. Or America is leading. Yeah,

15:53

absolutely. But it's still in a

15:55

way where Nvidia doesn't exist without

15:57

TSMC. But

16:00

it's something that we

16:02

absolutely want to

16:04

have very advanced boundaries that

16:07

are located in America. So

16:09

I think that there is also

16:11

kind of a fiscal tailwind for

16:13

Intel, regardless of who wins. And

16:17

so that's another kind of

16:19

margin of safety there. Got

16:21

it. And people who subscribe to

16:23

their sub stack, which I recommend people do

16:25

check out, and they actually see your baskets,

16:28

they will learn, as I did, that you're

16:30

a lot more diversified and you have a

16:32

lot more names than people who just follow

16:34

you on Twitter or listens to this podcast

16:36

or the OddBots podcast may see. It's not

16:38

just like, Nvidia 25%, SMCI 25%, you were

16:40

very diversified. How

16:44

much dispersion, is that

16:46

the fancy word? Have you seen between

16:48

the names? So for example, I know

16:50

in the S&P, it has been a

16:52

very narrow rally, unlike the rally of

16:54

2021, when it was the small cap

16:57

speculative stocks leading the way, it really has been

17:00

Nvidia, Meta, Google,

17:03

other giants leading the S&P

17:06

500. And the Russell, the

17:09

small cap stocks is not doing great at

17:11

all. Your comments on that one. And two,

17:13

have you seen that within your names where

17:15

it's, I thought that each of these names

17:17

would be up 15%, but really, most

17:20

of them are flat. And one of them is

17:22

up a ton. Not really. I

17:25

think that there are certainly names

17:27

that didn't perform as expected.

17:30

As far as the diverse, just to, this

17:32

is kind of a two part question, but

17:35

as far as diversification goes, I kind of

17:37

view it as being diversified

17:39

on a security level, but concentrated on a

17:41

thematic level. So the way that I think

17:43

about it is essentially, I have,

17:47

it's varied over the past year, but between 10 and

17:49

20% of my book in artificial

17:51

intelligence, right? What

17:54

I want to avoid is essentially

17:56

a theme that moves so fast, like AI

17:59

does. And then there's a

18:01

development where it significantly

18:04

affects single security that I'm

18:06

concentrated in. And

18:08

that doesn't mean that I really

18:10

shy a ton away from concentration.

18:12

A Celestica, SMCI, AOI, at one

18:15

point, these made up anywhere from

18:17

2% to even 10% of the

18:20

portfolio in a single name. But that was

18:22

not because you bought 10% of it. It was because

18:24

it started off as a 1% or 2% position and it went up

18:27

so much, correct? Yeah, exactly. The

18:29

idea that Nvidia has been

18:32

kind of like the sole beneficiary, I

18:34

can see where people might say that if they're

18:37

looking at the S&P and then you get the

18:39

whole S&P 493 versus the

18:42

S&P 7. But

18:44

there have been plenty of good returns to be

18:46

had elsewhere. If

18:48

I look at like a waterfall of the

18:50

contribution rate of every name

18:52

in this basket, Nvidia obviously is

18:55

pretty up there. But

18:57

it shares that spot with a lot

18:59

of other names like Verdiv,

19:03

for example, Credo,

19:05

Orista Networks, Oracle,

19:08

Amazon, Amphenol,

19:10

Fabernet. This is the point of

19:13

the strategy that I use, which I

19:15

think people kind of are

19:18

taken aback by how many securities there are.

19:22

The thing that I'll always say is if

19:24

you own the S&P, you own 500 names. And

19:27

if you can just

19:29

own 300 and end up

19:31

outperforming, that's pretty

19:34

good, especially when it makes it so that

19:36

your drawdowns are not life threatening.

19:39

Yes. And so that is

19:41

important consideration that you are diversified in

19:44

number of securities, but not thematically. And

19:46

so, yeah, if someone owns a lot

19:48

of real estate in China, it's not

19:51

all of their hopes are on one property, but if

19:53

you own apartment buildings in Shanghai and in Beijing,

19:55

they're correlated. And Broadcom and Nvidia

19:58

have been correlated on the way up. up

20:00

and if you go down, you know, these

20:02

names probably will go down together too. Four

20:05

guidance is brought to you by

20:08

VanEck. The VanEck Morningstar-wide Moat ETF

20:10

ticker MOAT has outperformed the S&P

20:13

500 for over a decade. How?

20:15

Moat strives to achieve a simple

20:17

but challenging task. Buy quality stocks

20:20

when they're undervalued and sell them

20:22

when they're overvalued. Visit vaneck.com/MoatFG to

20:24

learn more. That's vaneck.com/MoatFG. Now

20:27

the disclosures. All investing is subject to risk,

20:29

including the possible loss of money you

20:31

invest. Visit vaneck.com to carefully read a

20:33

prospectus before investing. The VanEck Morningstar-wide Moat

20:35

ETF is distributed by VanEck Securities Corporation,

20:38

a wholly owned subsidiary of VanEck Associates

20:40

Corporation. Thanks. Let's get back to the

20:42

interview. Any final comments,

20:44

Sauterny, before we move on to the

20:46

main topic of our discussion, which will

20:48

be, drum roll please,

20:50

the election. Yeah. Well, you

20:53

know, I figured, right, that since

20:55

everyone was so kind of angry

20:58

about me being super bullish on AI the

21:00

first time that I came around, that we

21:02

could talk about something a little less controversial,

21:04

like US politics. All right. Let's do it.

21:06

And with the caveat that I,

21:09

on this program, don't talk

21:12

about politics unless it impacts the

21:14

financial market, because I know, you

21:16

know, some part of my audience is going to be

21:18

on one side and another is going to be on

21:20

the other side. And if they met each other and

21:22

started talking politics, they probably wouldn't agree

21:24

with each other very much. But they do agree

21:26

with me when they're listening to me, you know,

21:28

about listening to Joseph Wang talking about quantitative tightening, or

21:31

about the construction of the CPI

21:34

index. So I think that I

21:36

intentionally shy away from politics. But

21:40

in this topic, it is

21:42

coming up and it undoubtedly will have

21:44

an impact on the fundamentals of markets,

21:46

definitely, and probably on the short term

21:48

price action, probably not definitely on that.

21:51

And you really are perfect to address

21:54

that because that you say you have

21:56

no political beliefs whatsoever. You only care

21:58

about how you're the only on

22:00

how the election will impact

22:03

different types of securities, different

22:05

currencies, interest rates. And

22:09

it is very speculative, but if

22:11

Trump wins, what asset is going to do well? You've

22:13

created a basket. If Biden wins, what's going to do

22:15

well? You've created a basket. Tell

22:17

us about your process in how

22:20

you even constructed these baskets. The

22:22

reason why I try to focus so

22:25

much on policy is because I think that the

22:27

idea of getting an edge from just when

22:30

you look at, for example, the Inflation Reduction Act

22:32

or the IIJA or the CHIPS Act, there

22:35

is a real discernible

22:37

edge in just reading those documents

22:39

and finding out kind of all

22:42

that money that's spent, it

22:45

ends up going to someone, right?

22:48

And a lot of times those,

22:50

some ones are publicly listed companies

22:52

and there's a direct impact on

22:54

their earnings. So a good

22:57

example that I use is you have a

22:59

company called Encore Wire. And then

23:01

in the Inflation Reduction Act, there were a

23:03

lot of green tax

23:05

credits. And part of

23:07

those green tax credits was something

23:10

called the Domestic Component requirement, where

23:13

if you're building a efficient

23:15

building, when, as you can see, infrastructure

23:18

investment has performed

23:21

a lot better than most other

23:23

areas, all these buildouts, you're

23:26

gonna need to put in American-made

23:28

components. And that goes beyond just

23:30

assembling the component in America and

23:32

putting it in. It

23:35

also includes basically sourcing from

23:37

companies that are doing

23:39

this in America. So you look at the

23:42

options for basically copper wire, and

23:44

the biggest one that produces their copper wire

23:46

in America was Encore Wire. So if you're

23:48

gonna build a efficient building, you're gonna need

23:51

a lot of copper wire. And

23:54

you're only gonna get the tax credit if

23:57

you buy from a domestic component requirement.

24:00

company. And so you look at how

24:02

Encore Wire reacted after the Inflation Reduction

24:04

Act text came out and there was

24:06

a delay basically while

24:08

people began to understand this. So I

24:11

focus a lot on policy because you look

24:13

at how Encore Wire has performed. I think

24:15

it's up 150, 200%

24:18

since then. And with an

24:20

election year, I just think

24:22

that... Listen, forecasting elections

24:24

is a lot harder than forecasting stocks.

24:27

And I'm not

24:29

someone that's going to tell you who I think is going

24:32

to win the election, but I

24:34

come in with kind of a Boy Scout mentality

24:36

of just be prepared and be

24:39

ready for whatever

24:41

pales might surface. And I look at

24:45

the election in

24:48

essentially two phases where

24:50

in this kind of first stage,

24:52

you have the market reaction to

24:54

the rhetoric from presidential promises or

24:57

debates and what politicians

24:59

say they're going to do. And that

25:01

has a lot of market moving impact

25:03

because people want to be positioned for

25:06

that. And then you have

25:08

the actual outcome and how

25:11

that will affect markets. So what I've done

25:13

is essentially created two strategies.

25:16

And the first one is

25:18

playing anticipation of the election

25:20

and playing what's

25:22

likely to be said or what has been said, how people

25:25

will react to when I

25:27

wrote it, rising Trump odds. Typical example there

25:29

was Trump came out one day at one

25:32

of his rallies in February and

25:34

said, if you're

25:36

a NATO member and you're not paying

25:38

your fair share, Russia can do whatever they want

25:40

to you. And the

25:43

idea that forcing

25:45

these kind of NATO companies to meet

25:48

their obligation to spend 2% of their GDP on

25:50

defense, that was

25:53

a huge delta, represented a huge

25:55

delta for European defense companies where

25:57

countries were absolutely not paying that.

26:00

So for a while, that was my

26:02

way of playing the election, just

26:05

long European defense companies. And

26:07

then as the election drew closer, I

26:09

started saying, OK, let me kind of

26:11

separate this out into where I think

26:14

there could be outperformance as we get closer to

26:16

the election. And then on the other side, there's

26:19

the actual impact of whoever ends up

26:21

winning. And the two areas that I

26:24

think that that really will show itself

26:26

is basically a trade policy and

26:28

the tax policy. So

26:31

that's pretty much how I've approached this so

26:33

far. So you

26:36

have a bunch of different

26:38

buckets within your election

26:41

basket. So for example, on the defense

26:43

side, if Trump wins, that's the

26:45

basket that benefits from Trump winning would be long

26:47

European defense stocks and short US defense stocks. Then

26:50

there's winners and losers

26:53

from increased tariffs. Then

26:55

there's bets on

26:57

Trump specific securities

26:59

such as his SPAC, Fannie

27:02

Mae, Freddie Mac, which some

27:06

investors think that Republicans are more likely to

27:08

deregulate that or repravitize Fannie and Freddie, which

27:10

got taken over in 2008, of

27:13

course. I think Republicans are

27:15

more likely to do it to Democrats. The question is, are they

27:17

likely to do it? Trump didn't do it in his first time.

27:19

Yeah, I don't actually care if they actually do it. The

27:23

thing is basically just playing the expectation

27:25

of it. This

27:28

is basically a basket that's just solely

27:31

the purpose is just to play election

27:33

expectations. If

27:36

Trump's odds are rising and there's a

27:38

higher chance of Republicans doing it than

27:40

Democrats, I

27:44

actually don't think that they'll do it either. But that

27:47

doesn't mean that if there's a higher chance,

27:49

then the securities probably should price that in. Stuff

27:52

like the regulation of

27:56

the financial sector, mergers and acquisitions in

27:58

a Trump scenario. of

28:00

immigration, that's kind of a classic one, you know.

28:03

And then the other thing

28:05

is basically, I think that no matter who

28:08

wins, there's gonna be some sort of protest

28:10

or whatever. So I found

28:12

a name that basically makes the storefront

28:15

glass for like retail. And

28:18

that seems like that'll do well if there are protests. Because

28:20

the glass will break and the store will have to buy more

28:22

glass. Yeah, I think so. But,

28:25

you know, the really interesting one, I think,

28:27

is kind of onshoring versus

28:30

nearshoring. If you

28:32

look at kind of the way that

28:34

the Democrats have approached the

28:36

pivot away from China versus the

28:39

Republicans. The Biden administration has definitely

28:41

favored nearshoring, especially to Mexico.

28:43

And I think that the Trump administration

28:45

would probably view that unfavorably because of

28:47

the negotiating leverage that it gives Mexico.

28:49

And, you know, their stance on illegal

28:51

immigration would also factor in. But,

28:54

you know, onshoring, the

28:56

idea of not shifting from using

28:59

one emerging market to another, but actually

29:02

bringing production domestically, I

29:04

think that really that's only feasible for a

29:07

first world country if you make significant progress

29:09

in automation. You

29:12

know, offshoring or nearshoring has solely been

29:14

driven by labor cost differences. And

29:17

you're not gonna get that benefit if you try

29:20

to, you know, onshore production of whatever it is,

29:22

you know, vehicles or pre-gas concrete or whatever

29:24

that is. So you look

29:26

at kind of like this, the way nearshoring is played

29:28

out in Mexico and, you know, it's been great for

29:30

them. But if Trump gets

29:32

in, I think that what we'll

29:35

see is probably a lot more

29:37

spending on kind of industrial automation, equipment

29:40

providers, advanced manufacturing. And

29:43

you can see a lot of America

29:45

first kind of policies about, you know,

29:47

supply chain resilience and trade tensions and

29:50

get some losers in kind of, you know,

29:52

Mexican manufacturing and logistics. And then on the

29:55

macro side, I think there would be some

29:57

implications obviously there for inflation. I'm not gonna...

29:59

go too far into that. But the one

30:01

macro view that I'd say I'd have on

30:03

the election, I think that if Trump gets

30:06

in, we'll probably see

30:08

one of the most epic two's ten steepeners

30:10

in the history of two's ten steepeners. So

30:12

you had an epic steepening in two's tens,

30:15

the spread between the two year treasury, the

30:17

ten year treasury, the ten year and the

30:19

two year. Now it's inverted two years higher

30:21

than ten year. You think that would change?

30:23

Why? If Trump got elected? I

30:26

think basically what you'd

30:28

see is, I don't know if you've

30:30

read Nick Timorellos' book Trillion Dollar Triage.

30:33

Oh yes, it's great. Yeah, great

30:35

book. And really rare opportunity to have the

30:37

sitting Fed share and then have such an

30:39

extensive look

30:42

into his thought process and

30:44

his values. And that was

30:46

really very helpful in the

30:49

fallout of Silicon Valley

30:51

Bank and the regional banking, whatever

30:53

you want to do, through a crisis, whatever.

30:55

Because you could basically go through that book

30:57

and control F like banking crisis. And

31:00

you would see what Powell thought during the,

31:02

I think it was the savings and loan

31:04

crisis where Powell had

31:06

this mentor blanking

31:09

on the exacts of it, but he was

31:11

very pro not letting the banks fail. And

31:13

so it really, if you haven't read it,

31:15

I would say it's definitely something that you should take

31:18

a look at. But I

31:20

think that one of the interesting parts

31:23

of that was this contentious relationship before

31:25

COVID, obviously of Trump and the Fed.

31:27

And that's not a secret to anyone,

31:30

but I do think that the

31:32

political pressure for the Fed to lower interest

31:35

rates would be extremely

31:37

significant. And I do

31:39

not think that there would be some degree

31:41

of caving on that. And at the same

31:47

time you have Trump's policies, which

31:49

are basically cutting

31:51

taxes without cutting spending.

31:54

And you have this kind of trade

31:57

policy tariff stuff. It's all

32:00

pretty inflationary stuff and obviously

32:02

dependent on whether the economy

32:05

can absorb that. But I do think that

32:07

in the beginning, what you essentially have is

32:09

the bond market

32:11

saying, well, if that's the

32:13

case, then why wouldn't I just

32:15

own the front end? So

32:18

I think what you could end up getting is basically significant

32:20

drop in rates on the front

32:22

end and with rates on the long

32:25

end staying the same or even going

32:27

up, kind of a bond market revolt.

32:29

So I think that that's like my

32:31

one, I don't really have a view

32:33

on like, how does the dollar do

32:35

if Trump gets in or Biden or

32:37

whatever? But as far as the

32:39

yield curve, I think that if Trump wins,

32:41

we get a pretty significant speed on me.

32:43

The US fiscal deficit was

32:47

large. Ever

32:49

since 2002, under Obama, 2008,

32:52

the economy clearly needed a

32:54

giant stimulus. And

32:56

as a Republican, when he elected in 2016,

32:59

ran a large post-cyclical deficit with tax

33:01

cuts, then a huge amount

33:04

of sub-sypsil stimulus in 2020, again, under Trump.

33:06

But of course, that was for the reason

33:08

of the pandemic. Biden is

33:10

also running a very large, smaller, but still very

33:12

large. And again, you could say that maybe

33:15

Biden's deficit now is

33:17

more large than Trump's was

33:19

with the consideration relative to the circumstances because we're

33:22

not in a, you know, you have unemployment rates

33:24

not at 20% now in the same way that

33:26

you could say Trump's first deficit in 2016. That

33:30

was a deficit when the economy was expanding,

33:32

unlike when Obama got elected. And clearly there

33:34

was a need for massive fiscal

33:36

stimulus because the economy was in a deep recession

33:39

and getting into it to a worse one. So

33:42

what is your outlook now on how

33:45

large the fiscal deficit will be under

33:47

Trump and Biden? And a reminder, viewers,

33:49

a fiscal deficit is just the

33:52

difference between how much the U.S. government

33:54

spends and how much it takes in

33:56

in taxes and tariffs since

33:58

2000 and 2001. the

34:01

US has been running a deficit. And

34:04

really since the 1950s, most of

34:06

the time the US is running a fiscal deficit

34:08

where a supplier of dollars, the world's reserve currency

34:10

to the rest of the world. But the deficit

34:12

has been increasingly large. And of course the deficit

34:14

every year gets tacked onto the debt. So just

34:16

explaining that for the audience, what is your outlook

34:19

on how large the

34:21

deficits will be under each

34:24

president, whether they're elected? You

34:26

said something interesting before I answer

34:28

that question, but the idea of when

34:31

the trickiest part of this election is

34:33

essentially that it's

34:35

a rematch, right? And that

34:38

kind of makes it very tempting to just go

34:40

and back test, well, how did stocks do when

34:42

Trump was president and how did stocks do when

34:45

Biden was president? And then call it a day.

34:47

And that's not gonna work. First

34:52

off, obviously the economy and external

34:54

factors are much more important than

34:57

whoever's president. The places

34:59

where election kind of focus, strategy will work

35:01

is gonna be more in the micro side,

35:03

I think, than the macro side,

35:05

just because you wanna be very specific

35:08

about the areas where things

35:10

will be different. This kind of gets

35:12

into the idea of not

35:14

playing the expectations surrounding the election, but

35:17

playing the actual outcome of the election.

35:20

And that

35:22

has a lot more to do with, like

35:24

I said, trade policy and tax policy. And

35:27

I think that we have

35:29

entered firmly into a dirigiste

35:32

bipartisan consensus, right? There is a- What

35:34

does that word mean? You use it

35:36

in your speech, what does that word mean? So there's

35:38

two sides, right? Lazé faire

35:41

and then dirigiste, right? And

35:43

lazé faire in

35:46

French means kind of like let alone, just

35:49

like let the market do

35:51

what it will. And that used to be kind of the

35:54

policy of the GOP, where

35:57

state intervention for curbing.

36:00

having inefficiencies or whatever,

36:02

doing counter-cyclical stimulus.

36:05

It was much more kind of let the market sort

36:08

it out. And obviously

36:10

there's still differences in how

36:12

these parties approach it, but TRGs

36:15

is kind of the opposite of

36:17

that, where you have a very

36:19

involved government that has like indicative

36:21

planning, government credit policies, subsidies,

36:24

kind of fostering the

36:27

areas where you think there's going to be the most

36:29

productivity gains. And both

36:31

sides are profligate spenders.

36:36

There's the era of

36:38

like the conservative fiscal policy, I

36:40

think has been disrupted

36:43

by the paradigm of

36:45

the US increasingly kind

36:47

of being a little more populist, where

36:50

you spend money because you

36:52

want people to vote for you. And

36:54

that's, I don't know if that's

36:56

cynical or whatever, maybe that's

36:59

my own political bias in the sense

37:01

that like I don't favor either party,

37:03

but it is something that I

37:05

think we have seen very significantly. And

37:08

I think that the deficit, that's

37:10

one area where it might

37:12

be a little larger under Trump, but ultimately

37:16

the offsets of like, the tax policy

37:18

differences, one of the first things that's

37:22

definitely on Trump's agenda will

37:24

be extending the TCJA, which

37:26

is the- Tax Cuts

37:28

and Jobs Act. He has a very

37:31

kind of aggressive tax policy where you're

37:34

going to do tax cuts and you're cutting

37:36

the corporate tax rate and you're extending

37:38

this kind of, you know,

37:40

like 100% bonus depreciation in the

37:43

first year. And then

37:45

on the Biden side, it's a lot

37:47

more aggressive where you have raised

37:49

the corporate tax rates at

37:52

28%, quadruple the stock buyback tax to

37:54

4%, eliminate tax subsidies

37:56

for fossil fuel and real estate

37:58

and executive comp. And then

38:01

you had Janet Yellen,

38:03

there's the global

38:06

minimum income tax for corporations

38:08

guilty, GILTI. It's

38:11

currently under the TCJA, there

38:13

are incentives to repatriate foreign

38:15

income, but that expires in

38:17

2025. So Yellen

38:19

has expressed hope that

38:22

Congress will basically admit

38:24

us to these 140 countries signatories where

38:26

no matter what you do, like

38:29

for example, big pharma, they have a tax

38:34

loophole that I think is called

38:36

single malt, which is basically hiding

38:38

your intellectual property kind of revenues

38:42

in Ireland that's owned by a company

38:44

in Malta. I don't know, I'm not

38:46

an accountant, but there are all these

38:48

loopholes where you can kind of shift

38:50

overseas and that would be the

38:53

tax burden on companies would be pretty significant.

38:55

So what I've done is I've looked at

38:59

what were these companies effective tax rates between the time

39:01

when the TCJA came in to 2018 just to eliminate

39:03

the COVID kind of, that's the thing

39:08

if you're analyzing these two things, you really have to

39:10

just take out COVID because this was the biggest bipartisan

39:12

thing ever. So

39:15

before the CARES Act basically, and

39:17

then examining who were the biggest

39:19

beneficiaries of kind of

39:21

the Trump tax plan versus who were the

39:23

biggest beneficiaries of Biden's tax plan. And

39:26

who was hurt and by either. So I

39:28

think that that's kind of an area where

39:31

you can really anticipate

39:33

well because it's pretty clear what they're going

39:35

to do from a tax policy perspective. And

39:38

while large cap companies don't really trade on

39:40

after tax earnings, it does make an impact

39:42

of how much money

39:44

people have and how much money these

39:46

companies have to kind of invest and

39:48

do capex and do R&D. I think

39:50

that when you're talking about

39:53

forecasting the deficit, I

39:55

don't really know how to do that as a play out. Does

39:57

Trump's tax policy make it so that

39:59

since there's... There's more money to invest

40:01

that we get more productivity and the

40:03

government gets more tax revenue and or

40:06

does does cutting taxes while

40:09

spending the same amount, which I'm pretty sure he

40:11

will. Does that increase

40:14

the deficit? I don't really know. But

40:16

what I do know on the tariffs

40:19

and tax policy, those are two places

40:21

where whenever the outcome comes out, that's

40:24

what I'm going to be playing. I

40:27

don't know if you've seen the

40:29

tariffs that are proposed by Trump, but they're

40:31

pretty aggressive. It's like a 10

40:34

percent tariff on anything that's not US made

40:36

and then like 60 percent tariff on some

40:38

Chinese goods. All of this

40:40

kind of informs also how

40:42

I want to be positioned in China, right? Where

40:45

I don't know if you saw recently

40:48

Michael Pettis on ... Did you see that?

40:52

I did, yes. He basically said

40:54

that we're nearing the third plenum

40:57

and we are

40:59

increasingly likely to see Chinese

41:02

consumer stimulus, right? Like

41:05

stimulating consumer demand, basically

41:07

like helicopter money or probably not

41:09

as extensive as what we did

41:11

during COVID. But the general idea

41:13

is basically you have this situation

41:16

where consumers aren't levered

41:19

up that much. They're spending that

41:21

much and the household savings high. So

41:24

you kind of incentivize that spending. I

41:28

think that that sets up pretty nicely where even

41:30

if Trump wins, in China it seems like you

41:33

want to be positioned for the kind of

41:35

localization of the Chinese economy. The

41:39

idea that if they are going to do that ... If

41:42

the stimulus materializes and it looks a lot

41:44

different than the previous

41:46

Chinese attempts at stimulus, which

41:49

obviously did not end super well,

41:52

you can basically create a

41:55

reasonable allocation. This is the interesting

41:57

part of how I run, or at least this interesting part of

41:59

the world. interesting to me. But the

42:01

interesting part is basically how different

42:03

themes kind of inform one another,

42:05

right? Where my

42:08

allocation in China is directly going to be

42:10

impacted by what I'm thinking

42:12

on the election or my

42:14

what I'm thinking on AI, for example,

42:17

you know, there's a there's a lot

42:19

of really contentious stuff going on where

42:22

China, the if

42:25

we're constantly saying, okay, China, you can you

42:27

can buy the generation of chips that's too

42:29

below, you know, whatever Nvidia has out right

42:32

now, you know, they're not just going to

42:34

say, okay, cool, you know, they're not going

42:36

to just relegate themselves to constantly being at

42:38

a disadvantage in this technology. So

42:42

I think when I

42:44

first put this China trade on,

42:46

which the last time I was

42:48

on here, I think I was complaining about how badly it had

42:50

done. But, you know, it recently

42:52

has a for a little

42:55

while, very small amount of time, but for

42:57

a little while, it was outperforming NASDAQ. Since

42:59

I put it on, I think that that

43:01

is back to underperforming the

43:03

NASDAQ slightly, but performing a bound line

43:05

with the S&P. The way

43:08

that I look at it is from the AI

43:11

side, okay, I'm going to focus a little more on the

43:13

necessities of China having to build

43:16

out its its semi domestic semiconductor

43:18

industry, do I

43:20

think that they will necessarily like be successful in

43:23

whether it's, you know, focusing on a six

43:25

or trying to create

43:28

a GPU that, you know, I don't know. But

43:30

what I do know is that they're going to

43:32

try and if you look at like Japanese exports

43:35

of semiconductor capital equipment, semi

43:37

cap to China, that line

43:39

has gone parabolic, right? Because

43:41

China realizes that

43:44

the only way that they're going to have some independence here

43:46

and not be beholden to a US company, which is, you

43:48

know, distasteful for them is

43:50

going to be by doing their own

43:52

build out and trying to basically create

43:54

something that puts them a little ahead.

43:57

You know, do I think that China will ever in

44:00

the next five years will create something that's better

44:02

than what NVIDIA has for the purposes of doing

44:04

AI and machine learning? Probably not, but can

44:07

they create something that is

44:09

better than what's needed to,

44:12

you know, exceed what

44:15

they're allowed to buy from the US? Or can

44:17

they focus more significantly on whatever

44:20

inference or another

44:22

area of AI or custom silicon?

44:24

This is all super interesting and

44:27

it kind of

44:29

informs how all

44:31

these themes, you know, fiscal informs

44:33

playing the election and informs kind of AI

44:37

through like the ChIP-SAC stuff and

44:39

China informs AI, AI informs China.

44:41

And, you know, so I think this

44:44

is like what I personally love about it,

44:46

right? I love like this massive

44:49

puzzle and sometimes

44:51

you get moments where you kind of see where the

44:53

pieces are going to fall or at least you think

44:55

you do. And that's my favorite part of it.

44:58

And we'll talk about your China basket

45:00

in just a sec. But you

45:02

have a basket of companies that benefited from

45:04

the TCGA, the

45:07

Trump's tax cut when he was president, where

45:09

you look at what their tax rates were

45:11

before the TCGA. And then after,

45:14

I'll just take an example. A good example of

45:16

like how hard this stuff is to predict was,

45:19

you know, the performance of like

45:21

clean energy stocks versus energy stocks

45:23

when when Biden got

45:25

elected. Because it's like

45:28

like if it's 2021 and

45:30

you're sitting there and you're looking at

45:33

a Democrat win instead of a Republican and

45:35

you know what the Democrats priorities are and you

45:38

say, all right, well,

45:40

if a Democrat is getting in, you know, that's

45:42

probably not great for energy. And that's probably pretty

45:44

good for clean energy. So

45:46

I'll go along, you know, ICLN and I'll

45:48

go short XLE. You would

45:51

have gotten destroyed. Right. And that's like

45:53

a great example of how the

45:55

economy is always going to be the

45:58

macro economy is the primary driver. And that's

46:00

a lot more important than who wins the election. So

46:04

you have to be really selective about not just

46:06

doing these kind of blanket things and

46:09

blanket assumptions of what the Republicans and

46:11

what the Democrats are like and finding

46:13

areas where there's an outsized impact from

46:16

specific policy rather than just how

46:18

you think things will do because

46:20

X, Y, or Z administration. So

46:22

the other themes are you call

46:25

idiosyncratic Trump beta, which is

46:27

just stocks that are associated

46:29

with Trump. So for yeah, if

46:32

any made ready Mac, rumble conservative

46:34

app, Trump's back, Trump

46:37

crypto tokens or NFTs, although I don't believe

46:39

that those are part of it. That might

46:41

be an interesting thing to talk about because

46:43

it's I don't know if you

46:45

saw recently, you know, Trump came out and was

46:47

like very supportive of crypto, right? And

46:50

so the most interesting part of this that

46:53

really in hindsight is pretty obvious.

46:56

The Biden administration does not

46:58

want crypto to become a

47:00

single issue, right? Where it's like

47:03

that's an issue where people vote along the lines

47:05

of where everybody that owns a significant amount of

47:07

crypto is going to say, well, you know, here's

47:10

how I care about X, Y, or Z

47:12

policy. But I a lot of my networks than crypto,

47:14

so I'm going to vote for Trump. You know, so

47:16

what you saw right after that was the approval of

47:18

the Ethereum ETF by the SEC. And

47:21

I can I can literally imagine the conversation

47:23

that was had between, you know, Biden and

47:25

Gensler where it probably just

47:27

called him and was like, listen, I do not

47:29

want to become the tough on crypto candidate because

47:32

that is not going to fare well. So you

47:34

have to get some something done here. Republicans

47:37

are coming out in support of crypto. Paul

47:39

Ryan, very notable deficit hawk back

47:41

in the day, you know, talk

47:43

about how we're printing so much more. We're borrowing

47:45

so much money. Our grandkids are going to be

47:47

saddled with debt. Now he is a

47:51

stablecoin promoter and talking about how stablecoins are

47:53

necessary because they can help us increase the

47:55

amount of because it's able coins will buy

47:57

U.S. Treasury debt, which is true. But

48:00

it's interesting, he's kind of acknowledging that

48:03

the private sector can print

48:05

money to buy the government

48:08

paper that the government prints when it borrows.

48:10

So a little maybe inconsistency there. And it's

48:12

similar to banks can just buy debt

48:15

and then issue credit against them. What do you think just

48:17

about the pure S&P 500, the

48:19

stock market? Does it do better or worse under

48:22

Trump or Biden and why? I

48:24

don't know. But I have thoughts

48:26

that are kind of contingent

48:28

on specific things happening.

48:30

You know, like I think it's

48:33

going to depend on what Trump can get through

48:35

on the trade policy side. I can tell you

48:37

for sure that, you know, if Trump got everything

48:39

he wanted on the trade policy side, I don't

48:41

think that like the tax cuts and kind of

48:44

supportive policies would be

48:46

enough to offset

48:48

the inflationary impulse. And so

48:51

I think that really that's

48:53

something where you

48:56

we're going to need a little more information leading up

48:58

to the election about exactly where

49:00

their priorities are. And and

49:02

kind of it's going to that probably

49:04

will be a lot more driven,

49:06

not by who wins the presidency, but

49:09

by what the what

49:11

the House and Senate are, because, you

49:14

know, stocks do better historically under

49:16

divided governments. So I think that

49:19

probably the best outcome for stocks

49:21

would be Trump

49:24

presidency and then and then basically

49:26

Democratic control of the Senate. But

49:29

again, I don't necessarily have a

49:31

strong view on that yet. Why

49:34

is that if stocks like large

49:36

fiscal deficits because they're a stimulus

49:38

to the economy, wouldn't the scenario

49:40

where you'd have the largest fiscal

49:42

deficit be Republican in

49:44

the Oval Office and, you

49:46

know, running the Congress? Or Democrats in the Oval

49:48

Office and running Congress? Sure.

49:51

Yeah. And, you know, that gets

49:53

into kind of an interesting question

49:55

of inflation. Is

49:57

inflation bad for stocks or is the central bank

49:59

real? reaction bet for stocks. Is

50:02

it actually because inflation,

50:04

probably inflation itself is

50:06

not the worst thing

50:08

for stocks. You

50:11

look at the 70s and the

50:13

period of time where the Fed

50:15

kind of like burns reaction. There

50:19

were years of pretty good returns in

50:21

there, but it was obviously the Fed's

50:24

reaction to inflation that drove stocks

50:26

down. Yeah. I

50:28

think you're right. There were years of good performance in the 1970s, but

50:30

I think the 1970s was one of the worst

50:36

performances in real terms, one

50:39

of the worst decades for stocks, probably even

50:41

worse than the 2000s, which included the great

50:43

financial crisis. Well, but you

50:45

look at companies margins in the

50:48

beginning, especially during the

50:50

1973, 1974,

50:53

as inflation went

50:55

up, the margins

50:57

on the S&P went up as well.

51:01

I think that you would have

51:03

to basically answer how

51:06

much can Trump believe the Fed, if you get

51:09

this large fiscal, a stimulative,

51:11

pro-cyclical fiscal deficit. Again,

51:14

all of this kind of comes back to

51:18

the actual effect of

51:20

the election on

51:22

the overall economy,

51:25

which I think really is not

51:28

extremely significant. The place where I want

51:30

to be positioned is in the micro

51:32

side. I want to be kind

51:35

of hyper focused on the impacts of

51:37

trade policy and tax policy on the

51:40

collections of single companies. The

51:44

overall economy is going to be much

51:47

more important, and that is where you get

51:49

this disconnect

51:52

of, well, what are stocks going to do? You

51:54

could have said any number of things about

51:56

how stocks are going to do under Biden or Trump

51:59

or So, the thing

52:01

is, predicting the entire stock

52:04

market off of just a single election

52:06

is going to

52:08

lead to some pretty wrongful

52:11

conclusions versus just trying to predict

52:13

the impact of specific

52:17

policies. A good example of this

52:19

is, imagine it's 2021 and Biden has just won over

52:21

Trump. And

52:25

your conclusion as well, the Democrats

52:28

are a lot more accommodative

52:30

of clean energy and

52:33

they're not that friendly to fossil

52:35

fuels. So I'm going to

52:37

go long clean energy stocks like

52:39

ICLN as an example of an

52:41

ETF and then I'm going to go short energy, XLE. You

52:45

would have gotten destroyed, right? The

52:47

returns would, I have

52:50

a chart of it, but it's pretty much down

52:52

only. That's a great example of saying,

52:57

Trump wins and then stocks go up or

52:59

Biden wins and then this happens. The

53:01

broad index is always going to be

53:04

driven more so by what's

53:06

going on overall in the economy and a bunch of

53:08

external factors. And that's why I think that really the

53:10

risk reward here is probably in just playing specific

53:13

kind of policy decisions.

53:15

Sattraini, you have created a basket.

53:18

Interestingly, and I like this, the

53:20

basket that you, so you haven't put on a lot of

53:22

these trades yet, the basket that you've created that is a

53:24

back test. So when you didn't have these positions on, actually

53:27

it tracks, but

53:29

it doesn't track that well. And

53:31

I actually kind of like that because what drives

53:34

me crazy is when people say, I've created this

53:36

index going back to 1980 and it tracks this

53:38

thing perfectly over the past 40 years. So if

53:40

we had a time machine, this would, but

53:43

it's not in sample. You have to

53:45

track it from when you actually have a position

53:47

on because you're going to be back testing until

53:49

you get the back test that you want. And

53:52

it's easy, not easy, but it's, it takes a while,

53:54

but it is doable to just create the perfect back

53:56

test. So I actually like that your back test doesn't

53:58

look that great. that the

54:00

trades that you actually have on such as AI,

54:02

those have tracked AI, physical beneficiaries,

54:05

extremely well. So I do think that

54:07

you, as I called you in our

54:09

first interview, the title, the world's hottest

54:11

stock picker, a very talented stock picker,

54:13

who's trying to pick a theme, as

54:15

you acknowledge, this is an incredibly hard

54:17

task. And so I'm just going to

54:19

ask you somewhat of a challenging question.

54:21

Like you have the the tariff basket.

54:23

So you have let me just find

54:25

this our house, you have, you have,

54:27

for example, long tariff winners, you have

54:29

BJ holes clubs and our house, they

54:32

produce a lot in the US. And then there's your

54:34

short basket for the short tariff

54:36

losers is target and Nike, which produce a

54:38

lot of stuff outside of the US. How

54:41

are you trying like, let's say if 5%

54:43

of the price delta in

54:46

target and our house are

54:49

based on this factor, how do you

54:51

isolate that from the 95% like, let's

54:53

say there's just a huge boom in

54:55

consumer spending, that benefits our house and

54:57

target, or there's a contraction that benefits

55:00

both, how are you trying to isolate

55:02

that factor? It seems very difficult. That's

55:04

where the basket composition comes in. And

55:07

this is an evolving thing, we'll see, it's

55:10

going to be based on, you know, rhetoric in

55:13

the debates. And I

55:16

think that you kind of get a

55:18

little bit less risk of just, well,

55:20

you know, what if this single company

55:22

is able to diversify more away from

55:24

China, or they anticipate

55:27

that the tariff risk will be

55:29

higher? This is where having that

55:32

diversification within kind of a

55:34

theme works out. So for

55:37

example, the tariffs would be long

55:39

DJs, long Ross

55:41

and like the

55:44

long back to the good examples like back

55:46

in Dickinson, right? Like long back in Dickinson,

55:50

and then short CAH. And

55:53

that's because of like specific tariffs

55:55

on syringes, basically.

55:57

And the thing is, like, the

56:00

bipartisan agreement on US protectionism when

56:02

it comes to China, it's pretty

56:05

significant. There's no real

56:07

China doves in the government anymore. The

56:11

way that I view this is,

56:13

I'll look at Biden just

56:15

passed some tariffs on China

56:18

that were a lot more specific than what

56:21

Trump's would be, but if

56:23

I can create a basket of tariffs

56:26

with longs like Beck and Dickinson and

56:28

then short like Cardinal Health and have

56:30

this kind of exposure where it will

56:33

likely get impacted by tariffs

56:35

that are already existing under Biden

56:37

that will only get expanded under Trump. That

56:40

gives me another kind of margin of safety.

56:44

As far as, I'm not going to

56:46

outright just short Nike because of tariffs,

56:48

but in a

56:50

collection with Target, Restoration

56:53

Hardware, Florine DeCors, it creates

56:55

an overall smooth out where you're

57:02

more accurately playing just tariffs

57:04

rather than specific company decisions.

57:07

Got it. Then how

57:09

do you think the

57:11

tariffs on Chinese vehicles

57:13

as well as European vehicles is going

57:16

to impact American car

57:18

manufacturers as well as let's say Tesla?

57:21

That's a pretty interesting one. It

57:23

seems like Europe originally

57:25

was a little more accepting of

57:27

the Chinese EVs and they might

57:30

be pivoting a little bit away from that. Have

57:34

you seen one of those electric vehicles? The

57:38

Chinese one? Yeah. I

57:40

have not. I know they're very cheap though. They're

57:43

super cheap and even better than the ones

57:45

we have here. It's

57:48

a shame really that if

57:51

you look at what happened with Japanese imports,

57:54

it elevated where

57:57

you can't have one country that's constantly just sending

58:00

these cheaper vehicles. Eventually, the

58:02

domestics have to keep

58:05

up. But Tesla has always been pretty

58:07

good at playing, whether it's

58:09

government subsidies or government policies. And

58:13

that's something where I don't think that we're ever

58:15

going to... The US is never going to let

58:17

in these Chinese EVs because

58:19

it would completely destroy the US electric

58:22

vehicle market. Just because there's

58:24

no way to compete. They're so

58:26

far ahead on production. And so

58:28

I think that the Chinese kind

58:30

of electric vehicle market, it's probably going to stay

58:33

in China. Sattrin, there

58:35

are some people who would consider

58:37

you to have a, oh,

58:40

this guy's focused on AI pie in the

58:42

sky thing. So I want to ask you

58:44

about what do you think about the Tesla...

58:46

Tesla's as an AI company and

58:48

Tesla as a beneficiary of

58:51

robotaxes, as well as the

58:53

most recent ARK open letter

58:55

where the current stock

58:57

price of Tesla is, let me get it, $178.

59:01

So ARK has an expected value in Tesla in 29

59:03

of $2600 per share. Their

59:08

bull case is 3,100 and their bear

59:10

case is $2,000. Again,

59:12

I repeat, the stock price now is $178. So their bear case is a

59:14

10X. What

59:17

do you think about Tesla? I haven't really seen Tesla in any

59:19

of your baskets. This is great,

59:21

Jack. I like how you had

59:23

me on when everyone hated AI and I

59:26

got to be bullish on AI. And now

59:28

I get to alienate all the real Super

59:30

Bulls who are also bullish Tesla because Tesla

59:32

is a car company. I

59:37

think that the robotaxi thing...

59:39

I said back and even before I was on the

59:41

podcast the first time, I said, you want to be

59:43

long AI and short Tesla. And so

59:46

it was good because I got to really piss

59:49

off the bears by being bullish

59:51

and then really alienate

59:54

everyone that was also very bullish because they

59:56

were likely to be bullish Tesla as well.

1:00:00

So, Tesla is, how

1:00:03

can you be involved in a company with

1:00:05

such like crazy governance? The

1:00:08

idea that I think Elon

1:00:10

Musk saying basically, well, I'll take my

1:00:12

AI efforts elsewhere. Can

1:00:15

you be an AI company if you're like constantly under

1:00:17

the threat that if you don't pay this guy enough

1:00:19

that, although I will

1:00:21

say that with the

1:00:23

shareholder meeting, I'm pretty sure they'll vote

1:00:25

his way. And I think that's probably a

1:00:27

good- They did, they already did. Oh,

1:00:30

okay. Yeah, there you go. I think that that's

1:00:32

actually, I mean, that's reasonable. You

1:00:35

made a deal. I'm

1:00:38

a business person, right? If you

1:00:40

say to your CEO, listen,

1:00:42

here's these crazy ridiculous targets that everyone on

1:00:45

the street thinks are just insane, but if

1:00:47

you hit them, we're going to pay you

1:00:49

an absolutely wild amount of money. If you

1:00:51

agree to that, then you have already agreed to

1:00:53

that. Do I think that being

1:00:56

able to keep Elon Musk

1:00:58

incentivized is going to really

1:01:01

elevate Tesla to the level of AI competitor?

1:01:04

No, I don't think so. Why not? I

1:01:07

just think that the, I mean,

1:01:09

the hyperscaler environment right now, and

1:01:11

also just the environment of AI

1:01:14

and companies like

1:01:16

that, just to

1:01:18

go off on a little bit of a tangent, I think that when

1:01:21

you look at the landscape right now

1:01:24

and all of these gains that have kind

1:01:26

of accrued to the hyperscalers and this fierce

1:01:28

competition in AI. A hyperscaler

1:01:30

just for the audience, how do you

1:01:33

define it? It's like a huge cloud

1:01:35

company like Amazon or Microsoft. Apple, Microsoft,

1:01:37

Google, yeah, Amazon. Yeah.

1:01:41

I know this is a little bit of a tangent, but something

1:01:44

that I wrote recently at the end

1:01:46

of May was that what

1:01:49

we're going to see now is that

1:01:51

competition has to heat up because the

1:01:54

way that it looks like AI is progressing

1:01:56

is essentially going from

1:01:59

here's chat. GPT. And you know, you

1:02:01

can use that as a productivity tool. And here's

1:02:03

like all these various startups and everything with their

1:02:05

AI tools. But the real value of AI is

1:02:07

that, you know, the

1:02:10

hottest new coding language is English, right?

1:02:12

And being able to interact regularly

1:02:14

and effectively with words

1:02:17

with, you know, like

1:02:19

an AI assistant, and have an AI

1:02:21

assistant that is using, whether

1:02:24

it's agents, or these kind of, quote

1:02:26

unquote, specialist models that are really

1:02:29

aggressively focused on their

1:02:32

smaller models, but they're really aggressively focused on specific

1:02:34

tasks, and have being

1:02:36

able to walk down the street and not and

1:02:39

work and not be on your phone and looking

1:02:41

down and just have, you know, an earphone and

1:02:43

say, Hey, you know, what was that meeting that

1:02:45

we had last week about so and so? And

1:02:47

then it tells you and you say, Okay, you

1:02:49

know, let's email that guy, let's find a place

1:02:51

in my schedule where we can do that. And

1:02:54

then as you interact with this,

1:02:56

it's going to increasingly kind of get

1:02:58

to know you better and get to

1:03:00

anticipate your needs and do inference and

1:03:03

be able to kind of infer things. And if

1:03:05

that's where we're going from here, the the

1:03:09

biggest component of who wins in in

1:03:11

these companies for for kind of consumer

1:03:13

and even be to be

1:03:15

to be focused AI, it has

1:03:17

a lot to do with trust. And when

1:03:20

you look at a company like say,

1:03:22

you know, Google or Apple, right, and

1:03:25

Google has the stock price has

1:03:27

pretty much shrugged off these kind of faux

1:03:29

pas that it's had with AI. I don't

1:03:31

know if you have been following this half

1:03:33

on June 2, I

1:03:36

published and that we were going to sell Google

1:03:38

in our AI basket and go long Apple. And,

1:03:41

you know, that worked that worked out pretty

1:03:43

well with the WVTC. And I think that

1:03:45

that what we're going to see from here

1:03:47

is basically, there's going to

1:03:49

be more dispersion in the Mag 7 because of this.

1:03:52

And, and, you know, so if you ask me, like,

1:03:54

well, you know, how is Tesla not

1:03:56

an AI company? Well, you know, I don't

1:03:58

really think that Tesla has a has a a real

1:04:01

kind of expansive trust that it can

1:04:03

tap into. It obviously has

1:04:05

a very passionate user base. Yeah, that's awesome.

1:04:08

But it's kind of, have

1:04:11

you ever seen the movie Gattaca? Yes. So

1:04:16

the way that I think that this goes forward is basically

1:04:18

if you are not

1:04:20

using AI and

1:04:23

you're not kind of, listen, not like

1:04:25

right now, although maybe in some areas

1:04:28

right now, but as AI

1:04:30

progresses to kind of this

1:04:32

AI assistant agent, whatever

1:04:34

you want to call it, there's going

1:04:36

to be a point in time where if

1:04:38

you're not using it, your productivity will be

1:04:41

so much worse than the next guys. And

1:04:43

that's why I think that these products are

1:04:45

going to be unbelievably sticky, right? Because of

1:04:47

that kind of inference that it's doing where

1:04:50

Apple has the potential here to,

1:04:52

do you know what single sign on is? No.

1:04:54

SSO. So you know like when you go to

1:04:57

a new website and it's got something at the

1:04:59

bottom that's like sign in with Google, sign in

1:05:01

with Apple. Yeah. Yeah. So if

1:05:04

you could imagine like an SSO, but like

1:05:06

for your entire life where

1:05:09

you have this one ecosystem and that's

1:05:12

where it kind of knows everything about you

1:05:14

and it can anticipate your needs.

1:05:16

And so whoever kind of wins this, not

1:05:22

that it's a zero sum game, but

1:05:24

there is a certain aspect of that

1:05:26

where whoever makes it so

1:05:29

that you have an AI

1:05:31

tool that is going to be better at knowing

1:05:33

you than any other solution. Even like

1:05:35

I said on the business side where

1:05:38

it's something like Salesforce. Salesforce was

1:05:40

positioned to be like a pretty

1:05:42

significant AI beneficiary, but it's also

1:05:46

really exposed to this

1:05:49

kind of AI and for instance just anyone

1:05:51

being able to have your own

1:05:53

CRM. And have

1:05:55

AI do that instead of having a

1:05:57

specific solution for it. And so. So

1:06:00

these ecosystems will be created. And I don't think that Tesla

1:06:02

is going to be the one to do that. And

1:06:05

CRM, customer relationship management also happens

1:06:07

to be the ticker of Salesforce,

1:06:09

which has historically been an excellently

1:06:11

performing stock. There

1:06:13

has been a bear market in software that

1:06:15

I think has taken a lot of people

1:06:17

by surprise. And is it because the reason

1:06:20

you just mentioned that, why do I

1:06:22

have to pay some company, $100, $1,000

1:06:25

a month to do this advanced processing

1:06:27

when I can just download a chatbot for free?

1:06:29

Or very free and... I

1:06:32

don't know. And the

1:06:34

software bear... You can... Is it

1:06:36

AI cannibalization? Is it that there are

1:06:38

so many variables here about why software

1:06:41

has underperformed? It's very

1:06:43

difficult to nail down a single

1:06:45

reason why. But I have

1:06:47

also kind of played my hand

1:06:50

with software and it

1:06:52

was very underrepresented in my AI

1:06:54

allocation. And the reason why was

1:06:56

because back in

1:06:58

June, I basically said that it would be too

1:07:00

difficult to anticipate who's going to win and who's

1:07:02

going to lose. Because it's not so much about

1:07:05

having upside because of AI. It's about not having

1:07:07

downside. And when

1:07:09

you looked at all

1:07:11

of these gains accruing to the hyperscalers and you had

1:07:14

all these software companies,

1:07:17

if you were going to say, oh, who's most likely

1:07:19

to be harmed by AI? Well, in the

1:07:21

long term, a lot of software companies. And I did

1:07:23

include a couple of them. But as I've narrowed it

1:07:26

down, I've focused a little bit away from that. I

1:07:28

think that you have the luxury right now basically waiting

1:07:30

to see which of

1:07:32

these software companies can kind of

1:07:34

integrate best. I think that CloudFlare

1:07:36

is kind of interesting just

1:07:39

from the CDN cybersecurity angle. And

1:07:42

I'm not someone that's going to be

1:07:44

buying 52-week lows. I think that the

1:07:46

amount of money that's been bought buying

1:07:48

off the 52-week low list is really

1:07:51

just astronomical. I

1:07:53

would much rather live by that, that

1:07:55

Jesse Livermore quote, the most expensive move,

1:07:58

the most expensive 8th of any move. are the

1:08:00

first and the last. And so I'll wait to

1:08:02

see who kind of makes it out. Because

1:08:07

right now in software, there's

1:08:11

going to be some amazing, amazing buys.

1:08:13

But there's also maybe 30% of software

1:08:15

companies out there right

1:08:17

now, just donuts. And I don't

1:08:19

really necessarily feel like rolling the dice on that yet.

1:08:22

And would you consider shorting any of these

1:08:24

companies you consider donuts? What are some SaaS,

1:08:26

software as a service, SaaS donuts, that you

1:08:28

have your eye on? I did that

1:08:30

already. I had like an

1:08:33

AI winners and losers

1:08:35

basket. And when we

1:08:38

got down to October

1:08:40

of last year, that was when I really

1:08:42

basically said to everyone, if you miss the

1:08:45

first wave, this is the time where you

1:08:47

want to. That was when I

1:08:49

said on Twitter, we're never going to see below 400

1:08:51

on Nvidia again. Although I guess I was

1:08:53

wrong because of this thought. But

1:08:57

that's around when I kind of covered the

1:09:00

shorts in software. But some of

1:09:02

the companies, like Expensify was

1:09:05

one of them. And I don't

1:09:07

really want to talk about that because it's

1:09:09

like right now, it's just so. You're not

1:09:11

short those things now. Yeah. Got

1:09:14

it. You said something about the Magnificent Seven.

1:09:16

So you think a lot of that will

1:09:18

be in the Magnificent Seven. So the hyperscalers,

1:09:20

which is Amazon, Google, and Microsoft, and

1:09:23

Meta. And then there's Nvidia, which is selling

1:09:25

them hyperscalers, all of

1:09:27

their stuff, which people have identified as a

1:09:29

weakness, which is all of the Nvidia's increase

1:09:31

in revenue is coming from these companies, even

1:09:34

though these companies have tons of money. Yep. That's

1:09:36

a serious risk. And something

1:09:38

that if you're playing this now

1:09:40

where the

1:09:42

margin of safety has gotten extraordinarily

1:09:45

diminished, you really have to

1:09:47

pay attention to the efforts

1:09:49

on the part of the hyperscalers

1:09:51

to develop custom silicon and basically

1:09:54

make their own chips that can

1:09:56

do this, the application-specific integrated circuits,

1:09:58

ASICS. to

1:10:00

also look at the efforts

1:10:02

in China, whether or not that's

1:10:04

a possibility or a probability. I

1:10:07

don't know, but it's something that you absolutely

1:10:09

have to track because what happens if Nvidia's

1:10:11

kind of dominance in this is disrupted by

1:10:13

that? I don't think that that's something

1:10:16

that's gonna happen in the next 12 months,

1:10:18

but it's something that there's only so long

1:10:20

that you can kind of maintain these extraordinary

1:10:23

margins, selling the

1:10:26

best in the industry. You

1:10:28

look at, I don't know, we can throw it back

1:10:30

to history and look at Portugal

1:10:33

when they invented the archibas.

1:10:36

That lasted maybe 15 years of 90%

1:10:38

margins from being able to

1:10:41

make the only musket. But eventually

1:10:43

you get disrupted. And

1:10:45

that's not a if question,

1:10:48

it's a one question. And the

1:10:51

when is basically gonna be

1:10:53

what is the extent to this? And

1:10:55

obviously Nvidia is winning right now. And

1:10:59

it's just something definitely keep an eye on

1:11:01

where you're kind of, well, how

1:11:04

is this gonna change? And you

1:11:06

talked about those companies, the hyperscalers efforts

1:11:08

to create their own chips in our

1:11:10

conversation in late January or early February

1:11:12

of this year. What

1:11:15

progress has been made on

1:11:17

that and what companies are helping Meta

1:11:20

and the hyperscalers do that? If this

1:11:23

happens, right tomorrow, Meta is,

1:11:27

oh, we created our own custom silicon and we don't

1:11:29

need to buy Nvidia chips anymore. I

1:11:32

think you would wanna look, again,

1:11:34

TSMC, right? Just because TSMC, probably,

1:11:37

Marvell, Micron, you're

1:11:43

still gonna need some of the

1:11:45

same components to create these chips. So

1:11:48

that would probably, it would shift some of the

1:11:51

balance of power to the suppliers

1:11:53

for ASICs rather than the supplier of

1:11:55

the finished product chip. And

1:11:57

in our second conversation earlier this year, or

1:12:00

you said that just because AI is

1:12:02

a thing doesn't mean that semiconductors aren't

1:12:04

cyclical anymore. Semiconductors are a cyclical industry.

1:12:07

There's being a boom, there being a

1:12:09

bust. Where are we in these?

1:12:11

Do you still stand by that statement? And where

1:12:13

are we in the semiconductor cycle now, you know,

1:12:15

five, six months later? No, yeah, absolutely.

1:12:17

I stand by that statement. And I think that that's

1:12:20

the you want to talk about like ringing

1:12:22

a bell at the top. When everyone is

1:12:24

starting to say, well, semiconductors aren't cyclical anymore.

1:12:27

That's when you want to start getting worried. It's

1:12:30

still going to be an economic activity thing. And

1:12:33

all this capex spending, it's

1:12:35

still related to you have

1:12:37

to have a certain kind of optimism

1:12:39

about the overall environment to continue kind

1:12:42

of spending this aggressively

1:12:44

on new technology. What

1:12:46

do you think the odds are that this is

1:12:48

the top in the AI type

1:12:51

stocks? Well, I think that that's a

1:12:53

pretty interesting that's a dynamic that's played

1:12:55

out. I don't know how many times

1:12:57

where, you know, the the buying pressure

1:13:00

on Nvidia is basically driven by people

1:13:02

who sold it, you know, 30 percent

1:13:05

ago, who said

1:13:07

it was too expensive then and then are

1:13:10

drastically underperforming because they're underweight Nvidia relative to

1:13:13

the index and then are forced to kind

1:13:15

of buy it back. And then

1:13:17

the people that are selling it to them are

1:13:19

people who think it's too expensive now. And it's

1:13:21

kind of this circle where, you know, the

1:13:25

kind of price action that we've seen on Nvidia for the

1:13:27

past, you know, couple of months is something

1:13:29

that it's not sustainable forever,

1:13:31

obviously. And but it all

1:13:34

it comes back again to the other side of

1:13:36

that Jesse Livermore quote, where the most expensive Ace

1:13:38

of any move are the first

1:13:41

and the last. As long as Nvidia

1:13:43

is going up every day, I'm not going to sell it. I

1:13:46

will, you know, and there are like periods

1:13:49

of consolidation stuff. And this is this kind

1:13:51

of when you get into

1:13:54

like the mania aspect where it starts to get

1:13:56

disconnected from fundamentals, that's when it just starts to

1:13:58

become a thematic kind of trend. following thing

1:14:01

where this

1:14:03

is clearly in a very strong

1:14:05

trend and I'm not going to try to fight that.

1:14:08

But I do think that there

1:14:11

are going to be, especially from prices

1:14:13

right now, there will be better opportunities

1:14:15

for outsized returns just like we said

1:14:17

on Micron back when it was trading

1:14:23

at $70. And

1:14:25

now it's at $140. And that's a double.

1:14:29

Obviously, you probably could

1:14:31

have made the same return buying Nvidia, but

1:14:33

what was your risk? Not

1:14:35

talking about ex-post, talking about ex-ante, what

1:14:37

was your risk in buying Micron when

1:14:39

it's kind of like coming out of

1:14:41

this cyclical trough versus buying Nvidia when

1:14:44

it's gone up every single day? I

1:14:46

think that I'm not buying more Nvidia right

1:14:48

now. Can I ask you, what's

1:14:50

your cost basis on Nvidia? You can either give

1:14:53

me post split or pre-split. I mean pre-split, it's

1:14:55

around $250 to $300. This

1:15:01

flow is 10 to 1, right? Or 9 to 1? Yeah,

1:15:03

10 to 1. 10 to 1. Okay.

1:15:06

So now it's a hot 13. So your cost basis, what you

1:15:08

got in it was? Call it $300. Yeah. So

1:15:12

that's $30. And now it's at $130.

1:15:14

So it's up quite a lot. And this is a thing

1:15:16

that in finance, people

1:15:18

who have a lower cost basis have different

1:15:20

behavioral incentives than someone who buys

1:15:22

at the highs. And likewise, if you buy a

1:15:24

SPAC for $10 and it goes down to a

1:15:26

dollar, you say, oh my God, there's no way

1:15:28

anyone would ever sell it at a dollar. But

1:15:31

if you're part of the company and you got

1:15:33

shares at $0.05 or a founder share, you would

1:15:35

sell. So that cost basis

1:15:38

thing is important. What would you

1:15:40

have to see for you to become bearish on Nvidia?

1:15:44

Because going back to our first conversation, James,

1:15:46

you said that this is

1:15:48

a bubble. And to quote the George, you're a big

1:15:50

fan of George Soros, when you see

1:15:52

a bubble, the first thing you do is buy. You saw this

1:15:54

as a very. No, no, no, no. When

1:15:56

you see a bubble forming. Yes.

1:15:59

Right. That's correct. That's the

1:16:01

bid. That word does a lot of

1:16:03

heavy lifting there because I think if

1:16:05

you said it that way, when you

1:16:08

see a bubble buy to anyone who

1:16:10

bought the top in 2000, they're

1:16:12

going to very significantly disagree with you.

1:16:14

When you see a bubble pouring, yes.

1:16:17

Yeah. When you're underwater

1:16:19

on Microsoft for 13 years, and then

1:16:21

there's, like you said, the behavioral incentives

1:16:23

and how likely are you to hold

1:16:25

Microsoft underwater for 13 years and just

1:16:28

dollar cost average into it, you

1:16:31

need to be really talented and have a lot of

1:16:33

foresight if you're going to do something like that. What

1:16:37

would cause me to be bearish on NVIDIA? In

1:16:40

terms of the overall market

1:16:43

environment, if we went into a recession,

1:16:45

yeah, I would be very

1:16:47

bearish on NVIDIA. In

1:16:50

terms of intrathomatic and developments

1:16:52

in AI, it comes back

1:16:54

to basically disruption. NVIDIA

1:16:58

has been the disruptor, and

1:17:00

it's a very common theme where

1:17:02

you go from being the disruptor to the disrupted.

1:17:04

If there's signs that there are

1:17:07

viable products that can unseat NVIDIA

1:17:09

in this pole position, that's

1:17:12

when you start saying, well, on

1:17:14

one hand, do I want to continue owning NVIDIA? But on

1:17:16

the other hand, what is the risk to the overall AI

1:17:19

theme here? Listen,

1:17:22

I'm bullish on AI, period,

1:17:24

right? But

1:17:26

I'm still in the market. If I think that

1:17:29

the stocks that are going to benefit from AI, that in a

1:17:31

year I'll be able to buy them 50%

1:17:33

cheaper, I have no qualms about selling

1:17:35

and getting back in. That's

1:17:38

something that I think probably

1:17:40

will happen. Over

1:17:42

the next 10 years, it's a market, right? But

1:17:48

I think it's going to be a very positive AI's

1:17:50

progression, which will be a technology that

1:17:52

we all use in our day to day

1:17:54

life. Right now, you can make

1:17:56

the decision whether or not you're going to use chat GPT, or

1:17:59

maybe where it doesn't really make sense for

1:18:01

you to, over

1:18:03

the next 10 years, every

1:18:05

single person in the world will be using AI.

1:18:08

But does that mean that the

1:18:11

stocks are only gonna go up from here? No, of course

1:18:13

not. It's a financial market. We

1:18:15

will get way too optimistic and then

1:18:17

we'll get way too pessimistic. And so

1:18:20

I'm not gonna sell because

1:18:23

we're too optimistic. I'm gonna sell because

1:18:25

we're too optimistic and threats start materializing

1:18:27

to that optimism. And those

1:18:29

threats would be a recession. So you divide the

1:18:31

threats into macro and non-macro. A recession as

1:18:34

a threat, spending goes down,

1:18:36

as well as the non-macro threat

1:18:38

to Nvidia is that all

1:18:40

the hyperscalers create their own chips. Yep,

1:18:43

essentially. And also, or

1:18:45

maybe China creates its own chip and

1:18:47

then China has a decision to make. Do we keep this for ourselves

1:18:49

or do we sell

1:18:51

it on the global market? And that's

1:18:54

something, although that is

1:18:57

a whole different conversation about the incentives there and

1:18:59

whether or not they would make that decision that

1:19:01

would take an entire other podcast to discuss. Now

1:19:04

let's go over to China. You mentioned

1:19:06

you added some companies to your China

1:19:08

basket and specifically on the

1:19:12

sub-localization of artificial intelligence. Basically

1:19:14

China can't buy AI chips.

1:19:16

So you added a

1:19:18

bunch of companies that are there. So

1:19:20

Sotri, in your changes to your Chinese

1:19:22

equity basket, I see five companies that

1:19:25

I, someone who is familiar with

1:19:27

a few names in the Chinese stock market

1:19:29

have never heard of. Nauru, Pyotech,

1:19:31

Hygon, Cambrakhan and Huahong.

1:19:35

Are these companies all in the AI

1:19:37

semiconductor world and why have you added

1:19:39

them to your basket? So we just

1:19:42

spoke about basically hyperscalers

1:19:45

and Nvidia having them by the balls. And

1:19:47

that's not a great position to be

1:19:50

in, obviously. You're

1:19:52

very incentivized to get out from under

1:19:54

that, especially when

1:19:56

you're running a business or you're, you

1:20:00

don't want to just be at the mercy of a

1:20:02

single company and have that be your input cost. So

1:20:06

consider that but on a nation

1:20:09

state version, right?

1:20:11

If you're China, you

1:20:13

know, I mean, you can

1:20:16

buy Nvidia chips, but you

1:20:19

can buy the most advanced Nvidia chips, but not on the

1:20:21

scale that you need to not with not with all the

1:20:23

sanctions, you know, sanctions aren't 100% effective for anything.

1:20:26

But sanctions on a product

1:20:28

that has really, really significant demand in,

1:20:30

you know, the domestic

1:20:32

market here, you're not

1:20:34

going to be able to compete really,

1:20:37

although I

1:20:39

will say that, you know, they have been

1:20:41

relatively effective in kind of finding ways around

1:20:43

it. And as demand kind

1:20:45

of softens for

1:20:47

these chips, and on

1:20:50

the less advanced side, we'll

1:20:52

see probably more of them finding their way to

1:20:54

China. But still, this

1:20:57

can be perceived for China as a

1:20:59

national security issue, right? Like a, it's

1:21:01

basically a state of emergency where if

1:21:04

Nvidia chips are going to result in

1:21:06

the US developing military technology that can

1:21:08

outplay them in a physical battlefield or

1:21:11

a virtual battlefield, and they have no

1:21:13

way of catching up and are constantly

1:21:15

at the mercy of, you know, buying

1:21:17

the second best chip. The

1:21:20

answer is pretty straightforward, right? You have to create

1:21:22

your own. And I mean,

1:21:25

that's, but that's a pretty tall order, you know, these

1:21:27

chips are the result of

1:21:29

decades of R&D with a unifying

1:21:31

goal supported by deeply ingrained relationships.

1:21:34

And, you know,

1:21:36

so basically for the for, it's

1:21:39

kind of a unique situation, because China

1:21:41

can't just copy, right, they

1:21:44

have to create their own, they have to,

1:21:48

they probably have the exact specs and, and,

1:21:50

you know, corporate espionage is good enough

1:21:52

where they probably know exactly what it is,

1:21:54

but you need the actual infrastructure

1:21:57

to be able to create these things. And

1:22:00

And so I see an environment

1:22:02

where China is going to have to

1:22:04

continue spending aggressively on WFE,

1:22:06

wait for fab equipment, and redouble

1:22:08

their efforts to create Chinese AI

1:22:11

without this Nvidia handicap. And this

1:22:13

goes beyond just having AI. It's

1:22:15

also AI's development right now

1:22:17

is very focused on the English language.

1:22:21

And that's going to

1:22:23

continue to be a thing until someone

1:22:26

is properly incentivized to do something else. You

1:22:29

know, the best way to kind of describe

1:22:32

ASICs is, have you

1:22:34

ever seen like a crypto miner,

1:22:36

like a Bitcoin miner? So

1:22:38

it used to be, you know, in the very

1:22:41

early days, you would be able to

1:22:43

mine Bitcoin on like your Celeron CPU,

1:22:45

right, and like your Dell XPS computer.

1:22:48

And then it evolved to basically you would buy

1:22:50

a milk crate and you would stick a motherboard

1:22:52

in it and some PCIe risers and put like

1:22:54

a terrible CPU in it. And

1:22:57

then you would just buy six

1:22:59

Nvidia consumer GPUs and stick them

1:23:01

on the milk crate and that would be your

1:23:03

crypto miner. And then some

1:23:06

Chinese companies basically figured out that they

1:23:09

could do this better. And you had

1:23:11

companies like Bitmain and Canaan and they

1:23:13

created application specific integrated circuits. And the

1:23:15

application it was specific for was mining

1:23:17

crypto. And, you know, now

1:23:21

that's still a function of floating point operations and

1:23:23

not not to say that, oh, if you can

1:23:25

create a Bitcoin miner, you can create, you know,

1:23:27

an AI ship that's going to be better than

1:23:29

what you can buy, what you're allowed to buy

1:23:31

from Nvidia. But China

1:23:33

is going to have to make some very

1:23:36

strategic investments and, you know, they can't just

1:23:38

spin up a few new fabs and churn

1:23:40

out a bunch of knockoff hopper chips. They

1:23:42

need to create an entire stack. And

1:23:46

that's an uphill battle, but it also

1:23:48

involves a lot of domestic spending. So

1:23:50

when you look at the

1:23:53

Chinese economy and the likelihood of like Chinese

1:23:56

fiscal stimulus, I think

1:23:58

that you. What

1:24:01

I've been doing with this kind of approach

1:24:03

to China that has played out pretty well

1:24:05

is what

1:24:08

I've called the Chinese equity barbell,

1:24:10

basically, where on the left

1:24:12

side of the barbell you have supply

1:24:15

side stimulus beneficiaries, and then on

1:24:17

the right side you have demand side stimulus

1:24:19

beneficiaries. And then for a while, what I

1:24:21

said the bar of the barbell was, we

1:24:23

were short three or four

1:24:26

times as much notional Chinese Yuan as

1:24:28

we were long Chinese stocks. And

1:24:32

so I covered that specific

1:24:35

aspect, but I'm still approaching it from

1:24:37

there is one of two ways that China

1:24:39

is going to do stimulus more aggressively. And right

1:24:41

now it's looking like consumer demand

1:24:43

focused stimulus would be the most

1:24:46

advisable course of action, but they're still going

1:24:48

to do a certain level of manufacturing stimulus.

1:24:50

And if you're China and you're looking at where

1:24:53

I'm going to stimulate my manufacturing

1:24:55

economy, am I going to stimulate the

1:24:57

part of it that's primarily focused on

1:24:59

exports and primarily very exposed

1:25:02

to maybe Trump gets

1:25:04

in and now everything that those manufacturers

1:25:06

are making has a 60% tariff on it when it's

1:25:08

sold to America? No, you're going to focus

1:25:10

on places that allow you to bolster

1:25:13

your own domestic economy. And I

1:25:15

think that the biggest area of this

1:25:17

would be in AI chips

1:25:19

and kind of fostering

1:25:21

innovation and keeping chips

1:25:25

in house and ensuring that you have your own AI ecosystem.

1:25:27

And they're already doing this. If

1:25:30

you look at a chart

1:25:32

of Japanese semi

1:25:34

cap exports to China,

1:25:37

it is a parabola basically.

1:25:40

And so what I've

1:25:42

done is essentially for

1:25:45

a while, these names have been obviously

1:25:47

getting slaughtered, right? And there's always an

1:25:49

overhang and kind of a risk of-

1:25:52

The five companies I listed you said. Yeah.

1:25:55

There's always a risk of basically

1:25:58

ending up on the entity list. like

1:26:00

SMIC, which is China's

1:26:02

TSMC. But I

1:26:05

think that for now, it's

1:26:07

an asymmetric bet where you

1:26:10

just look at the entire ecosystem of

1:26:12

what the ecosystem looks like in the

1:26:14

West, where ASML, LAM, applied

1:26:17

materials, and then CLAC,

1:26:19

and then you look at how

1:26:21

is China going to bring this

1:26:24

in country and what companies benefit

1:26:26

from that. NARA does

1:26:28

a deposition and etching. PO-TEC

1:26:32

does a chemical vapor

1:26:34

deposition, and HIGON makes CPUs, and

1:26:36

KAMBRKON is basically could

1:26:42

be, quote unquote, Chinese Nvidia. If then

1:26:46

you can't buy Huawei, but that's one of them.

1:26:49

Huawei is a semiconductor boundary.

1:26:51

So I think that right

1:26:53

now, that's primarily where I'm focused

1:26:56

on the supply

1:26:58

side of beneficiaries of whatever China decides

1:27:00

to do on fiscal. I

1:27:02

could. This could be off base. But did you get

1:27:04

a chance to look through or do you have views

1:27:06

of the short report by

1:27:09

Scorpion Capital on laser tech, which

1:27:11

that person alleges is a

1:27:13

colossal fraud and ticking time

1:27:15

bomb? And I believe it's a

1:27:18

competitor to ASML and its

1:27:20

customers are TSMC, Intel, Samsung. That's

1:27:22

the Japanese one, right? Yes. Yeah.

1:27:25

I mean, I applaud them for that's

1:27:28

a colossal undertaking to do a short report

1:27:30

on a Japanese company. And yeah,

1:27:33

and the way that the governance works

1:27:35

over there. And I mean, that is

1:27:38

a labyrinth. I haven't read it yet.

1:27:40

I didn't own the company. Luckily,

1:27:43

I think the most impressive short report I've

1:27:45

seen ever was

1:27:47

the Hindenburg or Donnie one. But

1:27:49

I wrote something about, you

1:27:52

know, kind of like financial history and talking

1:27:54

about the Dutch

1:27:56

East India Company and a really.

1:27:59

great story about how the Cape of

1:28:02

Good Hope was discovered because a guy

1:28:04

had basically shorted the Dutch East India

1:28:06

Company, one of the original people in

1:28:09

there. It's too long

1:28:11

to get into now, but the thing is, when you

1:28:13

are- Is that Isaac Demare? Yes. Yeah,

1:28:15

it was. Yeah, so you know. I

1:28:19

thought that that wasn't going to work because it's

1:28:21

kind of like a national champion type thing where

1:28:25

people will- just like how Holland

1:28:27

basically said that the

1:28:29

stock price of the VOC is direct

1:28:32

to how we are perceived on the national

1:28:34

stage. But when it comes to some semi-cap

1:28:37

company in Japan, I think your

1:28:39

only risk there is understanding it, but clearly they

1:28:41

did the work to understand it. So I don't

1:28:43

really know. I haven't read that specific report, but

1:28:46

if it is a fraud, then yeah, best of luck to them. Got

1:28:50

it. So I want to close by

1:28:52

asking, so yeah, on your first appearance

1:28:54

almost exactly a year ago, in our

1:28:57

YouTube comments from my subscribers, Blocker,

1:28:59

you get some negative comments. This

1:29:01

is before your portfolio went up

1:29:03

100%, but he said, just another

1:29:05

pumper talking his book looking for

1:29:07

X liquidity. Him

1:29:09

saying that long money is underweight technology. I

1:29:11

stopped listening right there. Literally everyone

1:29:14

is max long tech. If you're long the S&P

1:29:16

500, your ball is deep in tech because the

1:29:18

index is cap weighted. I mean,

1:29:20

that latter party is true, but that doesn't mean you

1:29:22

can't get more cap with tech, which is what happened.

1:29:25

Jack, you have the best macro channel on YouTube. Thank

1:29:27

you. However, you know that it's just

1:29:29

as much substance or anything to teach your audience. On

1:29:32

Twitter, I asked people for questions

1:29:34

and someone said, how does it feel to

1:29:36

live rent free in the heads of so

1:29:38

many perma bears? What do you think?

1:29:41

I mean, as long as I don't have to declare

1:29:43

it on my taxes. No. Honestly,

1:29:45

I think if you

1:29:49

have a theory and

1:29:51

you get significant pushback from it, that's kind

1:29:53

of what if you can find a place

1:29:56

where you can get quality pushback, right? Where

1:29:58

people are not just. attacking

1:30:02

the general idea of being long

1:30:04

technology. But if anything, that was

1:30:07

part of my thesis, which is just the

1:30:10

overarching skepticism of tech is

1:30:12

going to make it so that this is

1:30:14

gonna be more impactful on a market level

1:30:17

because of the idea that

1:30:19

everyone is going to have to rush

1:30:21

to buy these things because they are

1:30:23

so jaded by what just

1:30:25

blew up in their face during 2022. So

1:30:29

that is one thing, but the

1:30:31

idea that if someone comes

1:30:33

up to you and really disagrees

1:30:35

with you, but disagrees with specific points of

1:30:38

your argument, I would say to,

1:30:40

if anyone is listening right now

1:30:43

and has a really specific

1:30:46

disagreement with something that I've said,

1:30:48

please, that's how you get better.

1:30:52

That's how you avoid, listen, I would

1:30:54

much rather be embarrassed in public for

1:30:56

being wrong on something than be proven

1:30:58

wrong in the order book and lose

1:31:01

actual P and L. So

1:31:04

it's important not

1:31:06

to shut yourself off from people that disagree

1:31:08

with you, but at the same time, it

1:31:10

has been a little weird getting

1:31:13

used to having a larger following and

1:31:16

you're constantly just barraged with

1:31:19

people that just

1:31:21

disagree with you solely because you're

1:31:24

you or because they're positioned differently.

1:31:27

That is less helpful, obviously. Yes,

1:31:29

yes. But I had a friend

1:31:31

who said something, or actually you

1:31:33

can follow him on Twitter, his

1:31:35

handle is qfresearch. He

1:31:37

said something pretty interesting about, he's

1:31:39

always comparing, which

1:31:41

people have done since this started to

1:31:44

comparing like Cisco in 2000 to

1:31:46

Nvidia now. And his

1:31:49

contention for the past year,

1:31:52

even longer than I've been in it, has been like, this

1:31:54

is not the same yet. And

1:31:57

that like, if you wanna see about...

1:40:00

more difficult to predict the

1:40:02

qualitative aspect of the behavioral aspect where

1:40:04

you say, okay, well, you know, but

1:40:07

during, you know, last October,

1:40:09

when we were, you know,

1:40:11

going down and that kind

1:40:14

of cynicism and skepticism was

1:40:16

returning, would I have had

1:40:18

the guts to buy

1:40:20

the dip there if I didn't already

1:40:22

have this pre-existing hedge? And,

1:40:24

you know, I'll tell myself that the answer is

1:40:28

no, because then I can, you know, go

1:40:30

forward with the idea that I'm protected. Right.

1:40:33

And when you buy a put on the S&P,

1:40:35

you're both making a

1:40:37

bet on implied volatility, which is probably going to

1:40:40

be a negatively absolutely

1:40:42

returning vehicle because it's upward

1:40:44

sloping. It will trade at a premium

1:40:47

relative to realized volatility. You're going to lose most of

1:40:49

the time, yada, yada, yada, and as well as just

1:40:51

being short, you know, negative delta, the S&P 500. As

1:40:53

you said, it only works if you're a better stock

1:40:55

picker than the S&P 500, which is very hard. I

1:40:57

would say 90, maybe even 95 or 98%, 99% of

1:40:59

people, and maybe even, you

1:41:04

know, 90% of professionals. That's not true for it. I

1:41:06

think you may have a good reason to

1:41:08

think why definitely in this environment, you,

1:41:10

I mean, you know, you have been picking stocks

1:41:12

that are crushing the S&P 500. You

1:41:14

hope that continues and,

1:41:17

you know, it may continue, it may not, but,

1:41:19

you know, that is why, you know, it works

1:41:21

for you. But if someone's like picking stocks in

1:41:23

their runtime and then shorting the S&P against that,

1:41:25

they're basically making a bet that they're better than

1:41:28

the S&P 500, which professionals routinely

1:41:30

do not do year in and year

1:41:32

out. James, what do you think about

1:41:34

when you buy an S&P, you know, $5,200 put, is that $5,400? The

1:41:39

S&P 500 stays at $5,400 or even goes

1:41:41

up or doesn't get close, doesn't

1:41:43

decline close to that strike. And every day,

1:41:45

the day's expiration gets lower and lower. You

1:41:47

can sell and keep on rolling or just

1:41:49

let it expire. Yeah. I think that in

1:41:51

order to get the next leg up, sometimes

1:41:53

you have to send puts to put heaven.

1:41:55

I'll let it expire and

1:41:58

then reassess. But normally, when that

1:42:00

happens, you can say, well, you

1:42:02

know, I guess I didn't have to hedge, right. And if something

1:42:05

happens, you know, this happened in

1:42:07

April with the kind of like

1:42:10

that geopolitical volatility thing. And, you

1:42:13

know, luckily around that time, my

1:42:16

overall allocation was still kind of,

1:42:19

it's, I think it still would have slightly

1:42:21

outperformed even if I hadn't hedged.

1:42:23

But there is no easier way

1:42:26

to generate alpha than to have

1:42:28

your base, like

1:42:30

your equity curve goes sideways when the index goes

1:42:32

down. It's getting increasingly difficult

1:42:34

to generate alpha against the benchmark and

1:42:37

as like the past month, you

1:42:40

know, I'm, I think I'm, I'm technically underweight

1:42:42

in video compared to the S&P, you know,

1:42:46

yeah, just, I mean, yeah, but how much,

1:42:48

how much of the S&P does Nvidia make

1:42:50

up right now? Well, because as Nvidia goes

1:42:52

up, you keep on selling and rebalancing it

1:42:54

a little bit lower, keeping it at a

1:42:56

certain weight, whereas S&P doesn't care. It's like

1:42:58

if, if Nvidia gets bigger and

1:43:00

bigger as a percentage of the S&P, that's fine.

1:43:02

And that's why we, you know, Apple is so

1:43:04

big. Yeah. So, so, you know,

1:43:06

it gets a little more difficult to to

1:43:08

outperform the index and keep your risk within

1:43:11

the parameters that you want it to be. It's

1:43:13

a lot easier to, if

1:43:16

you really have like

1:43:18

a strong reason to that,

1:43:21

the most effective hedge that I had all year

1:43:23

was if you were, I

1:43:25

think it was April 4th, Israel had retaliated against

1:43:27

Iran. Yeah.

1:43:33

Israel retaliated against Iran. They sent some missiles out

1:43:35

and there was a big kind of,

1:43:37

you know, oh, this is going to escalate into a regional conflict.

1:43:39

And I was not playing that game.

1:43:41

I was, okay. Like if that's what

1:43:43

we're talking about now, then I'm going to buy

1:43:46

some puts and, you know, and that

1:43:48

worked out well because not

1:43:50

necessarily because of that reason. And, you

1:43:52

know, I don't really like the idea

1:43:54

of like, oh, fading stocks

1:43:56

because of geopolitical reasons, because that doesn't work out

1:43:58

super well historically.