HomeVideos

What Investors Don't Understand About the Software Meltdown | Deiya Pernas

Now Playing

What Investors Don't Understand About the Software Meltdown | Deiya Pernas

Transcript

1824 segments

0:00

If you're invested in a mature SAS

0:01

company trading in north of seven times

0:03

EV to sales, I think there's just a lot

0:05

of vulnerability there. And what we've

0:06

seen so far really is just multiple

0:09

compression, I don't think we've seen

0:11

disruption risk really priced in SAS

0:13

yet. If you're investing in a SAS

0:15

company that isn't somehow using AI to

0:18

make the product better,

0:20

uh I I would rethink the company

0:23

strategy. Buying SAS for us has been

0:24

part of a broader kind of barbell

0:26

strategy which is really pairing what we

0:29

think are AI winners with AI losers

0:32

quote unquote that are mislabeled as AI

0:34

losers.

0:35

>> Welcome to Monetary Matters. Today I'm

0:37

speaking with one of the best performing

0:38

independent equity research providers

0:40

and not just measured by a paper

0:42

portfolio or a few big calls but by an

0:44

audited real money portfolio. Paris

0:47

Research has compounded capital at over

0:49

30% since they began tracking their

0:52

portfolio in 2017. They're coming off

0:54

another great year in 2025, up over 52%.

0:58

I'm joined today by co-founder Dea

0:59

Paris. Thank you so much for joining me

1:01

today.

1:01

>> Very happy to be here. Thank you for the

1:04

introduction and I'm looking forward to

1:05

a good conversation.

1:06

>> Dea, one of the things I I like about

1:08

your research is the way you approach it

1:10

like a portfolio manager and that

1:12

includes writing quarterly letters. So,

1:14

you just released your Q4 2025 letter

1:18

and one of my favorite parts was the

1:20

section where you posed 10 questions for

1:22

2025. Some of them are big picture

1:25

questions, others are more idiosyncratic

1:28

relating to, you know, very specific

1:30

sectors and positions, but I want to

1:32

start off with one of the ones that I

1:34

think is is the biggest and super timely

1:36

for what's happening in the market right

1:37

now. and that's will businesses continue

1:40

to allocate the same percentage of spend

1:43

to software. Why do you think this is

1:45

such a key question for the market in

1:47

general but also for the companies that

1:49

that you guys are invested in and

1:51

researching?

1:52

>> Yeah. So, uh we did list those

1:54

questions. They were meant to be

1:56

provocative. Uh there there's you know

1:59

there's a good amount of uncertainty no

2:01

matter how uh concrete your

2:04

understanding is of each and that the

2:07

software question is one of them. You

2:08

know are businesses going to continue to

2:10

spend the same roughly the same

2:12

percentage of their budget on software

2:15

and you've seen a significant amount of

2:17

market movement. There's been the SAS

2:20

apocalypse dubbed the SAS apocalypse. I

2:23

think that's a bit dramatic given the

2:25

movement so far. most SAS SAS type

2:28

companies are down about 10 to 50% year

2:30

to date. Um but there are a lot of

2:33

questions for software investors that

2:35

they haven't had to ask before like it

2:38

used to be the case you know analyzing

2:40

SAS business is one of the easiest

2:42

businesses you can analyze. You find a

2:44

company uh it's led by you know led by

2:47

founder product visionary uh you

2:50

understand the product you understand

2:51

the total addressable market uh yeah

2:54

there's some calculations around uh CAC

2:56

and LTV uh but after that you kind of

3:00

you you kind of have it all mapped out

3:02

you understand how they're going to grow

3:03

it's a capital light business model and

3:06

there's the the rule of 40 where a

3:08

company's growing a certain percentage

3:10

and if they have the margins to match it

3:12

you're kind kind of done. You can go to

3:14

the beach and watch this thing compound

3:15

for a while. Uh but now uh SAS investors

3:18

are having to start ask new questions

3:20

about you know disruption and um claude

3:25

and this year especially uh uh clog

3:28

which is anthropics coding AI at least

3:31

perception wise has seen a huge step

3:33

function improvement and the marginal

3:35

cost of producing software is rapidly

3:39

decreasing. It used to be that to build

3:41

a good SAS uh product, it took about

3:44

seven years, seven, eight years or so.

3:47

And if that time has been reduced to 6

3:50

months or less, then you start to have I

3:52

mean there's a whole a bunch of

3:54

questions about potential entrance and

3:56

how do SAS companies perform in an AI

3:59

native world and the market doesn't seem

4:01

to be doing a lot of discerning. We like

4:04

um

4:06

we like areas where it seems like the

4:08

market's confused. Uh it's not really

4:10

discerning enterprise SAS from uh

4:13

mid-market SAS from consumer SAS,

4:15

horizontal from vertical. There's

4:16

there's no sense of discernment at all.

4:18

It's just, you know, everything's

4:20

selling off. And that's the type of uh

4:22

environment that we think uh at least

4:26

for an active uh equity

4:29

uh analyst like ourselves, you can

4:32

develop a differentiated perspective and

4:35

um and if your differentiated

4:37

perspective is correct, there's that

4:39

opportunity to produce uh alpha and we

4:41

you know we can get on all the issues in

4:43

some of the software space but just

4:44

broad picture that's why we think it's

4:46

interesting. But this is a case where

4:48

you have a sector that has been sold off

4:50

tremendously. I mean we could go through

4:52

some of the battleground names out there

4:54

like Salesforce, HubSpot, like these are

4:58

the types of companies that are you know

4:59

off I think many of them over 50% from

5:02

their highs and there are people out

5:03

there who are just kind of buying

5:05

blindly betting on this mean reversion.

5:07

How do you you know specifically because

5:09

I know you guys are are are long in your

5:11

portfolio. How do you make sure you're

5:13

not just being a mean reversionist and

5:15

you're you're really trying to get into

5:18

the questions around does this company

5:20

have advantage in distribution? Does

5:22

this company what is sticky about this?

5:25

This is why we think there's going to be

5:27

uh uh sizable opportunities in the

5:29

software space because I don't think uh

5:32

the market is nuanced in oh like is

5:36

enterprise SAS just as disruptable as

5:39

some sort of consumer SAS and we don't

5:42

think so. We think for many of the

5:43

reasons you've said I mean enterprise

5:45

SAS uh um there's a lot more bells and

5:48

whistles to it. You know, companies care

5:50

about uh compliance, companies care

5:52

about securities, companies care about,

5:54

you know, workflows across departments.

5:57

Uh and u enterprise needs the confidence

6:01

of having a big time a big team behind

6:04

some of the key uh software that they

6:06

utilize. But that doesn't also mean that

6:10

the pricing isn't going to change. Uh

6:12

SAS, enterprise SAS typically works on a

6:14

per seat model. um you could see that

6:17

transitioning to a usagebased or

6:20

contract negotiation start to look

6:22

slightly different which is al which

6:24

could also disrupt growth. So even if

6:26

you're confident about the product you

6:28

know you got to think about the pricing

6:29

a little bit. Uh you talked about mean

6:32

reversion and uh yeah in our letter we

6:34

did critique kind of lazy type of value

6:37

investing. We we call it mechanical

6:38

value investing where uh you're looking

6:41

at companies that have just sold off low

6:43

PE and you're saying you know there's

6:44

usually some sort of uh problems. Um and

6:48

you're saying that well look you know

6:49

when the company turns around a lot of

6:51

this pessimism is extrapolated and

6:53

definitely going forward. Uh the market

6:55

tends to be overly pessimistic when

6:56

there's bad news and when there's when

6:59

the company starts to turn and there's a

7:01

little bit of good news we'll get a

7:02

reperic and uh we'll enjoy some alpha

7:05

that way. I mean that type of mean

7:07

version could work in certain

7:08

environments that are relatively

7:10

stationary.

7:11

Um but when you have environments like

7:14

the one that we believe we're in today

7:16

that there's uh there's a lot of

7:17

structural change um in many different

7:20

areas that we can get into, you have to

7:22

be very very careful about thinking that

7:24

things are going to go back to the way

7:25

they were. Um or we can just bet that

7:28

this temporary problem this company's

7:30

experiencing is going to be fixed. uh

7:32

that lazy type analysis doesn't work and

7:34

you have to have concrete justifications

7:36

why problems are going to be fixed or

7:37

problems are temporary uh and you really

7:40

have to anticipate what the future

7:42

contours of that business are going to

7:43

look like. So it so our whole investment

7:46

philosophy is about leaning into

7:47

prediction. Um and obviously you have to

7:51

understand the historical financials and

7:53

u you know just how the psych you know

7:56

is it a cycl type business ecosystem and

7:59

so on but you have to put everything

8:01

together in a forward-looking view. It's

8:03

not enough to just look at past trends

8:05

and expect them continue or expect them

8:07

to reverse there. that lazy type of

8:09

thinking uh is completely decimated and

8:13

you have to really be concrete in how

8:15

you're thinking about the future. So

8:17

when you look at, you know, the hundreds

8:19

of SAS names that are out there that are

8:21

that are down tremendously, what are

8:23

some of the the larger companies that

8:25

you look at and you say, "Yeah, I think

8:27

the market has this right. This business

8:28

is in trouble." Uh whether it's its

8:32

pricing as you said or the actual

8:35

product itself is is going to fall out

8:36

of favor as new entrance either pop up

8:39

or already have. If you look at the SAS

8:42

complex right now, it's still in in my

8:45

opinion and we tend to be more price

8:46

sensitive than others. Uh it's still

8:49

very overpriced. If you look at

8:51

companies like Salesforce, Adobe,

8:54

Autodesk, Service Now, and you just look

8:56

at median valuations as a whole, EV to

8:58

forward sales are still six to seven

9:00

times, which is expensive. We start

9:03

getting interested in SAS companies when

9:05

they're trading below five times EV to

9:07

sales. So despite the underperformance

9:10

of SAS last year, despite uh SAS

9:13

companies being down 10 to 15% this

9:15

year, we still think they're overvalued

9:17

and uh there's another I mean I mean if

9:21

you're invested in mature SAS company

9:22

trading at north of seven times EB

9:24

sales, I think there's just a lot of

9:26

vulnerability there and what we've seen

9:28

so far really is just multiple

9:30

compression. I don't think we've seen

9:32

disruption risk really priced in SAS

9:34

yet. So when you're saying multiple

9:36

compression that enterprise value to

9:38

sales, you're saying enterprise values

9:40

have come down. We haven't even really

9:42

seen the sales the denominator really

9:45

get hit. Have the concerns that people

9:47

have been talking about with disruption

9:49

to sales shown up in in that denominator

9:52

really at all in the industry.

9:53

>> You've seen some sales slow down

9:55

slightly with some, you know, horizontal

9:58

type SAS. uh it it appears that

10:01

companies are just a bit more cautious

10:02

about uh about SAS spend. So if it comes

10:06

to maybe a new SAS product or something

10:08

like that, it's like okay, how many

10:09

people do we actually need using this

10:11

product? Uh does a SAS pricing model

10:14

make sense? Um let's just be a little

10:16

bit more cautious. And that is that that

10:19

has translated in the sales figures of

10:21

some companies. So you've seen a slight

10:23

slowdown in kind of sales growth. Um, so

10:26

a lot of some of those companies have

10:28

come down as a result of that sales

10:30

growth, but given the disruption risk

10:32

that I'm referring to we haven't really

10:35

seen that much of that yet. So I think

10:37

that the um most of the complex is still

10:40

trading too rich and we're going to see

10:41

it uh generally start to come down. What

10:44

are the parts of the complex that you do

10:46

like that you believe we are going to

10:49

see the same allocation of spend to um

10:53

or or even growth in spend?

10:56

>> We think that uh more vertical type

10:58

software uh that really understands

11:00

their niche is probably better protect

11:02

protected than something that's more

11:04

horizontal. We think enterprise SAS is

11:06

more protected.

11:07

um you know so but apart from that just

11:12

very broad uh just those very broad

11:15

guidelines it's very very nuanced and if

11:18

you're investing in a SAS company that

11:21

isn't somehow using AI to make the

11:23

product better

11:25

uh I I would rethink the company

11:27

strategy almost most companies we look

11:29

at are either trying to move up to

11:31

enterprise if uh because it you know

11:33

it's stickier there's been some softness

11:35

in uh kind of small medium business

11:37

space and they're using AI somehow to

11:39

make their product better or at least

11:41

they're trying. Uh a lot of them are

11:43

going to fail at it, but there it does

11:46

certainly appear to be that sense of

11:48

urgency in that in the SAS world.

11:50

>> Okay. And the way you construct your

11:52

portfolio is you have your core

11:54

positions which are about 5 to 15% each.

11:57

You have starter positions in the 3%

11:59

range and then you have your more

12:00

speculative trades in the 1% range. Have

12:03

any SAS names made it into any of those

12:05

buckets that we could talk about today?

12:08

>> Buying SAS for us has been part of a

12:09

broader kind of barbell strategy which

12:11

is really pairing what we think are AI

12:14

winners with uh

12:17

AI losers quote uh quote unquote that

12:20

are that are mislabeled as AI losers. So

12:22

they're not really AI losers, but the

12:24

market the perception seems to think so.

12:26

Sam Rush is a company we invested in

12:28

about four months ago. They originally

12:30

were known uh as an SEO provider for uh

12:33

small medium businesses and they

12:35

expanded that product suite uh and they

12:37

moved up in enterprise and they started

12:39

solving problem for enterprise customers

12:41

which is uh hey how am I showing up in

12:43

different LLMs? It wasn't just about

12:45

keywords anymore. If I'm if I'm ranked

12:47

high if somebody searches some keyword

12:49

am I ranked high in that list? It's more

12:51

about hey are what are customers asking

12:54

in LLM's and and how what are what are

12:56

LLM saying which is which is becoming a

12:58

more and more important topic. So they

13:01

were making uh large investments in this

13:03

area and it was about 10 to 15% of

13:06

revenues and that that part of their

13:07

business was growing uh about 30 to 40%

13:11

or so. So they had an AI solution, they

13:14

had traction,

13:16

they have a very large installed base of

13:17

enterprise customers which are going to

13:19

continue to penetrate. So the story for

13:21

us was that oh the market seems to think

13:23

that this company is an AI loser because

13:26

maybe companies are going to start

13:27

spending less on SEO and more on uh more

13:31

on LLM type commercial searches. But it

13:35

was very very clear to us that this

13:36

company's moving in the right direction

13:37

and they end up getting acquired by

13:39

Adobe to bolster their AI solutions. So

13:42

that that was a great example of a

13:44

company that was mislabeled as an AI

13:46

loser. So that's what we're one of our

13:48

prime areas of focus is trying to find

13:51

those types of companies.

13:53

>> So you mentioned in this mean version

13:54

section that there are structural

13:56

changes underfoot. I mean, we've talked

13:58

about one of them, which is AI and the

14:01

way that that companies are looking for

14:03

AI solutions and and changing, you know,

14:05

with the SEO as an example. Um, changing

14:08

what they're focused on. What are the

14:10

other structural changes that you are

14:12

really focused on that you think if

14:14

investors are not paying attention to,

14:16

they're going to have a really hard time

14:18

in 2026?

14:20

just general consumer behavior uh

14:22

enterprise behavior around AI, how are

14:24

enterprises using AI, how is that going

14:27

to change, uh their uh relationship to

14:32

labor moving forward and which areas of

14:35

uh which areas of labor are more

14:37

affected than others. Uh right now it

14:39

seems to be that white collar slowed,

14:41

entry level positions that can be

14:43

automated away are uh seeing trouble. um

14:47

consumers uh how are consumers going to

14:50

uh shop for things? How is that going to

14:53

change? I mean, the big question on

14:55

everybody's mind right now is commercial

14:57

intent and LLM. Exactly how that's going

14:59

to move over. Uh there's been a lot of

15:02

chatter about, oh, oh, OpenAI is going

15:04

to start putting uh you know, these buy

15:06

now buttons on these searches, but then

15:08

it's like, well, wait a minute. know, is

15:10

it true that consumers just want to type

15:12

a question in and have one product show

15:14

up and then purchase it? Or do they

15:17

still want the whole uh carousel of

15:20

choice that they, you know, or images

15:23

they can scroll through from something

15:24

like Google search uh to make purchasing

15:27

decisions. And if that's true, then

15:28

maybe we're back to the same kind of

15:30

Google search interface. So, uh I think

15:32

there's just a lot of questions about

15:34

how uh you know, how consumers going to

15:37

shop online. I think there's a lot of

15:39

questions about how their consumers are

15:41

going to uh consume entertainment

15:44

in an uh in an AI world like what what

15:48

the entertainment consumption is going

15:49

to look like social media how businesses

15:52

are paying for ads h how uh just how

15:57

many iterations can you know if you're

15:59

paying for an ad on meta and um just how

16:02

good those ads iterations are given

16:05

they're AI driven given that uh the

16:08

content generation can be so quick and

16:10

the iterations uh you know continue to

16:14

just in real time improve uh improve the

16:18

level of engagement.

16:19

>> There's a big regulatory component to

16:21

that too. I don't know if you saw that

16:22

study from I think it was NYU where they

16:25

said AI generated ads were you know

16:28

obviously more effective than both AI

16:31

plus human and human only generated ads.

16:35

But then all of that outperformance went

16:37

away and it actually went negative if

16:38

they had to disclose that the ad was

16:40

generated by AI. Okay, great. We have

16:42

all these AI ads. They're more

16:44

effective, but you have to put a little

16:46

sticker in the bottom that says this is

16:47

AI. Like it really affects the the

16:50

person's perception. So it's it's not

16:52

just does the technology work. It's are

16:54

the regulators going to let these

16:56

companies run wild or are there going to

16:58

be more guard rails

16:59

>> which I assume is going to follow its

17:01

regular pattern of being significantly

17:03

late. uh the space is just moving so

17:05

fast. I think the regulators are are

17:07

going to be so behind. I mean, we're

17:09

still talking about crypto regulation

17:11

today. Uh so um yeah, I agree there's

17:15

that perception of it. And then what

17:16

exactly is uh something created by AI?

17:19

What if there's a human element plus an

17:21

AI? Does that constitute is AI

17:23

generated? But yeah, I agree. There's

17:25

this perception of if it's AI created,

17:27

then it's kind of slop and well then,

17:29

you know, I don't I don't want to see

17:30

something that's totally commoditized.

17:32

Uh but then again if they don't have to

17:34

disclose that then you know the the uh

17:37

consumer is not going to be able to tell

17:38

the difference. Again the space is

17:41

evolving very fast. We have we don't

17:43

know exactly how this is going to

17:45

develop but these all these thought

17:46

experiments are interesting because at

17:48

least it keeps you uh flexible as the

17:52

you know as things continue to evolve.

17:55

>> What are some of the thought experiments

17:56

where you do have a house view at this

17:58

point? You've looked at this question,

18:00

you've asked it, and you think at this

18:02

point, obviously facts can change,

18:04

anything can change in a in a second,

18:06

but you do have a view for how this

18:09

question is going to to shape up over

18:11

the next year.

18:13

>> SEO like marketing a business spending

18:16

its marketing dollars on SEO, we think

18:18

that's going to be challenged for a

18:20

number of reasons. Uh primarily the

18:23

commercial search intent moving over to

18:26

LLM. Uh right now Open AI handles about

18:30

uh a trillion uh prompts a year versus

18:33

uh Google's 5 trillion or so. So it just

18:36

goes to show you just how

18:38

uh just the level of engagement and

18:40

attention uh that uh consumers are

18:43

investing in in LLM.

18:46

>> A first order thinking I would hear less

18:48

spend on SEO and I would think that's

18:50

going to hurt Google's ad business. But

18:52

at the same time, they're one of the

18:54

leaders in AI and they're making they

18:56

have their own um the TPUs, not the not

18:59

the GPUs.

19:00

>> So how does a view like that translate

19:03

into portfolio decisions?

19:05

>> You have to get down to the nuance a

19:07

little bit. You have to understand that

19:09

yes there is this trend that okay a lot

19:11

of consumers are on uh uh LLM's and what

19:15

exact like is the nature of what they're

19:17

searching on LLM structurally different

19:19

from what you know uh from Google

19:21

searches for instance and the answer

19:23

turns to be turns out to be yes

19:25

commercial search intent is still the

19:27

domain of Google usually if somebody has

19:30

a commercial search intent they're going

19:31

to Google first if they're if they're

19:33

looking for something more instructional

19:34

or more knowledgebased they're going to

19:36

lo And hence Google has uh has adapted

19:40

right you've seen anytime anybody

19:42

searches something Google they see the a

19:43

AI powered search answer uh and that but

19:46

that if you're an advertiser that maybe

19:49

that was part of your keyword well all

19:51

of a sudden that reduces clickthrough

19:53

rates because people as soon as they see

19:54

that AI answer they don't go through and

19:56

click through links which is bad for you

19:58

as a potential advertiser for that

20:00

keyword. So it turns out uh that in

20:03

certain areas it has greatly affected

20:06

traffic for some businesses and certain

20:08

areas not so much. So really

20:10

understanding that what types of

20:11

businesses uh are still uh going to be

20:15

able to advertise and see an ROI from

20:18

that and which types of businesses are

20:20

going to have to think of another

20:21

strategy because for for most businesses

20:23

they could just spend money on Google

20:25

ads and then just go away. It didn't

20:26

really have to be much very

20:28

sophisticated beyond that. But now

20:30

you're starting they're starting to have

20:32

to get more nuanced. Uh social is more

20:35

and more uh uh going to absorb some of

20:38

that spend. That's why we think the

20:40

penetration rate of social as far as the

20:42

overall ad spend is going to continue to

20:44

increase and that's something that we

20:45

think is concrete and is uh you know is

20:48

backed up by the way the data is moving

20:50

just more of the ad spend pie is going

20:51

to go towards social. Uh so that's kind

20:55

of a a trend that we would be a we would

20:57

be willing to bet on. So we our

20:59

prediction in some sense is that is that

21:02

uh social the the amount of spend going

21:06

social is going to increase and

21:07

companies that are situated within that

21:10

should see a tailwind benefit. Uh and

21:12

that's when we think about prediction

21:14

that's kind of our our flavor of

21:16

prediction. That's how we do it. It's

21:18

grounded in evidence. It's grounded in

21:21

not only anecdotal but you're marrying

21:22

the anecdotal with quantitative. you

21:24

have conviction in a trend and then

21:27

you're looking for companies that can

21:28

benefit from that.

21:29

>> What about the the types of products or

21:33

services that are not going to be or are

21:36

not currently being disrupted as much?

21:39

Um, is it high-end goods? Is it low-end

21:43

goods? Are these things where people are

21:45

searching for brand names and not a

21:47

product type? Like what are the types of

21:49

businesses that are not seeing that

21:51

click-through hurt as much?

21:54

businesses that I think that have spent

21:56

just more time uh and energy on their uh

21:59

keywords that have more of uh more of a

22:04

a PR presence as well. So given whatever

22:08

somebody searches on uh on Google,

22:11

they're they seem to be coming up a lot

22:13

more in the blue links. So, it's

22:16

companies that are larger that have

22:18

invested a lot of money and spend in

22:20

building out this uh SEO infrastructure.

22:24

Uh I think smaller companies are the

22:25

ones that are hurt from a lot of this

22:27

because they don't have the

22:28

sophistication or the spending dollars

22:31

to consistently rank with all these

22:32

keywords.

22:33

>> That's interesting to me because and I

22:35

you have small companies in your

22:37

portfolio, you have large companies in

22:39

your portfolio. In your latest research

22:41

email, you had, you know, a a massive

22:45

massive company and a couple of

22:47

subbillion dollar companies, but you did

22:49

have a really interesting appendix in

22:51

your letter talking about the number of

22:52

multibaggers in 2025 and just how big a

22:56

year it was for small companies. So,

22:58

it's interesting to me that you think

23:00

these small companies are going to be

23:01

hurt, but I see them in your portfolio

23:03

and clearly that's one of the things

23:06

that worked a lot last year. How do you

23:08

square this concern for the smaller

23:10

businesses with the fact that that's

23:12

what's working in the market?

23:14

>> For most of the small companies that we

23:16

invested in that we've seen outsized

23:17

returns in, those have not been product

23:19

oriented. It's been a result of uh large

23:23

uh data center type spends, companies

23:26

that are almost small industrials that

23:29

have pivoted uh you know smaller

23:31

companies by their nature are there's

23:34

just less you know inertia there. So

23:36

they they're able to pivot quicker than

23:37

larger companies to take advantage of

23:39

certain opportunities. Uh so it's been

23:42

company, you know, smaller type

23:43

industrial companies that have pivoted

23:44

towards these large trends and seeing an

23:46

influx uh in demand. Uh and Capstone is

23:49

one of these one of these companies that

23:51

we invested in last year. Uh they're a

23:54

maker of microturbans

23:56

and they're one of the dominant

23:58

manufacturer of uh gas power natural gas

24:00

powered microtur.

24:02

And this is a company that merged from

24:03

the industry last year. uh had new

24:06

management. They're really just like a

24:08

kind of a large assembler and they have

24:10

about a couple thousand vendors and they

24:12

went through every single vendor

24:13

relationship and try to you know

24:15

rationalize spend uh along with

24:17

rationalizing their pricing and that

24:19

went a long way to achieving their kind

24:21

of first quarter of uh profitability.

24:23

And it turns out they have a a perfect

24:25

type of product for data centers that

24:27

are looking for off-grid solutions.

24:29

Everybody knows about the strain power

24:31

grid. Everybody understands that data

24:34

center uh energy uh usage is going to

24:37

continue to increase. It's continue to

24:38

make a a bigger percentage of that

24:41

energy consumption and a lot of the

24:43

alternative energy providers are going

24:45

to benefit from that because the grid

24:46

isn't going to be able to do it itself.

24:48

So off-grid solutions like microurbons

24:52

uh can pivot towards this vertical and

24:55

enjoy uh you know enhanced revenue

24:58

growth. So that that was something. So

25:00

that that's really a just hey find the

25:03

right company and get there early type

25:05

story.

25:06

>> It's relatively small still even after

25:08

the tremendous year it had in uh 2025.

25:12

So what is your process like for

25:14

uncovering these smaller companies

25:17

>> you're already looking for the

25:19

companies. So uh it's like hey uh the

25:22

the over the overarching thesis is quite

25:24

clear. Hey, off-grid solutions are going

25:26

to become more and more important.

25:27

Alternative energy providers.

25:29

Find ones that are pivoting that have

25:31

the right leadership and cost structure

25:33

and so on where the pivot makes sense

25:35

and they're or ideally they're already

25:37

getting some sort of traction. We don't

25:39

like to uh like turnarounds or even

25:43

companies that have pivoted where they

25:44

have not seen traction yet. It's a it's

25:47

still a bit too nebulous for us. we like

25:49

to see some sort of traction before uh

25:52

we enter a position and we start getting

25:54

really excited about something when

25:55

we're the first ones to be talking about

25:57

it or thinking about it. Uh and if

25:59

you're first, there's going to be a

26:01

turning over. If you're correct about

26:03

your thesis, there's going to be a

26:05

turning over investor base. There's

26:08

there's going to be a very long rerating

26:10

cycle. And there's no better feeling

26:13

than being at the at the very beginning

26:15

of that wave. Uh it's it's unbelievable.

26:18

>> What does the beginning of the wave look

26:19

like? You talk about traction. What are

26:22

the types of things you like to see in a

26:23

company? It does it have to show up in

26:25

earnings or can it be press releases

26:28

about deals being signed that you know

26:29

are not going to show up in the

26:31

financials until quarters down the line?

26:34

>> It goes back to something that we think

26:36

about uh like what it's what we call

26:38

motor investing. We're looking for

26:40

companies where the motor of the company

26:41

is getting stronger. And what that means

26:44

is it's kind of like uh the potential

26:46

energy is getting stronger but but it

26:48

hasn't yet translated to kinetic energy.

26:50

It's almost like if you envision the

26:52

pulling back of a bow so to speak. It's

26:54

like there's more and more and more

26:56

potential but there it hasn't really

27:00

like it's it hasn't the income statement

27:02

yet or hasn't hit the revenue growth

27:03

yet. And that's what we really really

27:05

like to see.

27:06

>> Okay. Well, sometimes the bow snaps. How

27:08

how do you uh how do you manage risk?

27:11

How do you deal with hey there's a lot

27:13

of potential energy building here and

27:15

you know what somebody else took it all

27:18

they are the big winner this isn't the

27:19

winner

27:19

>> the most important part is having a

27:21

belief system that is dynamic and that

27:24

updates on the face of information and

27:27

you have to be very uh you know

27:30

attracted to disisconfirming evidence

27:33

and you continually be testing your

27:35

thesis and if new information emerges

27:37

that contradicts that you better pay

27:39

attention to it and ask

27:41

So, uh it's really about like you have a

27:43

belief, it's updated, it's it's fluid

27:45

based on new information and you have to

27:47

you have to rely on your own ability to

27:50

kind of update your uh your evolving

27:52

conceptualization of where you think

27:53

this business is headed for the like you

27:56

said, yeah, things go ary uh let's say

27:59

Capstone. If this is a company that's

28:00

already been bit up quite a bit for a

28:03

lot of fundamental reasons, uh they've

28:05

achieved profitability.

28:07

They've uh they've rationalized pricing.

28:09

So revenue growth has uh proceeded in

28:11

line there. The valuation right now is

28:15

starting to get a little rich if they

28:16

don't sign some uh uh some of these

28:20

larger customers that we're talking

28:21

about. So if it's you're also looking at

28:23

valuations too, right? So if let's say

28:26

valuations keep going up and up and up

28:28

and then uh they haven't really made any

28:30

more headway to the uh to adopting some

28:33

of these data center customers. Well

28:35

then that's a position we're going to be

28:36

looking to trim or to get out of or uh

28:38

or whatnot. We we don't think that's

28:40

going to we do think they are going to

28:42

uh those plans are going to come to

28:44

fruition. But yeah, it's a calculus.

28:46

You're balancing uh updating beliefs

28:49

with valuation

28:51

uh and and you're combining that all

28:53

together. uh to make real-time

28:56

decisions.

28:56

>> And I can tell from reading your letter,

28:57

it's definitely not 100% price based.

28:59

Like there some investors who say

29:01

position pulls back this amount like

29:03

we're done with it. You know, there was

29:04

a name uh that you wrote about in your

29:06

letter that, you know, has maintained

29:08

its its position in the portfolio coming

29:10

into 2026 despite, you know, a pretty

29:12

sizable draw down. So, you know, I think

29:14

everybody has had a stock that they were

29:17

confident on pull back in a way that

29:19

they weren't expecting, especially when

29:21

you love the fundamental thesis and the

29:23

market just doesn't agree. How do you

29:25

decide between being wrong and and

29:28

sticking with your guns um in in a name

29:31

that that pulls back a considerable

29:33

amount?

29:34

>> Remittly is a good example of this

29:35

because it's been our biggest position.

29:37

Well, it's alternated one and two for

29:39

quite some time and it's been a position

29:43

where we're probably a little early in

29:45

the name around 2023

29:48

when it was trading north of five times

29:50

sales and the whole thesis was quite

29:53

simple. Uh, remitt is a crossber

29:55

remittance provider. So, uh, you know,

29:58

think of migrants uh, you know, for

30:00

example, come to United States who send

30:02

money uh, back home. So, it it provides

30:05

a digital solution. It's a best-in-class

30:07

digital solution. It's a founder

30:09

company, Capital Light. Uh its main

30:13

competitors are dinosaurs like Western

30:15

Union and Moneygram. So, they're very

30:18

quickly taking share uh from these

30:20

players along with a very broad and this

30:23

goes back to kind of prediction uh what

30:25

we're talking about like really

30:26

understanding these broad-based trends

30:28

and having a lot of conviction on them.

30:29

The the other part of the the trend is

30:32

that okay well uh there's going to be a

30:35

lot of cash moving to digital. Um so in

30:38

the past when migrants would send cash

30:40

home to friends and family now they're

30:42

using a digital means send money home

30:44

like mobile. So clearly the you know the

30:47

best-in-class mobile player is going to

30:48

be a beneficiary of that. Add to that

30:51

the uh the developed world is aging

30:54

quite rapidly. Um birth rates are not

30:57

keeping up with replacement. If you look

30:58

at places like Italy, places like

31:01

Germany, uh Japan, uh those the median

31:05

ages are getting close to 50. Uh the

31:08

median age of those uh countries. The

31:11

US, I think the median age is something

31:12

like 39. Uh which is still uh you know,

31:15

which is still getting up there. US is

31:17

one of the few developed nations that is

31:19

keeping up birth rates are keeping up

31:20

the replacement. But you're having you

31:22

have the situation where the developed

31:24

world regardless of the national

31:26

populist rheto rhetoric uh that is going

31:29

around the globe right now they're going

31:30

to need migrants to plug those uh

31:33

population gaps

31:35

and because of that that is a trend that

31:37

is going to continue and again if you're

31:39

a crossber remittance provider you have

31:41

the best supplies digital solution you

31:42

are going to benefit from that trend.

31:44

Um, so what happened with remitt is we

31:47

ended up being completely right on the

31:48

fundamentals since we've invested in the

31:50

company. They've grown revenue at 30%

31:52

plus the entire time. The stock

31:55

unfortunately has just continued to sell

31:57

off stock and we're down uh as last year

32:00

down quite considerably on the name. Uh

32:03

currently we think it's trading at an

32:04

unbelievably attractive valuation. We

32:06

believe that the company's going to be

32:07

multibagger. uh you just this is a

32:10

situation where the price is moving one

32:13

way and the fundamentals are agreeing

32:14

with you and you just have to you have

32:16

to just keep adding unless you see

32:18

something some disisconfirming evidence

32:20

that it's going to prevent you from

32:22

doing so but you just have to keep

32:24

adding to the position. Uh again like

32:27

you have to be paranoid you have to

32:28

retest your thesis. Uh you know am I

32:31

wrong? Could I could I miss something?

32:33

Is there a potential terminal value risk

32:36

that I'm not thinking of with stable

32:37

coins or something like that or

32:38

whatever? And then you retest that

32:39

thesis and then after you think about it

32:41

in an objective way, then you go back

32:43

and you say, "No, I don't think that's

32:44

true. I'm I'm going to continue adding."

32:46

Um,

32:46

>> are they utilizing stable coins at all?

32:49

>> Uh, they're util utiliz they're uh

32:51

talking about utilizing stable coins in

32:53

a treasury function. Uh, not to help

32:56

directly with consumer transfers. Um so

33:00

yeah in a treasury function only stable

33:03

coin right now it's it's mainly a crypto

33:05

phenomenon. We uh we are very convicted

33:09

against stable coin kind of disruption

33:11

disrupting the normal kind of flows of

33:13

money given that uh you know people in

33:16

their local economies spend money in

33:18

fiat unless that changes unless uh

33:21

somebody in their local economy is going

33:22

down to the vendor and playing with in

33:24

some sort of crypto uh it's very

33:26

unlikely for us to see uh you know

33:29

stable coins as a disruptor of normal

33:31

commerce.

33:32

>> What about remittance? I mean it's one

33:34

of the main if you talk to any crypto

33:36

bullet it's far and away like one of the

33:38

main success stories that they like to

33:40

talk about because people always say

33:42

like what is the real world use case of

33:44

crypto and and that's probably the

33:46

number one that I have seen cited is is

33:49

remittance payments crossber transfers

33:52

is that actually showing up is it taking

33:54

share and you know we we talked about

33:56

wanting to see yes maybe these companies

33:59

are AI losers but they're making shifts

34:02

to address that in inside of their

34:04

companies and and I kind of see stable

34:06

coins in a in a similar in a similar

34:08

vein. Why aren't you concerned about

34:12

stable coins with remitt?

34:15

>> So that phenomena that you're describing

34:18

is completely non-existent as in in very

34:22

in very specific situations where you

34:24

have uh like maybe some African

34:26

countries where you have a total banking

34:28

system collapse. uh they're sending

34:30

money through uh cryptos, but the

34:33

they're uh it's these are definitely

34:36

exceptional cases. So the idea that uh

34:39

stablecoin is going to and I get why

34:42

people think that it's like oh you send

34:43

stable coin it's going to be a lot

34:44

cheaper than remittance or whatever but

34:47

turns out remittances are already uh

34:50

have come down uh you know like from

34:54

maybe about 10 years ago 80 90% as far

34:56

as cost goes. Uh so remittances uh you

35:00

know it's one two percent to send money

35:02

across borders. Um so the you just

35:06

haven't seen that in the data at all. Um

35:08

and and again going back to the main

35:12

uh explanation why I don't think you'll

35:13

see that is because when somebody's

35:15

sending money back home to their friends

35:16

and family, you know, a migrant is

35:18

sending money back home to friends and

35:19

family. They're sending them money that

35:21

they're going to need to use for to

35:23

purchase food, rent, utilities, what

35:26

have you. And that food run utilities is

35:28

spent. They're paying for that in their

35:31

local currency. So they want local

35:34

currency and they're going to use that

35:35

to spend local currency. And as long as

35:37

it's being done in a very very efficient

35:39

way, lightning fast, and the cost is uh

35:42

very very low, uh I don't see how stable

35:46

coins is going to be disruptor of that.

35:47

again

35:49

like um you know we're big on

35:53

you know our industry is funny because

35:56

uh you know clearly you have to kind of

35:57

sound like you you know you're confident

36:00

about your ideas and so on and you've

36:01

done your work in um and we are but that

36:05

is not without us always maintaining

36:07

some small reservoir of doubt in all our

36:09

decision making I mean uh it's just part

36:12

of being a good thinker is you have to

36:14

you have to maintain a bit of doubt with

36:17

everything especially since we're

36:18

talking about the future right uh so if

36:20

I see evidence contradicting that view

36:22

which currently there's absolutely none

36:25

uh then yeah I'll you know we'll revisit

36:28

figure out well why is this happening in

36:29

this area and maybe uh stable coins are

36:31

going to increase penetration overall

36:33

remittances and uh the thesis is broken

36:36

uh but you know that's certainly not

36:39

something we're seeing

36:40

>> and the other side is the person

36:41

receiving the remittance payment is

36:42

usually an older person

36:44

>> yes

36:45

>> so it's going to somebody who is maybe

36:48

not as familiar with crypto. And so

36:50

you're going to have to see some

36:51

improvement in the in the UX and the

36:53

infrastructure for these consumer-based

36:56

crypto apps if you want uh I I used to

36:59

have to travel to Puerto Rico a lot for

37:00

work and I like to call it a island uh

37:03

because it's it's interesting being a US

37:05

territory and that there's free

37:07

migration between the mainland United

37:09

States and Puerto Rico. So, anybody who

37:12

wants to come here, if they can afford

37:14

uh a plane ticket and they can afford

37:16

rent and find a place to live, can come

37:18

here. And so, you're talking about

37:19

median age being about 39. I know in

37:21

Puerto Rico it's about 45. Every time I

37:23

would be on the plane, I'd be sitting

37:24

next to either somebody, it was an old

37:26

person who was visiting their younger

37:28

family who had moved to the US or it was

37:30

a younger person who was coming back to

37:32

visit their older family who had

37:33

remained on the island. And um you know

37:38

you know these are people who are flying

37:40

many times for the first time in their

37:41

lives. Like I can't imagine them like

37:42

opening a crypto app to uh convert their

37:45

stable coins into dollars. And you know

37:48

there's not the local currency problem

37:49

because again it's Puerto Rico. It's

37:50

it's a US uh US territory and so they

37:54

they use dollars there. But um

37:56

>> just just very interesting to to see

37:59

because I think it is like a perfect

38:00

microcosm of the migration trend because

38:02

of the lack of barriers completely to to

38:06

what we see from other countries.

38:07

>> Yeah. And if you look at Puerto Rico

38:10

especially, that island has seen

38:11

significant population decline since

38:14

Maria. Uh I mean I think there's like

38:16

something like three three million or so

38:18

Puerto Ricans in New York. I know

38:20

economists have dubbed this a brain

38:21

drain where any young person with any

38:23

sort of agency is gonna go to where

38:27

there's more opportunity especi

38:28

especially if it's frictionless to do

38:29

that. So I I think that's important is

38:31

you're understanding that customers

38:33

these are migrants. They're not your

38:36

typical uh develop nation wellbanked

38:38

type customers. And this is another

38:41

reason why uh when I talk to people

38:43

they're like oh why don't they just do x

38:45

y and z and we'll be a lot better

38:46

cheaper. I'm like, well, you have to

38:48

understand that how migrants send money

38:52

uh you know which products and services

38:53

they trust

38:55

and uh a lot of that community work uh

38:59

is through word of mouth as well. So

39:02

it's unlikely after you've already

39:04

entrenched yourself in that in that

39:06

community to be easily disrupted just

39:08

because of the nuances of uh you know

39:12

just migrate type culture. Now you

39:14

mentioned that you are are very much

39:16

bottom up when it comes to the the

39:17

companies you invest in. A couple of

39:19

those 10 questions that you asked are

39:22

have big macro implications. When you

39:25

think about these questions, is it in a

39:28

macro through a macro lens and how

39:30

they're going to affect a broader

39:32

economy or are you looking for

39:35

companies that are directly affected by

39:38

that particular trend?

39:40

>> It's a bit of both. And uh like as far

39:42

as some of the macro trends we talked

39:44

about, what one of the trends that I

39:45

think is has not talked about enough is

39:49

the rebalancing in consumer in just

39:52

general consumer spend versus

39:55

uh 2020. Uh you know, and I I think that

40:00

uh one of the things that makes you a

40:02

good equity analyst is you have to be

40:04

somewhat of a social scientist. like you

40:07

you have to be observing the behavior of

40:08

others and saying, "Oh, is this behavior

40:10

a and observing your own behavior?" I

40:12

know we all like to think we're special,

40:14

different, but understanding that which

40:16

parts of my behavior uh are maybe not so

40:19

intentional and are are really scaling

40:21

every like people are doing this

40:23

everywhere. Like I noticed the other day

40:24

I was on uh Uber and I was I I ordered

40:27

uh Uber Eats and I I think I got an Uber

40:30

one uh subscription or something and I

40:32

noticed I'm spending a lot more just on

40:34

Uber with Eat, you know, eats. And then

40:36

I look at go through Uber's uh

40:38

transcripts and they're like, "Oh, Uber

40:40

1 customers spend, you know, two or

40:42

three times as much as your your normal

40:45

uh Uber customer." like oh well that you

40:47

know anecdotally that that immediately

40:49

gives that quantitative data just a

40:51

little bit more conviction given that I

40:53

I understand it at an anecdotal level.

40:55

Uh but as far as uh a broad macro trend

40:58

there's been a rebalancing consumer

40:59

spend you've seen wages have increased

41:02

cumulatively since uh over the last 5

41:04

years since the end of 2020 about 26%.

41:08

And there's been a lot of essential

41:10

categories that have outstripped wage

41:12

growth. When you're talking about health

41:14

care, vehicle insurance, uh shelter, uh

41:18

electricity, all all those uh categories

41:21

have increased 30 to 40%. I was pointing

41:24

the data here. Motor vehicle insurance

41:25

is up 63%. Uh electricity 38%. Meats,

41:30

poultry, fish, and eggs 31%. Again, this

41:33

is cumulative increase over the last 5

41:36

years relative to wages. So you've seen

41:38

a lot of these essential categories

41:39

outstrip uh wage growth which means

41:44

there's less discretionary capacity than

41:46

uh other cycles. So there's been a

41:48

rebalancing in consumer spend. Uh so if

41:52

you're looking at a a discretionary

41:54

category and you're saying to yourself,

41:56

well is this are the levels of volumes

41:58

going to get back to 2020? Uh as a

42:01

category probably not. uh but maybe

42:04

there's individual companies that uh

42:07

that are going to continue to do quite

42:08

well, but it's not going to be a

42:10

category argument. It's going to be a

42:12

more a share taking thesis at that

42:15

point. So that that's that that's kind

42:18

of uh I think that's a macro uh

42:22

structural macro change that I don't

42:24

think is talked about a lot and it get

42:25

gets back to the house housing uh buying

42:28

as well. Like if you is is is are is the

42:31

level of housing activity can you get

42:32

back to 2018? Well, 2018 tenure was

42:35

about uh 2.5 or so. Now the tenure is

42:38

about 4.25. It has come down uh quite a

42:41

bit. But even if it does come down 2018

42:43

levels after we get a a doubish bed

42:45

share that, you know turns on

42:47

quantitative easing uh or or a

42:49

significant amount of quantitative

42:50

easing gets there. Uh you know home

42:54

insurance, HOA fees, a lot of those have

42:56

outstripped wage growth. Uh again the

42:59

price of homes are up up about 40%

43:00

cumulatively which leaves you know

43:03

consumers have to put a lot more down

43:04

for the down payment. Uh obviously

43:07

consumer balance sheets have gotten a

43:09

little better since uh 2000 the end of

43:11

2020 but that tends to be lopsided into

43:14

uh you know higher net worth uh type

43:17

households. So, it's uh my guess would

43:20

be that no, the activity doesn't return

43:22

to 2018 levels, even though you're

43:24

starting to see like a lot of these like

43:26

rocket mortgage type companies trading I

43:28

I think very healthy valuations

43:30

expecting that same amount of activity

43:33

to return.

43:34

And I would bet well I mean just

43:37

psychologically I would bet against it

43:39

but you know we don't short positions

43:40

but it just seems to me that given this

43:42

rebalancing in spend that it's unlikely

43:45

that you're going to see that uh level

43:47

of housing activity given where uh

43:49

prices are at.

43:51

>> You're preaching to the choir. I'm a I

43:52

am a home h housing market volume bear

43:56

as well. I just don't I don't see how we

43:59

>> how we get back to those to those levels

44:01

either. Um, as far as the the consumer

44:04

trends that you were talking about,

44:07

>> what are the types of companies that you

44:09

think are going to um that are going to

44:13

take share? You said you don't think

44:14

it's going to be category.

44:16

>> I would not make a bet on a

44:19

discretionary category getting back to

44:21

former glory. That being said, we do

44:23

invest in some uh consumer discretionary

44:26

names. One of them uh is Doc Martens.

44:29

We've invested in quite some time. It's

44:31

been roughly flat. They've been growing

44:33

revenues mainly through pricing volumes

44:35

are slightly down. Uh

44:38

but our conviction in the uh in their

44:42

brand equity is so high that even if the

44:45

category is quite down, we think that

44:47

they're going to continue to maintain

44:48

share and grow when the category

44:51

rebounds uh and it's still the category

44:54

the boots category is still at a

44:55

cyclical low. uh and we do not expect

44:58

the rebound to get back to any former

45:00

levels at all for us for the uh for the

45:03

thesis uh to make sense. Um so I think

45:08

that if you're going to invest in a

45:09

consumer discretion type category, you

45:10

have to be really really confident in uh

45:12

their brand equity uh and confident in

45:14

their ability to take share in that

45:16

category because it's really a a share

45:19

taking type story. I I would not expect

45:23

uh the same level of discretionary spend

45:25

in certain categories that we that we've

45:26

enjoyed in the past.

45:28

>> H how does one maintain and get that

45:31

conviction? I mean, you look at a

45:32

Lululemon, something that a few years

45:35

ago people had tremendous conviction in

45:37

the brand equity and uh you know, maybe

45:40

I'm biased here in in Fashion Forward

45:42

New York, but the uh the good-looking

45:44

girls are wearing Alo Yoga, not

45:48

Lululemon. It's something your mom is

45:49

wearing. Um I'm just, you know, it it's

45:53

it can happen extremely quickly. what

45:56

looks like strong grand equity um you

45:59

know flips like how do you I mean I I

46:02

guess look Doc Martens has been around a

46:04

long time.

46:05

>> Yes. Exactly. Yeah. So Doc Martens

46:08

really been around since 60s. Uh this is

46:11

a company that the majority of its sales

46:14

are from a a timeless type of product

46:17

their uh 1460 boot. And it's if you look

46:21

stylistically if you look at it it's

46:23

kind of remained it hasn't changed all

46:25

that much throughout the years. So they

46:27

haven't had to make hu these huge

46:28

revamps and style or whatever to keep

46:30

the customer interested. Uh additionally

46:32

if you look at all age categories

46:34

there's still high level engagement all

46:36

age categories. So even the new you know

46:39

uh you're looking at new genz new

46:40

customers. It's always a problem for a

46:42

brand when it's like oh it's your

46:44

grandma's brand but not you know younger

46:46

generation. But there's something

46:48

special uh about the soul of the Doc

46:50

Martens brand, the kind of rebellious

46:52

spirit of it that refreshes uh every

46:55

generation, you know. Um so those

46:59

>> old punks and young punks,

47:01

>> it's Yeah, exactly. Yeah.

47:04

Um as far as uh Lululemon goes, again,

47:07

Lululemon would have been difficult for

47:08

us to invest in just purely on valuation

47:10

basis, but also it's a company that has

47:12

been around that long. So, any company

47:14

that has been any type of brand or

47:16

discretionary product that hasn't been

47:17

around that long, you always have to ask

47:19

yourself, well, is this kind of just a a

47:21

trend that's going to go away at some

47:23

point. Um, so clearly they were uh they

47:27

were kind of first to this kind of

47:28

athleisure

47:30

>> type of of a trend. And you've seen a

47:33

couple this this other competitor come

47:35

along and take a lot of share and and

47:37

some of it is because of the this the

47:38

fattish nation fattish uh nature of the

47:42

category. Um, so yeah, I think any any

47:45

type of brand that hasn't been around

47:46

that long, you got to be very very

47:47

careful ascribing a lot of brand equity

47:49

to it.

47:50

>> What about those brands that are on the

47:52

rise? I mean, some of the best

47:53

performing consumer discretionary names

47:55

are often those ones that are taking off

47:58

at the beginning of a fad.

48:00

>> That's typically something we stay away

48:02

from entirely. Uh, I mean, usually if

48:05

something is fat is fattish, uh, it's

48:08

going to be, uh, trading quite rich. Um

48:12

so and fash is something that is

48:15

notoriously unpredictable for us. We're

48:18

looking for something that's concrete,

48:19

something that uh we can have some sort

48:22

of predictive insight into and anything

48:25

that tends to be uh fad driven. Uh yeah,

48:29

that we we as much as we can we try to

48:31

stay away from those categories

48:33

>> with that limited discretionary spend.

48:35

Is it going to go towards goods or is it

48:38

going to go towards experiences? I

48:40

forget who I was talking to recently uh

48:42

who was saying that the younger

48:45

generation's

48:47

uh aptitude to spend more on experiences

48:49

is related or connected to this uh lack

48:53

of home ownership to some extent. It's

48:55

this idea that you know uh we're not

48:58

going to be owning a home. Uh there's a

49:00

kind of lack of stationerity and

49:04

that kind of lends itself to more spend

49:06

on experiences of goods or something. I

49:08

don't know if that's true. I have no

49:09

idea. Maybe it is. Um, but I'm not going

49:13

to say something either way about the

49:15

structural or cyclical cyclical nature

49:17

of uh, you know, goods versus uh,

49:20

experiences for uh, younger consumers.

49:23

You could make a big argument of social

49:25

media just the in social media influence

49:28

regarding uh, getting people excited

49:31

over experiences and travel. Uh, you

49:34

obviously that is going to persist. You

49:36

know, it's funny. I remember uh when

49:37

people this was uh pre- pandemic when uh

49:41

Are you a golfer at all?

49:43

>> Yeah.

49:44

>> Okay. Um

49:45

>> captain of the high school golf team.

49:47

>> Oh, nice. Okay, cool. Um so, do you

49:49

remember pre- pandemic when there seemed

49:52

to be a narrative that uh you know, golf

49:55

was kind of dying in a way because

49:56

people couldn't spend the time on a golf

49:59

course like attention spans were a lot

50:01

lower and people are just going to golf

50:03

a lot less. Um, and there was just a

50:05

kind of decline in the attention of the

50:06

game. Is that Do you remember hearing

50:08

about that?

50:09

>> Yeah. And I mean, it was partially

50:10

related to Tiger no longer being there.

50:12

I mean, you had this tremendous tailwind

50:14

of like a huge star who had propelled

50:15

the game, propelled everything forward.

50:18

And then there was a bit of an overbuild

50:20

too in the early 2000s related to the

50:22

tiger boom in golf. And so, you know,

50:24

that was a big sort of like cyclical

50:27

um sort of unwind there. But

50:30

>> yeah, I mean for me I I think that the

50:32

clear tailwind for golf has been the

50:35

influencer space.

50:36

>> Yes, exactly.

50:38

>> It's been tremendous and I've seen it

50:40

with with friends who I'm just like how

50:43

many hours a week are you spending

50:45

watching golf YouTube?

50:47

>> Um and you know to all of our listeners

50:49

who watch both thank you for allotting

50:50

some of your time to monetary matters

50:52

and to markets. But it's it's pretty

50:55

incredible how how much people are

50:57

watching golf much less than putting,

51:00

you know, five hours into playing it.

51:02

>> Uh it's it's unbelievable. I just

51:04

remember I would go to uh you know,

51:06

different driving ranges pre, you know,

51:08

precoid and they'd be empty. And now

51:11

it's like I can't find a spot. Um and

51:13

it's been like exactly you said a result

51:15

of this social kind of media trend which

51:18

which flew in the face of the narrative

51:20

which which seemed almost structural.

51:22

It's like, hey, well, people have a lot

51:23

less time. Um, they're just not going to

51:26

be golfing as much. Golf is a sport,

51:28

unfortunately. One, that just takes too

51:30

much time and attention. People just

51:32

don't have the bandwidth anymore. So,

51:33

there's a structural reason to for the

51:36

what we're seeing in golf. And then, you

51:38

know, fast forward a few years later and

51:41

because of this influencer phenomenon,

51:42

you know, the whole structure uh and uh

51:46

and popularity of the game has has

51:48

completely has completely changed

51:49

overnight. So obviously there's a huge

51:53

social part of this that makes it uh you

51:56

know very difficult to gain insight of

51:57

what what's happening or what will

51:58

happen. But yeah but going back to the

52:01

goods versus experiences we don't know

52:03

if this is a cycl sickle thing with

52:05

young people or we go back to goods. Uh

52:07

it seems to me that experiences are

52:08

going to be here to stay for a long time

52:10

but I don't know.

52:11

>> Well what's interesting about golf is I

52:13

I call them gear sports. They are sports

52:16

where part of it is the ability to get

52:18

the new toy, right? And it also ties

52:21

into the K-shaped economy thing of, you

52:25

know, whether it's the super nice road

52:27

bikes or mountain bikes, like there's a

52:29

new thing that you can buy and add on

52:32

that's going to change your game. And

52:35

you know there is an aspect of work and

52:37

putting in the time and the experience

52:38

side of it but people you know you get

52:40

to buy the outfit, you get to put on the

52:43

outfit, you get to put the outfit on

52:44

social media, you get to buy the

52:46

equipment that goes with it too. It's a

52:48

it's it's a good and experience category

52:52

and it's also

52:55

considered to be like more at the top of

52:58

the K. Um, so I think that it's it's got

53:02

a lot going for it in that. And you

53:04

have, you know, you have some like big

53:06

private equity deals like uh like the

53:08

Lab putters. Um, they got

53:10

>> bought out by private equity. You know,

53:12

certainly those guys are paying

53:14

attention to what is happening at um in

53:16

these golf markets.

53:18

>> That's insightful about Gear Sports. Uh

53:21

yeah, and maybe there's a social cache

53:23

element to it like you said as well. I I

53:25

don't know. Uh, it's also funny because

53:27

like my like I'm a big believer that uh

53:31

just with new generation of golf

53:32

equipment, it doesn't like it doesn't do

53:34

all that much to your golf game.

53:37

>> It's one of the secrets on tour. They're

53:39

all playing old. They all have a

53:41

favorite prov from like the

53:44

playing and it's just stamped with the

53:46

new stamp.

53:47

>> It's hard for me to relate because I'm

53:48

never going to be one of these guys that

53:49

goes out and gets tireless driver every

53:51

year. Uh because I just don't think it

53:53

matters. uh you know I I think that you

53:56

know if you look there's a lot of

53:57

YouTube videos on them testing all old

53:58

equipment versus new equipment and

54:00

there's you know a lot of times there's

54:01

very there's very little uh difference

54:04

uh but yeah no I think the gear that

54:06

that whole that whole uh ecosystem is

54:09

part of uh the allure for a lot of

54:11

people and what's what's grown the game

54:13

which is you know really interesting to

54:14

see

54:15

>> it's pretty hard to play golf here in

54:17

New York so I have switched more to

54:19

tennis but it has the same dynamics and

54:22

I also think not that, you know, women's

54:24

golf isn't also very popular, but like

54:27

women's tennis, they're the highest

54:29

earning female athletes in the world.

54:30

They have some of the biggest

54:32

endorsement deals and uh like we can

54:34

just, you know, come right out and say

54:36

like tennis outfits like look good. And

54:38

so there's like when we talk about the

54:40

social media, the signaling of what part

54:43

of the K and the K-shaped economy you're

54:45

in, like the attendance at the US Open

54:48

is at record levels. there are people

54:50

there who couldn't name

54:52

>> couldn't name players um and it's just

54:55

like a huge thing of like oh I am of

54:58

this class of person I live in the New

55:01

York City area I am supposed to go to

55:05

the US Open whether you're a man or a

55:07

woman we just also had the rider cup

55:10

which I think was a little bit I think

55:12

golf is still skews male in terms of

55:14

this trend but similarly it was just

55:16

like if you are a red-blooded male male

55:20

of this part of the K of the K-shaped

55:22

economy, like you should be going to the

55:24

RDER Cup with your boys. And if you're

55:26

not, then you're missing out. Um, and

55:29

you're, you know, it it's it's pretty

55:32

incredible, uh, what I've seen. It's

55:34

something that I like very much focus on

55:36

because it's affecting how much it costs

55:37

me to play tennis.

55:40

>> Yeah, humans are Yeah, that's bizarre

55:43

kind of behavior. Uh it's it's so

55:45

interesting because you would think that

55:47

uh like it's it's counterintuitive and

55:50

I'm also always interested in situations

55:52

where it's like oh that's interesting. I

55:55

wouldn't have thought that that would

55:56

develop that way like like for example

55:58

like you have the internet you have

55:59

social media and it can you know

56:01

whatever fits your persuasion you can go

56:03

find you can nurture your interest and

56:05

it seemed seems to me that that type of

56:08

ecosystem would lead to uh you know a

56:11

more heterogeneity in interest and it

56:14

seems like what you're describing is the

56:15

opposite you're getting a lot of like

56:16

clotting around uh just kind of these

56:19

social expectations uh which you know

56:22

just just bizarre I've never I'm I I

56:25

I'll never, you know, cease to be

56:28

surprised by uh just behavior of uh

56:31

consumers. Like it's always it always

56:33

surprises me.

56:34

>> Yeah. It's no longer the Christmas card

56:37

with you and your family and your new

56:39

starter home. It's we had 12 honeydeuces

56:43

and we got seats at Arthur Ash for the

56:46

big tennis match. Like that's that's the

56:48

new status symbol Christmas card.

56:50

>> Wow.

56:52

>> Yeah. There you go. Let's go to maybe

56:54

some of the more commercial trends that

56:56

you are wondering about and one of them

56:58

is is will commercial travel go back to

57:00

COVID levels really interesting uh I

57:03

think it was Delta just reported and

57:04

they just had an overall drop in

57:07

passengers but reported you know great

57:10

profits and revenues because of the

57:12

strength of business class. So I think

57:15

it ties into this commercial travel

57:16

question. It ties into these K-shaped

57:18

economy questions. How are you thinking

57:20

about air travel and um aerospace in

57:23

general in the portfolio whether it's

57:25

the defense side of things or commercial

57:27

air travel?

57:28

>> I think we ascribe to just the general

57:30

NBA view that those are very you know

57:32

capital intensive just structurally

57:34

difficult businesses uh very

57:36

commoditized. I mean if you look at

57:38

airlines they're one of the uh like

57:41

seats that has experienced some of the

57:44

lowest inflation since 2020. I think

57:46

cumulatively it's around 10% or so. Um

57:49

so you know costs have been up and

57:51

prices they haven't been able to offset

57:52

the prices uh just given the extreme

57:54

commoditization in the space. Um as far

57:57

as the the defense side of things is

57:59

totally different. Uh I think we're very

58:01

interested in uh in increasing our

58:03

defense exposure. It's just hard to find

58:05

the right name and a lot of these

58:07

companies have been bit up quite a bit.

58:10

Uh, as far as going back to the

58:11

commercial travel question, I think it

58:12

links back to

58:15

uh uh just commercial uh or commercial

58:20

real estate and are are people coming

58:22

back to the office uh in significant

58:25

ways? I know there's been a lot of

58:26

announcements uh with companies uh

58:29

bringing people back to the office. I

58:30

haven't seen the latest data on that. I

58:32

I mean, you probably know better than me

58:33

in New York. Uh it seems like there's

58:35

some sort of hybrid model that is

58:37

probably more uh that has gained

58:39

popularity where people come back three

58:41

days or four days uh in the office and

58:43

they have um you you know maybe they can

58:46

work from home on Friday. I I thought

58:49

that uh you know maybe before people

58:51

started coming back to the office a

58:52

little bit. I thought it was one of the

58:53

hardest trends to try to predict whether

58:56

this was kind of structural or people

58:57

are going to come back. Uh, I mean I it

58:59

seemed to me be likely that you know

59:01

rents in commercial buildings would

59:03

continue to come down and then it would

59:04

just be more attractive for people to

59:05

kind of oh like I'll just get an office

59:08

but that still wouldn't be like people

59:10

coming back to the office full-time that

59:12

it'd be them just having an office more

59:14

of a hybrid model. Uh but I I don't know

59:16

what what are you saying?

59:17

>> It's definitely hybrid. There is a lot

59:19

of return to office. I mean especially

59:21

now as people are it's one of the ways

59:23

that you can kind of silently downsize.

59:25

you can put out a big return to office

59:27

memorandum and you have all these people

59:29

that were hired and instead of having to

59:32

lay them off and and in a not that I

59:35

don't think New York has like a

59:36

requirement that you pay people

59:38

severance but you know it's pretty

59:39

common and so if you can say you're

59:41

returning to office and it's either quit

59:43

or you're fired for cause save a lot of

59:46

money on that and you don't have to

59:47

announce you don't have to do the big

59:49

layoff announcement which looks bad for

59:50

your business that you have to explain

59:52

so return to office is a way I think

59:55

also So to do silent layoffs. Um.

59:59

>> Yes.

59:59

>> But as far as the commercial travel

60:01

goes, like I just booked a trip for like

60:05

to go to the eye connections conference

60:07

in Miami, uh, the big allocator

60:10

conference. And, you know, we were

60:11

talking about it, Jack, my business

60:13

partner, and I, and we're like, yeah,

60:14

it'll be great to get this trip in.

60:16

Like, we don't really have to travel a

60:18

lot for work. You started out in this

60:20

business with, you know, Real Vision. We

60:23

would fly around the country to do

60:26

in-person shoots with two cameras and

60:29

there was a production crew and all

60:31

these sorts of things and you know the

60:33

podcast industry was was was

60:36

>> was around but it was much more nent and

60:39

you know you are seeing some return to

60:41

the in studio just as the the ability to

60:44

make content has people have the

60:46

equipment they've seen how the business

60:49

model works and so you are seeing a bit

60:51

more um sort of like upscaling of the

60:54

quality back towards that inperson

60:57

instudio type of model. Um but it's

61:01

still, you know, you have to have an

61:03

established brand to do that or or

61:04

you're just lighting money on fire um in

61:07

in a lot of ways. And there's a big

61:09

difference between traveling to do that

61:11

and setting up shoots on location and

61:15

having a studio and just saying, you

61:16

know, let us know the next time you're

61:18

in New York. We'll get it done. Um

61:20

>> but you know I think it's less traveling

61:23

to shake hands, show face, you know,

61:27

sign the deal and actually that

61:29

experiential travel of like oh you know

61:33

more conferences, more events like that

61:36

type of corporate travel I see as as

61:38

being like pretty strong. Um especially

61:41

because there is less in office. It's

61:43

it's a way to for these people.

61:46

>> You have a team that is dispersed across

61:49

the country. Uh you're going to all come

61:52

together for a conference and it's where

61:54

you're going to show face and build that

61:55

camaraderie. You know, seeing a lot more

61:57

like corporate off-site events where

62:00

you've got the the team dispersed across

62:02

different regions and you know they're

62:04

going to get together for a week and do

62:06

team building and stuff like that. So

62:08

playing into that experiential sort of

62:11

trend is is definitely what I see, but

62:14

all anecdot not anything hard.

62:17

>> Yeah. And that that's interesting like

62:19

so you're almost making a case for

62:20

here's my reasons why why commercial

62:23

travel is going to get back or even uh

62:25

exceed uh previous levels. And you know

62:29

I mean you have these two counteracting

62:31

forces. the other forces you described

62:32

where uh yeah there needs to be just

62:34

more of a cadence of people meeting in

62:36

person given the work from home

62:38

phenomenon but then you still have more

62:40

people from work from home and generally

62:41

when there's more people in an office

62:43

there's probably going to be more travel

62:45

uh so those two things counteract each

62:47

other and I don't know how it nets out

62:50

but it does seem reasonable to me that

62:52

uh

62:53

>> well I look at it as more like we're

62:54

sending the whole team like corporate

62:56

travel used to be we've got one guy

62:58

who's on the road and he's on the road

63:00

every week and is flying to a new place

63:02

to meet clients and we've got these like

63:04

teams that are kind of always on the

63:06

road. And so, how does that person who

63:10

used to take 50 flights a year is now

63:12

taking 10. How does that square with all

63:15

right, we're sending everybody to this

63:17

conference or to this event and we're

63:19

putting everyone on a plane, you know,

63:21

40 people. Is that enough to make up for

63:24

the loss from the road warrior? Yeah,

63:26

>> because I think that is lost. Like we

63:28

are doing

63:30

>> the the initial call, even signing the

63:32

deal, right? Um doesn't necessarily need

63:36

to be in person in the same way.

63:37

>> Well said. Yeah, it's it like how that

63:40

squares uh is is interesting how that'll

63:43

shake out. I I don't know uh which which

63:45

one of those forces will win out. And

63:47

going back to what you previously said

63:48

about companies, uh, yeah, like I'm not

63:51

sure if companies actually mean it when

63:54

they say come back to the office or like

63:55

you like you said, it's more of a tactic

63:58

uh for headcount reduction. Um, you

64:02

know, because it seems to me that

64:03

executives on down seem to be h, you

64:06

know, happy with flexibility. I'm

64:08

unsure. I know there's some productivity

64:10

hits in some areas and companies have

64:12

kind of had have spent a lot of time

64:14

thinking about how to maintain

64:15

productivity or productivity or even

64:18

enhance productivity uh in kind of like

64:20

this Zoom type work from home world. Um

64:24

so yeah, I think uh yeah, it's

64:28

interesting. I I wouldn't be surprised

64:29

if we just stay at these same levels for

64:31

a very long time.

64:32

>> Yeah, I I do think there's some of a

64:34

reset to the new normal. But the the the

64:36

other big question you have about white

64:37

collar headcount could very much affect

64:39

that if the difference between the

64:40

person who stays around and doesn't is

64:42

they're the one who lives in town and

64:44

and comes into the office and has that

64:46

facetime versus the person who is remote

64:49

regardless of whether it's um their

64:51

individual productivity is is better or

64:54

worse. It's just we don't need as many

64:57

people because more stuff is being done

64:59

by AI. Um, is that going to mean, you

65:02

know, that that those people in office

65:04

are going to to come out on top? You

65:07

know, these are these are all these are

65:08

all the big questions that I think we

65:10

have. I'd like to close with a big

65:13

answer, right? Is there anything you

65:16

know in in the questions you asked that

65:18

you have the most confidence in looking

65:21

forward that that really is already

65:23

taken up position in your portfolio and

65:26

and you barring a major change you see

65:28

that staying the case for 2026.

65:31

>> I think the allocation to cyber security

65:33

uh is going to meaningfully increase for

65:36

businesses. Uh if you look at uh this is

65:39

a good stat um since about 2023 the

65:43

number the amount of cyber security

65:45

attacks of businesses has been going up

65:46

by something like 30% a year. So because

65:50

you know bad actors are using AI in

65:52

different ways to kind of attack uh

65:53

companies. So the amount of spend that's

65:55

being allocated

65:57

by companies the amount of energy and

65:59

focus that has to be put into cyber

66:01

security is is going to continue to take

66:03

more share. uh you know there's like

66:06

large companies uh you know crowd

66:08

strikes paloto networks of the world

66:10

that will continue to benefit I think

66:11

from those trends they're already very

66:13

very large uh I think PaloAlto Network

66:16

is something likeund 100 something

66:17

billion dollar company um and what I

66:20

think is interesting is how how those

66:21

big companies compete uh and cyber

66:24

security is one of the hardest it's one

66:27

of the hardest companies to run well if

66:28

like running cyber security given the

66:30

threats are always changing so these big

66:33

cyber security forms

66:34

have adopted almost like this big pharma

66:37

model where you know as soon as this new

66:40

cyber kind of threat comes out they look

66:41

for uh a small company that's been

66:44

founded and funded uh to order fight

66:46

like you know uh maybe API security is

66:50

now a big issue a company that's already

66:52

doing that they go and acquire that

66:53

company and it's really like this serial

66:56

kind of acquire model uh this how these

66:58

kind of big big companies are able to

67:00

kind of uh maintain agility in a very

67:03

very dynamic uh cyber security world. We

67:06

would like to increase uh or to uh to

67:11

add some cyber security portfolio. It's

67:12

just a notoriously uh difficult industry

67:15

given just the levels of dynamism and

67:17

change.

67:18

>> Are you looking for those smaller

67:20

companies that are going to be acquired

67:22

or is it the big companies but they're

67:24

just not at the right price for you?

67:27

>> The big companies just aren't at the

67:28

right price point. uh we we're not going

67:31

to invest in I I don't know what uh like

67:33

some of these big companies trading at.

67:35

I I won't be surprised. They're north of

67:36

10 times sales. I think I think uh

67:39

Cloudflare is still north of I think

67:41

that's around 20. I I think it they're

67:43

trading at obscene valuations in our

67:44

opinion. Um so yeah, that we can't touch

67:48

them. But yeah, smaller companies

67:51

um potentially they're trading something

67:53

reasonable is definitely something we're

67:54

going to be evaluating.

67:55

>> Well, DA, we'll leave it right there.

67:57

Everybody can find your work at

67:58

parisresearch.com. You're also on

68:00

Twitter. Are there any other places

68:02

people can find you?

68:03

>> Mainly uh Twitter pronouncresearch.com

68:07

where we publish our research.

68:08

>> Well, thank you so much. We'll do it

68:09

again soon.

Interactive Summary

Paris Research, a high-performing independent equity research provider, discussed their investment strategies focusing on identifying "AI winners" and "mislabeled AI losers" within structurally changing markets. They critically view the current SaaS complex as overvalued, despite recent corrections, noting that disruption risk from AI is not fully priced in. They favor vertical and enterprise SaaS companies that integrate AI, providing Samrush as an example of a successful "mislabeled AI loser" investment. The discussion also covered broader structural shifts, including changing consumer behavior influenced by AI, a rebalancing of consumer spend where essential categories outpace wage growth, and the evolving landscape of advertising (less SEO, more social media). They also shared their "motor investing" philosophy, emphasizing dynamic belief systems and looking for companies with building potential energy. A key conviction is Remittly, a cross-border remittance provider, which benefits from global migration and digital payment trends, and they dismiss stablecoins as a significant disruptor. For 2026, they anticipate a meaningful increase in cybersecurity spending due to AI-driven attacks, seeking smaller, reasonably priced companies in this sector.

Suggested questions

10 ready-made prompts

Recently Distilled

Videos recently processed by our community