HomeVideos

How I Built a 1.4-Billion-Dollar Quant Fund - Deepak Gurnani on Founding Versor Investments

Now Playing

How I Built a 1.4-Billion-Dollar Quant Fund - Deepak Gurnani on Founding Versor Investments

Transcript

1146 segments

0:00

Deepo, thank you so much for doing this.

0:04

Welcome to Colombia.

0:05

>> Thank you. Thank you for inviting me.

0:08

>> You founded the hedge fund division of

0:11

Invest Court. Was there any one moment

0:15

while you were doing that when you said

0:17

I want to build my own quant hedge fund?

0:20

>> Yeah. So maybe just if I may give you a

0:22

little bit of s background, right? So

0:24

the when I first started I I was at

0:27

Invesco for 20 years. Uh when I started

0:31

working at Invest uh initially I started

0:34

as an analyst supporting an internal

0:37

quantitative hedge fund effort and uh

0:42

at that stage I sort of got introduced

0:44

to hedge funds and got very interested.

0:46

So interested on the cognitive side was

0:48

sort of natural given the education and

0:50

my interest and uh uh I would say pretty

0:53

much throughout the 20 years stay uh I

0:57

had a fair amount of flexibility

0:59

uh in being able to work on different

1:02

kinds of strategies. I think it was

1:03

towards the end of that tenure uh where

1:07

I felt that uh uh because I had broader

1:10

responsibilities I was not spending uh

1:13

as much time on specific qualitative

1:15

strategies. That's where sort of the

1:17

interest to get back exclusively to

1:20

quantitative strategy sort of started. I

1:22

would say towards the last I would say a

1:24

year or two is where it was but pretty

1:26

much throughout the period uh I focused

1:28

on amongst other thing quantitative

1:31

research and strategies.

1:32

>> Can you talk a little bit about your

1:34

background in you know before founding

1:36

the hedge fund division of of invest you

1:39

know education I guess what it was like

1:41

growing up because you grew up in India

1:42

right? So I was I was fortunate from an

1:44

education standpoint. Uh I attended some

1:47

of the prestigious schools. So there's a

1:48

prestigious Indian Institute of

1:50

Technology. I attended that. Uh and

1:52

there's a Indian Institute of Management

1:54

which is again a a good school. And uh I

1:57

was sort of fortunate to have attended

1:58

both those schools. Uh I think it became

2:00

very clear I was very interested in

2:02

mathematics and programming and uh

2:05

somehow uh I think it was inevitable

2:09

that I would let end up in a

2:10

quantitative field uh not necessarily

2:13

hedge funds. I didn't know hedge funds

2:15

uh in the 80s when I was when I was

2:17

studying. So I worked initially with the

2:20

city group for five five and a half

2:22

years and

2:23

>> that was in India or

2:24

>> in Europe mostly in Europe not in the US

2:27

not neither in India. uh I started in

2:29

India and then I worked in Europe uh

2:30

various locations in Europe for about

2:32

five five and a half years. Uh that's

2:35

when I got introduced to Invest Corp and

2:38

uh Invesc ran uh some internal hedge

2:41

funds quantitative which is what really

2:43

attracted me to invest corp. Uh I guess

2:46

I was at the right place at the right

2:48

time. I started as an analyst and in

2:51

about five years or so I was the co-head

2:54

and co-CIO of a new hedge fund effort at

2:57

Invest Corp and that's how sort of I got

3:00

I took on a leadership role and then

3:03

developed the business along with other

3:05

partners as well in 2008

3:07

>> I became the sole head and the sole CIO

3:11

>> group I mean five years from analyst to

3:13

head of the division is pretty

3:14

>> I call it right place at the right time

3:16

>> pretty crazy the other thing And this is

3:19

something Dwayne mentioned to me and I

3:21

really wanted to ask you about this cuz

3:23

if I'm correct the hedge fund division

3:25

of Invest Corp was founded from um out

3:29

of the sale of of Gucci. Am I correct?

3:32

>> So the hedge fund there was a hedge fund

3:35

business when I first joined. Yeah. It

3:36

went through changes, right? It went

3:38

through changes and uh uh one of the

3:42

notable changes was the sale of Gujits

3:44

in the public, right? I'm not saying

3:46

anything that is not in public domain.

3:48

Uh and uh yes the the hedge fund effort

3:52

that I was charged with as a co-head

3:54

cocio sort of started around the same

3:56

time.

3:56

>> Okay.

3:57

>> In '9697.

3:59

>> Okay. Okay. And what was it like you

4:01

know so in those five years what did you

4:03

do? I mean you say right place, right

4:05

time and I guess you know there's an

4:07

audience of very driven young people who

4:09

are clearly interested in in in the

4:11

markets and in building um you know

4:14

asset management businesses or at least

4:16

working in the space. Um how did you in

4:18

those five years progress from analyst

4:21

to eventually co-heading the division?

4:24

So I think uh the way that sort of

4:27

effort started was the previous effort

4:30

uh which was already in progress was a

4:32

more mature effort when I first joined

4:34

as an analyst.

4:36

>> That effort was wound down and uh a

4:39

group of folks were identified uh to

4:42

head the new effort that started in

4:45

9697. Okay.

4:47

>> Uh that effort uh was being headed by

4:50

the chief financial officer. Okay. at

4:53

the firm. Uh I was working very closely

4:55

with the chief financial officer at that

4:57

time and with the treasury and so I was

5:01

one of the two or three senior people

5:03

who were sort of put on that. Uh I think

5:05

I just with hindsight it's been a while

5:08

back right? Uh I think I did have uh

5:11

very strong quantitative skills and uh I

5:15

think that sort of complemented the

5:16

other skill set and I think that was one

5:18

of the reasons uh where sort of that

5:21

opportunity came up

5:22

>> uh and then you have to execute on the

5:24

opportunity I guess right and uh we

5:26

executed as a group we executed it that

5:28

well. Yeah.

5:29

>> And uh that really helped progress the

5:31

effort.

5:31

>> And so what year was that? What year did

5:33

you start as an analyst there?

5:34

>> 93.

5:35

>> 93. And so in 98 you, you know, started

5:39

heading the the division.

5:40

>> Uh no, I was the co-head and co-CIO from

5:43

9798 onwards.

5:45

>> Uh I became a partner in 2000.

5:48

>> Okay.

5:48

>> Uh and in 2008, I became the sole head

5:51

and the

5:52

>> uh the CIO of that division. And so what

5:56

did the quant world look like back then

5:58

versus the 2000s, 2010s and then

6:01

obviously now today

6:02

>> the the quant effort uh has s obviously

6:07

sort of it has evolved right uh it has

6:10

evolved along many dimensions. I think

6:13

the amount of data

6:16

uh that is available today is

6:19

significantly more than the kind of data

6:22

that was available in the earlier days.

6:24

Right? So uh uh the data the statistical

6:29

methods that were used uh are

6:32

significantly advanced and also I think

6:36

just the sheer amount of research

6:39

uh that has been done in the field uh

6:42

you know I and my colleagues have done a

6:43

significant amount as well but clearly

6:45

the industry has developed industry in

6:48

those days was the hedge fund industry

6:50

was much smaller uh focused more on high

6:54

net worth individ individuals uh and

6:56

then today obviously it's last numbers I

6:59

read it's a $5 trillion industry so a

7:01

lot of that effort the research effort

7:04

has also sort of increased in that area

7:07

so I would say clearly on all of the

7:09

dimensions whether it's the amount of

7:10

data uh the kind of data that is

7:13

available uh the kind of research uh the

7:17

understanding of hedge fund strategies

7:19

today is significantly different from

7:22

what it was uh when I first started Now

7:25

uh that itself was an opportunity in

7:27

those days right so in the early days

7:29

because the strategies were not that

7:31

well understood.

7:32

>> So using a quantitative approach to

7:35

implementing hedge fund strategies

7:38

>> uh gave you an edge and I I guess that

7:41

sort of helped me and contributed to my

7:44

personal sort of professional growth at

7:46

the firm. Now what you mentioned about

7:48

the industry I mean it's still very

7:49

opaque but if you look back then and

7:52

compare it to today it's night and day

7:54

like I recently I was chatting to a guy

7:57

on Saturday and he was talking you know

7:59

I asked the question he said that's a

8:00

great question and it's funny the only

8:02

reason I'm able to ask that question is

8:04

because there's so much information out

8:06

there versus I can't imagine you know

8:08

90s even early 2000s I mean I think not

8:12

much was known and I guess I'd like to

8:14

hear about what sort of stuff you guys

8:17

at invest corp were doing um you know

8:19

late 90s early 2000s what sort of

8:21

strategies were you guys running

8:23

>> so I think the big change uh from a

8:25

quantitative research perspective I can

8:27

sort of focus on that happened after

8:29

2000

8:30

>> okay

8:31

>> right uh I would say the first few years

8:35

uh or the years prior to that was more

8:37

uh from my p personal perspective as

8:39

well as of the group that I worked with

8:41

was getting a deeper understanding yeah

8:44

because it was a relatively new

8:45

industry. Uh we were relatively new to

8:47

that as a group.

8:49

>> Yeah.

8:49

>> And uh so the first few years was spent

8:51

in developing a much deeper

8:54

understanding

8:55

>> of the of the industry. Uh I would say

8:58

2000 was a turning point where we did

9:03

see significant amount of institutional

9:06

interest uh into the hedge fund

9:09

industry. uh so pension funds globally,

9:12

sovereign wealth funds started looking

9:14

at hedge funds more closely. Uh I would

9:16

say starting from 2000 onwards and uh

9:19

that sort of encouraged me and my and

9:22

the team that I worked with to do more

9:25

quantitative research into different

9:28

hedge fund strategies because there was

9:30

very little uh research out there. uh it

9:34

is very opaque uh in those days

9:37

certainly much more than what it is uh

9:38

today and uh investors uh who were used

9:43

to quantitative research in uh long only

9:47

equities or fixed income investing were

9:50

looking for something similar on the

9:52

hedge fund side

9:53

>> and that really spurred a lot of

9:55

interest on our side to do that research

9:58

share that with potential investors and

10:00

I think did contribute to the success of

10:02

that division. over time.

10:05

>> So you're running you're running a b a

10:06

bunch of strategies at at um at invest

10:09

corp. I had coffee with Nishant on

10:12

Monday and he said you know you're also

10:16

heavily involved in the fund of fund

10:18

business. Um and I guess I want to

10:20

return to the first question I asked at

10:22

the start of this conversation about

10:25

spinning out right and and and building

10:28

Verser what is today a $ 1.4 billion

10:31

quant fund. Um, I guess take me to

10:35

that exact moment and what was it like

10:38

going from, you know, obviously very

10:40

successful heading the hedge fund

10:42

division uh of Invest Corp to to taking

10:45

that risk and building your own thing

10:47

obviously with founding partners but

10:49

still very much betting on yourself. The

10:51

motivation at that stage was to uh set

10:55

up on my own more to focus on

10:58

quantitative strategies only

11:01

>> and uh focus on a few niche quantitative

11:05

strategies. So that's what we do and

11:07

sort of that enables me to get into the

11:12

day-to-day working of the strategies.

11:14

Right? So I was sort of more interested

11:16

in in doing that getting back to actual

11:20

hands-on day-to-day investing.

11:22

>> Yeah.

11:23

>> As compared to necessarily managing a

11:25

larger portfolio uh with other PMs and

11:29

other strategies within that.

11:30

[clears throat] Right. So I think that

11:31

was the motivation. Uh I think I clearly

11:34

uh because I started that uh the

11:37

business here uh after a fairly long

11:40

career. I was sort of more realistic in

11:43

what I was getting into. So I knew uh

11:46

what are the risks that were being taken

11:48

uh and uh uh what I was getting into

11:51

right nothing sort of totally prepares

11:53

you for it until you actually go through

11:55

the process. But uh uh I think I had a

11:58

fairly good idea right of what it would

12:01

entail and so on. And uh I think is

12:05

really a focus on going back to

12:08

quantitative strategies only which is

12:11

really the reason why I had joined

12:12

incot.

12:13

>> Yeah.

12:13

>> I think that was the main motivator and

12:16

I looked at it and said that uh

12:18

>> we can set up a successful business with

12:21

like-minded partners. We could just

12:23

continue to do that for the rest of our

12:25

career. Yeah,

12:27

whenever I speak to someone who, you

12:29

know, worked at a big company and then

12:32

started their own fund, they always talk

12:34

about how aggregating all the different

12:38

moving parts together and, you know,

12:40

because you're not just running money,

12:41

you're building a business around it.

12:43

Um, they talk about how challenging that

12:44

was. And I guess, you know, obviously

12:48

you were very experienced before doing

12:50

that, but can you can you take me

12:52

through that? You know, what was it

12:53

like?

12:54

>> So, I think uh one of the activities

12:57

that I also did at Invest Corp uh when I

13:01

uh amongst several responsibilities

13:04

right being the head of the group was

13:05

seeding new hedge funds

13:07

>> so uh where you know we'd allocate

13:10

capital to uh upcoming new hedge funds

13:14

uh and uh uh so I had some experience

13:17

also seeing right how those uh evolved

13:21

and uh so I think the first uh thing I

13:25

would suggest setting up any business,

13:26

but certainly when you're setting up a

13:28

I'll speak more from the perspective of

13:31

a hedge fund, but I think it's true for

13:32

any business.

13:33

>> Yeah,

13:34

>> it is a marathon. It's not a sprint,

13:36

right? So, I think you have to view it

13:37

that way. So, you have to have a steady

13:41

uh speed, right? You can't be too slow,

13:42

right? Otherwise, you'll never get to

13:44

the finish line, but you don't want to

13:46

be so fast in a as in a sprint because

13:49

uh you'll sort of fail, right? Or or

13:51

complete much before the finish line as

13:53

well, right? So I think that's the first

13:55

thing. Uh the second thing is uh I think

13:58

very important is to know the value

14:01

proposition. So one of my favorite

14:03

questions every time uh I would uh this

14:07

is back in my days at invest when I was

14:09

seeding a hedge fund. Every time

14:11

somebody came in and said I want to open

14:12

up a new hedge fund give me some

14:14

capital. The first question sort of that

14:17

I would ask is uh what is your value

14:20

proposition? Because the last thing the

14:22

world needs is yet another hedge fund.

14:25

>> There are, you know, thousands, tens of

14:28

thousands of hedge funds,

14:31

hundreds of them coming up, uh,

14:33

thousands of them closing down. Uh, so

14:35

having a clear value proposition as to

14:38

what do you what is the differentiator

14:41

that you bring uh to the clients I think

14:44

is very important.

14:45

>> Yeah. Without that I would suggest to

14:48

people that do not even go down that

14:49

path because you know you will not be

14:51

able to differentiate yourself. Then

14:54

whatever amount of time and capital that

14:57

you think it will take to succeed it

15:01

takes at least twice as long and it

15:04

costs you at least two to three times

15:06

what you initially thought even with the

15:08

best made plans. So you have to be as a

15:12

founder or a set of founders, you have

15:14

to be ready to grind it out and uh if

15:17

you put too strict a timeline uh or you

15:20

have too short a runway budget from a

15:22

budget point of view, you're almost

15:24

destined to fail.

15:26

>> So I think that sort of has been uh the

15:28

conclusion that it always takes longer.

15:31

Uh you have to uh have a clear value

15:34

proposition. And I I think the other

15:35

thing I should mention uh which has been

15:37

constant throughout my tenure in hedge

15:39

funds is that the industry is constantly

15:42

evolving

15:44

and a lot of times when people say oh

15:46

you know the industry is at a at a it's

15:50

due for a major change. data set every

15:52

few years it has always changed

15:55

>> and uh so uh even the value proposition

15:59

that we initially started with you have

16:01

to evolve that you have to be relevant

16:04

in the environment that you are in and

16:07

adjust you can't constantly change but

16:09

you have to evolve in a very thoughtful

16:11

manner so I think those are sort of the

16:13

three or four points that I mentioned

16:15

are really the challenges I think it's

16:18

true for any business but certainly for

16:20

hedge funds that is based on our

16:22

experience at War.

16:24

>> Oh, would you talk about with regards to

16:26

evolution like absolutely and I um you

16:31

know whenever you read articles about

16:32

the pod shops and the war for talent and

16:34

all these different things um I think

16:37

that's just a natural that just follows

16:40

from these firms trying to ruthlessly

16:43

evolve and and I guess I guess I'd like

16:46

to hear about how you try to do that at

16:49

Verser. Um, obviously

16:53

you're running all these different

16:54

strategies,

16:56

doing doing well. Um, how do you as the

17:00

as the head of the firm wake up every

17:03

day and and and and have I guess a a

17:05

process for for analyzing the business

17:07

and thinking, okay, we got to do this,

17:09

we got to do that. What does that look

17:11

like?

17:12

>> Yeah. So I think the the way I sort of

17:14

look at splitting the effort is uh a

17:18

part of the day does go into structured

17:21

research meetings right so we have

17:22

products you have to look at the

17:24

performance uh we have to look at uh new

17:27

ideas that are coming up implement them

17:29

right so that's one part the second part

17:31

is you have to generate new research

17:33

ideas right so new research ideas where

17:35

you I have to use my experience uh look

17:38

at what's out there in the research and

17:40

I think certainly from a quant one of

17:42

the under appreciated aspects is looking

17:45

at the market. So I do spend a fair

17:47

amount of time studying the markets and

17:50

uh incorporating that into new ideas.

17:52

And then the third aspect is also

17:56

looking at ways how to evolve the firm

17:59

and evolve the business. Right? And and

18:01

it's not like

18:02

>> you sit every day and say look I'm going

18:04

to spend an hour thinking about what

18:06

right? So I can illustrate it with some

18:08

examples. So in when we started the firm

18:11

in uh 2013

18:14

uh around that time uh I actually got

18:16

introduced to uh cloud computing uh

18:21

actually by my son Nishant who you know

18:23

right so he was a undergrad at Princeton

18:25

in those days and uh we were starting

18:27

the firm and he introduced me to cloud

18:30

computing

18:31

>> and uh it made a lot of sense and I said

18:34

for a firm a small midsize firm starting

18:37

out do We have an edge in maintaining a

18:40

data center and maintaining

18:42

contingencies if the data center goes

18:44

down being a quad we need 24/7 right

18:47

literally availability of systems and

18:50

when I met in those days it was only

18:53

Amazon web services and we met them and

18:56

we liked it and we signed them up pretty

18:58

very early in fact to the extent that

19:01

very recent uh a few years back I should

19:03

say around covid time they did a case

19:06

study Amazon web services did a case

19:07

study on Of course we have a relatively

19:09

small firm as we were one of the first

19:10

firms right so that was again you can

19:12

call it lucky you can call it strategy

19:14

right but we sort of right place right

19:16

time I always say that right so we did

19:18

that I think the second evolution which

19:20

has influenced warser is use of

19:23

alternative data

19:25

I was invited to attend a alternative

19:28

data conference back in 2017 here in New

19:31

York City uh I did not have much

19:33

exposure to alternative data before that

19:36

I attended that event it was I think one

19:38

or two days and uh again it was an eye

19:41

openener it made a lot of sense right so

19:43

just to explain uh we do primarily

19:47

equities right different forms of

19:48

equities so in equities uh you have uh

19:51

fundamental data balance sheets income

19:52

statements right so you call it

19:54

fundamental data uh there is price data

19:57

right equity prices move during the day

19:59

you have tickmatic data volume order

20:01

book so that's technical alternate data

20:04

is everything else so a lot of it is

20:06

unstructured data like things like news

20:08

uh it's uh things like when there are

20:11

earnings calls uh those there are

20:13

transcripts written of those earning

20:15

calls you can apply uh AI methods to

20:18

analyze the calls and transfer convert

20:21

them into quantitative scores credit

20:23

card data satellite images uh weather

20:26

patterns uh and that has exploded now

20:29

right so in 2017 it was still relatively

20:32

new and uh I you know as it happened I

20:36

attended that conference it was a

20:37

two-day event I met with a few vendors

20:40

of course now the number of vendors has

20:41

also exploded there are at least 10 such

20:44

conferences that happen in a year in New

20:46

York City alone and uh so that's when we

20:50

started working on alternative data and

20:53

realized that alternative data is going

20:55

to be an edge for analyzing equity

20:58

prices uh or equity models going forward

21:01

and so we did that and uh I would say uh

21:04

the third thing which went along along

21:06

with alternative data was that the

21:08

traditional statistical methods uh would

21:11

not work as well. So we started using AI

21:15

and machine learning methods pretty much

21:17

in the inception of the firm but it went

21:19

together with using alternative data and

21:21

because we were using cloud services we

21:23

could scale up the infrastructure

21:26

>> uh in a in a very flexible manner where

21:28

we could use extra services uh servers

21:32

and computing power when we needed it

21:34

and shut it down when we didn't need it

21:36

right and that flexibility so a

21:38

combination of that I think was again a

21:41

part of our evolution so I would over

21:43

the last 5 to 6 years. So I think it's

21:45

just being open and receptive to ideas,

21:49

being sort of sensitive to what's going

21:51

on in the industry and being willing to

21:55

make the jump, right? In a in a

21:57

thoughtful manner, right? And uh making

22:00

an assessment of what we think will work

22:02

in the future, right? You don't always

22:04

get it right, but I think we've been

22:06

somewhat fortunate. I think it's part

22:08

fortunate, part strategy that we have

22:10

been at the front of some of these

22:12

evolutions in the last several years.

22:15

>> I mean what you say about

22:18

being early to cloud computing I think

22:20

that's 2013 thereabouts 2013 and then

22:24

that providing you the infrastructure to

22:28

you know build out you know an an

22:30

amazing research pipeline for applying

22:32

all these alternative you know applying

22:34

all these different sources of

22:35

alternative data. Um, that sounds like

22:39

either luck or genius or both. Um, and I

22:43

guess you talked a little bit about it

22:45

there where you I mean you said

22:48

applying these things and being very

22:50

thoughtful about them.

22:54

Where do you see the next frontier like

22:56

where where are your eyes right now?

22:58

Where are you looking I guess um for for

23:00

Verser's next evolution so to speak? So

23:03

I think it is uh it is certainly

23:06

expanding

23:08

use of application of AI and machine

23:12

learning methods to investing in

23:14

equities. Uh and uh that's really where

23:18

our focus uh continues to be. But when

23:21

we say equities uh we do different

23:23

variations, right? So we do single

23:25

stocks, right? We do what is referred to

23:27

as statistical arbitrage strategies. We

23:29

do that. We also do event driven.

23:31

>> Yeah. uh where uh these are companies

23:34

that go through mergers or being spun

23:36

off. Uh we do those as a part of our

23:38

strategies and we also do equity index

23:40

futures. Yeah. Uh as a part we also do

23:42

other futures but predominantly equity

23:44

index futures. So uh I think our big

23:47

focus right now is on expanding the use

23:50

of alternative data uh expanding the use

23:53

of AI and machine learning methods. uh

23:56

to that. I think that continues to be a

23:58

big focus area for us and especially I

24:01

think uh although we started work on

24:04

this much earlier uh with uh uh the

24:07

advent of chat GPT and the increased

24:09

focus on AI uh there's a tremendous

24:12

amount of research and tools that are

24:14

being available and uh that gives us the

24:18

opportunity to use those tools and uh be

24:21

able to further refine our products. I

24:25

mean you talk there about you know using

24:27

all the best tools refining your

24:28

products um that entails a lot of

24:30

research and you know I was talked to

24:34

Nishan Dwayne and also another partner

24:36

adversary Yash and he was you know they

24:38

all talk about how your team is you know

24:42

partially based in New York partially in

24:44

Mumbai um offices in both locations.

24:48

I guess my question is and and I asked

24:51

this to every boutique hedge fund

24:54

manager is you have all the big players

24:57

call it Citadel, Mill, Millennium, Jane

25:00

Street um and they're all able to they

25:02

all they have, you know, they they have

25:04

tons of money, right? They can hire the

25:06

best talent. Um I mean you look on the

25:10

news and you see the salaries of you

25:12

know internet Jane Street or or any one

25:15

of these other firms and they're getting

25:16

higher and higher every year. I guess my

25:19

question for you adverser is

25:24

how do you maintain uh you know how do

25:27

you attract the best talent to work for

25:29

you? Um how do you compete in in that

25:32

domain? Because I imagine it's it's it's

25:34

it's not easy.

25:36

>> Yeah. I think there's a the the implicit

25:39

question right yeah sort of there is

25:42

>> this debate on large funds versus small

25:46

funds

25:46

>> has been there for a very long time

25:50

>> right and uh so I I sort of uh address

25:54

that firstly right in a couple of

25:56

different ways right I think uh if you

25:58

look at innovation in other industries

26:01

right let alone hedge fund if you just

26:03

look at innovations in technology ology

26:06

or you look at innovation in uh in

26:10

medical medicine field, it's typically

26:13

the smaller firms

26:15

>> are the ones that innovate,

26:17

>> come out with either new technologies or

26:19

new drugs uh biotech firms and they then

26:23

get acquired by some of the larger

26:25

pharma firms or some of the larger

26:27

technology firms and uh uh the the

26:31

larger firms then sort of incorporate

26:33

that. Right? So there again I think we

26:36

consistently see and I'll address why

26:38

right but we consistently see that

26:40

smaller firms uh right so if you are if

26:44

you are a open AI right now they're much

26:46

bigger but when open AI first started

26:48

work and uh they had a partnership with

26:50

Microsoft again I have no idea why

26:53

Microsoft

26:55

decided to back open AI rather than I

26:58

think open AAI had what a thousand or

27:00

2,000 engineers at that point in time

27:03

Microsoft oft had over 150 200,000 I

27:06

think I've got the order of magnitude

27:07

correct even if the exact numbers are

27:09

wrong and uh Microsoft I I don't know

27:12

why but didn't say I'll allocate 5,000

27:14

engineers we'll develop our own chat GPT

27:17

equivalent right they didn't do that

27:19

right they said we going to work with

27:20

open AI which was a much smaller firm at

27:22

that point in time right

27:24

>> and uh and then you know so uh this

27:28

advancements were done and there are

27:29

other firms that have done that also

27:30

right so I'm saying this debate about

27:32

large firms firms having more scale uh

27:35

more technology more resources versus

27:38

small firms has been there forever right

27:40

but still small firms so why does that

27:42

happen right I think it really boils

27:45

down to uh motivation and alignment

27:48

right so uh again it's not true for all

27:52

large firms but it's it is true that the

27:54

larger the firm the larger the group

27:56

right uh there are more institutional

27:59

challenges to make a change within that

28:01

firm right there is so the word

28:03

bureaucracy does get used uh and once

28:06

you are a firm that has successful

28:09

products uh you tend to be more careful

28:11

and say uh why do I want to deviate from

28:14

something that has worked well and uh so

28:16

I think the nature of resources that

28:19

work for smaller firms uh tend to be

28:22

somewhat different from folks that who

28:25

work for the larger firms and I think

28:28

the challenge for smaller firms such as

28:29

Verser is to get the alignment correct

28:33

Right. So the environment that people

28:35

work in right so if we if folks who work

28:38

at vers when they work on a investment

28:42

problem uh they can actually see the

28:45

implementation from the

28:47

conceptualization of the idea to the

28:50

actual implementation and results.

28:53

As compared to that, if you are working

28:55

in a larger firm and you're working on

28:57

alternative data, for example, uh you

29:00

might be one out of 10 people working on

29:04

a form of credit card data that is being

29:07

used and uh you may have very little

29:10

visibility on the work that you do. Does

29:13

it eventually make it into the portfolio

29:15

or not and how does it work and so on.

29:18

So there's a siloed because it's a much

29:20

larger firm. There's a much larger

29:21

setup. So I think uh secondly as a firm

29:25

right vers we are clearly also aligned

29:28

100% owned by the staff uh the founders

29:31

as well as additional staff. So I think

29:34

between that so it attracts a different

29:37

kind of people and uh we have been

29:40

fairly successful in attracting and

29:41

retaining uh good people and uh uh you

29:46

know using cloud computing enables us to

29:49

have scalable infrastructure

29:52

uh in terms of we are not certainly

29:54

lacking in terms of experience several

29:56

of the partners have 20 plus years of

29:58

experience so I think we're no different

30:01

in that a smaller firm firm in pharma,

30:04

biotech or a smaller tech firm being

30:07

incentivized to innovate and come out

30:09

with new methods, right? I think we're

30:11

very similar in that regard

30:14

for sure. And I remember cuz I wanted

30:18

to, you know, prepare for this and make

30:21

sure that I could ask good questions. So

30:22

I called Yash, partner adverser, and I

30:25

was and he mentioned something that I

30:27

found very interesting. He said, "People

30:29

who work with you rarely ever churn. You

30:31

know, they stick with you." And

30:32

obviously, you know, Jun on the junior

30:34

level, maybe there's some churn, but

30:35

people who stick around, you said

30:37

generally they really like working with

30:39

you. And I guess my next question is is

30:43

more about how how you spot talent,

30:45

right? Um because obviously

30:49

um adversary you can't run the same type

30:51

of business as a citadel or a millennium

30:53

where you can you know run a essentially

30:56

a a people business where you hire

30:58

people churns through them make sure

30:59

they're generating you know great

31:02

riskadjusted returns if they're not cut

31:04

them and you know you have an army of

31:06

pods and but the whole thing runs runs

31:09

well essentially right for you you're

31:11

you're you're a smaller firm and and so

31:13

I think that on some level. Uh any given

31:17

hire can be a huge value ad, but I think

31:20

um it's also a bigger risk, right? They

31:22

can do more damage than they could at at

31:25

at a larger shop. And so I guess from

31:27

your end as as the founder of of Verser,

31:30

how do you screen talent? You know, how

31:32

do you when do you you know, see someone

31:34

and say, "Okay, this guy's this guy's

31:36

going to be exceptional. We need him."

31:38

>> So I think uh we like I think with most

31:41

people, right, so we will start with the

31:43

hard skills, right? So the hard skills

31:46

that are relevant uh for what we do are

31:48

mathematics, statistics and programming

31:52

uh right because if uh if somebody does

31:54

not have those skills then really

31:56

there's not much they can do in a quant

31:58

for right so you start with that it is

32:00

necessary but it's not sufficient right

32:01

so we start with that then we also look

32:04

for people who have demonstrated some

32:07

ability to solve hard research problems

32:11

uh and the problems need not be in

32:13

finance or investment

32:15

Right. So uh right so for example we do

32:17

take people with background in physics

32:20

we do take people with background in

32:23

biology. Folks who have tackled hard

32:26

research problems where uh it entails

32:29

working with a lot of data. The data is

32:31

not very clean. It needs to be uh uh

32:34

sort of massaged in a in a using

32:37

statistical methods. Uh the signal to

32:40

noise ratio is very low. So there's a

32:42

lot of noise in the data. Uh those are

32:44

the similar to the kind of financial

32:46

investment problems that we do right. So

32:49

that's second. Third, then once you

32:51

start getting into the softer skills, it

32:53

becomes harder and it takes time.

32:55

>> Uh is to look for folks who are willing

32:59

to learn. Right? So I always sort of say

33:01

that look uh once you go through your

33:04

college education uh directly there's

33:07

very little of that you'll use in

33:09

certainly in the setup that uh a quant

33:11

hedge fund but you have to it gives you

33:14

the skill set to continue to learn right

33:16

so we need to determine whether people

33:17

are continuing to learn and are

33:21

have the ability to generate new ideas

33:24

that takes time right so that typically

33:27

takes time so when a person comes on

33:28

board it takes us you know a year or two

33:32

uh to determine that and last but not

33:33

the least teamwork. Uh different firms

33:37

have very successfully implemented

33:39

different structures. There are some

33:41

quant firms who are very successful as

33:44

silos where people don't talk to each

33:46

other uh but they're very successful. Uh

33:48

and there are firms where there's a lot

33:50

of collaboration and teamwork. The model

33:53

that we have chosen is of teamwork. So

33:56

the so again I sort of put that in the

33:59

softer skill set. So it's really a

34:01

combination of those. The harder skills

34:04

I would say are somewhat easier. You go

34:07

through tests etc. We do the same thing

34:09

and it always pleases me because there

34:12

are times where I ask folks who are

34:14

already working with us uh to suggest

34:17

some of their friends etc. uh who might

34:19

be uh interested and when they apply and

34:23

I ask them what happened they said no

34:24

the tests were too hard they failed

34:26

right it it really pleases me because it

34:29

says that look we are uh we are we have

34:32

the right amount of sort of you know uh

34:35

caliber of people that we are we are

34:37

looking for but uh the softer skills do

34:39

take time so I it typically takes a year

34:42

or two so we do get some churn I would

34:44

say in a year first year or two because

34:47

I think it does become clear uh whether

34:49

uh whether the folks who joined us have

34:52

the skill set and more importantly do

34:55

they have the right temperament uh

34:57

because there's a high pressure

34:58

situation right you you have to deal

35:00

with the volatility of P&L and again in

35:03

a smaller firm you're a lot more exposed

35:05

to those day-to-day

35:07

>> pressures than in a larger firm

35:10

>> out of all the things you said there the

35:13

one thing that stood out to me because

35:14

I've heard that you know technical high

35:16

bar um soft skills, teamwork,

35:20

but I think one of the things that I

35:22

don't hear talked about very often is

35:26

the ability to generate great ideas

35:29

and I guess I want to hear it from you.

35:32

How do you build a culture or how do you

35:35

invest in people so that at the end of

35:38

those one two years of let's call it

35:40

general training, they're able to really

35:43

contribute in a in a different way to

35:44

everyone else.

35:46

>> Yeah. So I think the the important thing

35:48

is for uh the senior folks the more

35:52

experienced folks right uh me included

35:55

to continuously emphasize in research

35:58

meetings that everybody's equal ideas

36:01

will be evaluated on the merit of the

36:03

idea and not who brought up the idea

36:06

because the human tendency is to say

36:08

look you're the head of the firm or

36:11

somebody is head of research and it's

36:13

their idea so we have to do that right

36:14

so I

36:16

I do think we have been able to uh put

36:19

in a culture like that. Of course,

36:20

you'll have to talk to some of the other

36:22

folks to see if they feel agree the same

36:24

way or not which is what I'm seeing. But

36:26

I I do get the sense right that so once

36:29

I I think that environment has to be

36:31

there where folks have to feel that if

36:35

if they come up with ideas and the ideas

36:37

get shut down. What I tell people is uh

36:39

there's no guarantee that all ideas that

36:41

you come up with will be implemented.

36:43

But what we guarantee though is that

36:45

every idea will be evaluated and full

36:49

merit with full consideration and that

36:52

that we commit as a firm as a culture is

36:55

to take the ideas and I tell people that

36:58

don't feel bad if your idea doesn't get

37:00

implemented eventually because it may be

37:03

because of a number of reasons. uh a lot

37:06

of ideas that we come up with fail

37:08

anyway. Right? It's not like uh every

37:10

idea that I come up with succeeds. Not

37:12

at all. Right? So the idea is to the the

37:15

whole objective is to get people in an

37:18

environment where they feel comfortable

37:20

in suggesting new ideas goes through a

37:23

evaluation process and people are not

37:25

penalizes for penalized for ideas that

37:27

may not work out. I would rather shoot

37:30

down an idea at the evaluation stage

37:32

than for it to get into the portfolio

37:34

and see a lot of damage being done to

37:36

the portfolio returns and I think that's

37:38

a continuous process uh that we do.

37:41

>> Absolutely. And you know Deepo we've

37:44

talked a little bit about your story um

37:48

um IIT which we have some people in our

37:52

financial engineering masters program

37:53

who came from IIT and I can tell you

37:55

they're top-notch. So, um, we talked a

37:59

little bit about your, you know, Invest

38:01

Corp and then eventually building

38:02

Verser.

38:05

We've got a lot of people here who are

38:06

very driven. You know, they want to

38:09

maybe one day do something like you're

38:11

doing. Um, what advice would you give?

38:14

I'll make it specific because I feel

38:16

like normally people ask these questions

38:18

very generally. Let's say someone here

38:20

wants to one day start a quant fund,

38:22

right? What would you do? tell me the

38:26

step one, step two, step three.

38:28

>> So, look, uh I think I started uh my

38:33

fund after uh 20 years uh at Invest

38:37

CCOP. That's certainly on the longer

38:39

side, right? So, when I I I've seen I've

38:41

been fortunate I've been associated with

38:44

several hedge funds uh from their

38:46

inception, right? So uh clearly uh I

38:51

think if I sort of go back right you

38:54

have to have a value proposition

38:56

right as the last thing the world needs

38:58

is yet another hedge fund. So there is a

39:02

value proposition needed right. So for

39:04

that value proposition

39:06

uh it will differ from person to person

39:09

right look I think it's much harder

39:13

right it's doable but much harder

39:15

straight out of college to have a value

39:17

proposition it could happen there have

39:18

been instances very few but I think the

39:21

path that most folks take is to uh work

39:25

for a few years uh get to understand the

39:28

business they get to understand

39:30

different strategies and determine

39:33

what exactly is their edge or what is

39:38

their interest what is their edge uh and

39:40

then go to that value proposition aspect

39:43

right but you have to start I would say

39:45

the first starting point is a value

39:46

proposition that let's say you're

39:48

looking to start a fund what is the

39:50

product and what is why should somebody

39:55

invest in that product I think once you

39:58

answer that then you can look at the

40:01

next steps right it's not just enough to

40:03

say I want to start up a new fund. I

40:05

want to start up a fund, right? What is

40:07

what kind of product and what kind of

40:09

edge will it do? And yes, once you do

40:11

that, then there are other things that

40:14

need to be rolled out, right? Where

40:15

would the initial capital be come out?

40:17

Uh if you're running a quant fund, it

40:19

also requires capital to build out the

40:21

systems and uh hire the initial set of

40:24

people before you get to profitability.

40:26

Uh I think there are well established uh

40:29

mechanisms for that right within the

40:31

industry. But each person if you go to

40:34

somebody and say I want to start a fund

40:36

uh give me capital. They're going to ask

40:38

the same question that I'm asking. What

40:40

is the value proposition? What is the

40:42

product? And why should I as an investor

40:45

bet on you? And the more convincing an

40:48

answer you have uh you are you're going

40:51

to be more successful in being able to

40:53

do that. Right? So I think uh there's no

40:55

sort of straight answer for you need x

40:57

number of years, 5 years, 3 years. It it

41:00

will differ. It'll depend upon the

41:03

individual. It'll depend upon the

41:04

opportunity. I have seen a big

41:07

variation, right? But very few people

41:10

just come out just starting off without

41:12

any experience, right? Very few. There

41:14

are some, bro, there are uh uh you

41:17

mentioned citadel, right? Look, look at

41:19

the background there, right? Certainly,

41:20

right? Remarkable. But there are very

41:22

few like that, right? uh I think most of

41:25

the times the pattern I have seen is

41:28

folks who have worked uh either uh in a

41:31

in a bank or in another hedge fund have

41:34

worked for a few years developed an

41:36

expertise

41:38

and uh feel that they have the uh

41:42

ability and the motivation to uh set up

41:45

and set up a firm uh on your own

41:48

right fortunately I think nowadays uh

41:50

there are there is a sort of a midway

41:53

mechanism as well where you could go to

41:55

some of the firms that you mentioned

41:56

right uh some of the build the part shop

41:59

model right uh where

42:02

>> I think there are increasing number of

42:03

people are using that as well

42:05

>> right so they may work at a hedge fund

42:07

they develop a skill set uh and then get

42:09

approach a multi-large multi-manager or

42:12

a mid-size multi-manager platform and

42:14

get capital to start trading that could

42:16

be a good introduction to also before

42:19

you decide to set up a an independent

42:21

firm right so there are lots of

42:23

different variations that one can do.

42:28

>> If you were a college student today and

42:34

I shall rephrase the question because I

42:36

think I so I was going to ask if you're

42:37

a college student today, you know, what

42:39

would be the the path to to to doing

42:41

what you did and you kind of laid it out

42:43

there. um

42:46

work at a place where you can learn as

42:48

much as possible to where you can get

42:50

into a position where you you you have

42:51

that value proposition and then sell it.

42:55

Um, but

42:57

I think that it's very easy as a young

43:00

person to look at industries like quant

43:03

trading or like investment banking. Um,

43:06

you know, look at very very high status

43:09

industries, right? And and and run as

43:12

fast as you can at them because you have

43:15

the skills to do so. Um, if you were a

43:19

young person today,

43:22

would you run in that direction or do

43:25

you think that there are people who who

43:28

who I guess

43:31

go for something that is very often

43:34

viewed as the high status thing to go

43:36

for and maybe you're making a mistake by

43:39

doing that. I

43:40

>> I think the key is to determine uh what

43:43

is your interest,

43:46

right? So uh for example right I said I

43:49

was always interested in mathematics and

43:52

statistics and computing programming and

43:55

I actually wrote my first trend

43:58

following program

44:00

while I was still at college right never

44:01

ran any money on it uh or anything right

44:04

but so I think clearly it was an

44:08

interest right I don't know how it got

44:10

to that but it was I was very interested

44:12

in that so I think uh one has to be very

44:16

interested tested in the field right so

44:18

you use the word high status right

44:20

>> uh I wouldn't do it for that I

44:22

personally would not do it for that

44:24

reason uh and uh it is again human right

44:28

that some people will do that right uh

44:30

and then maybe develop an interest and

44:32

so on right again everybody has

44:34

different paths but I think determining

44:37

what is your interest now I'm not saying

44:39

that you will know you'll have it fully

44:42

uh outlined in terms of what will happen

44:44

in the next 20 years or 10 years, right?

44:46

Far from it because the industry

44:48

continues to evolve. But the starting

44:50

point has to be a level of interest,

44:52

right? So, if you're interested uh in

44:55

this particular field, why are you

44:57

interested? You need and do you really

44:59

enjoy doing it? You enjoy working on it

45:02

24/7

45:04

uh because you like it, right? I I guess

45:07

then that's the right reason. That's

45:08

personally my view. That's the right

45:10

reason uh to do that, right? And then

45:13

the opportunity in the first few years

45:15

certainly is to learn as much as

45:18

possible right again there is peer

45:21

pressure you know to uh because you know

45:25

when you're out of college you get

45:26

evaluated based on where you are what

45:29

kind of compensation and so on and so

45:31

forth that is very natural that's human

45:33

right but I do personally think that uh

45:36

knowledge and learning more if you are

45:38

interested in an industry learning about

45:41

that industry

45:43

uh the more you can learn about it in

45:45

the first few years uh I think is the

45:47

most valuable

45:49

personally that's again just speaking

45:51

personally and based on my personal

45:54

experience and I think also having seen

45:56

a lot of successful people uh within

45:59

this industry

46:01

>> last question before we open it up for

46:04

Q&A

46:06

um Deepak you've achieved a lot clearly

46:09

um you've built a very successful

46:11

business. Um, but you're still at it,

46:14

right?

46:16

What would it take for you to think to

46:18

yourself, I've made it. Are you already

46:20

there or is there I don't know some next

46:25

milestone that you'll shoot for and then

46:27

you can say you've made it.

46:29

>> Uh, I think that'll never happen. Right.

46:32

So there is no I have made it. I I just

46:35

don't get that sense. Right. Yeah. ctain

46:38

satisfied with what you know we've

46:39

achieved so far. uh in fact I was having

46:41

a partner meeting recently right and I

46:43

was telling people right that if if if

46:46

let's say if you decided not to if I

46:48

decided not to work on words sir for a

46:52

single day more uh I would still be very

46:54

satisfied right I would look at it and

46:55

say it was a success but I'm not

46:58

stopping and uh we certainly so I'm

47:00

saying there's a difference here we are

47:01

proud of what we have achieved uh but

47:04

it's not you you can't get complacent

47:06

and say I have made it and uh so that's

47:09

just again a personal philosy

47:11

I love that. Thank you very much. Um,

47:14

and all the best.

47:16

>> Thank you. Thank you very much.

Interactive Summary

Deepo, after a 20-year career at Invesco where he rose to become the sole head and CIO of the hedge fund division, founded Verser, a $1.4 billion quant fund. His motivation was to return to focusing exclusively on niche quantitative strategies and hands-on investing. He highlights the dramatic evolution of the quant industry, which has seen an explosion of data, advanced statistical methods, and research, growing from a small, opaque sector to a $5 trillion institutional industry. Deepo stresses that starting a hedge fund is a marathon, requiring a clear and unique value proposition, significant time, and capital. Verser's success is attributed to its early adoption of cloud computing, integration of alternative data following a 2017 conference, and continuous use of AI and machine learning. To attract and retain talent against larger competitors, Verser emphasizes motivation, alignment, direct impact visibility for employees, and a culture of teamwork, being 100% staff-owned. He looks for hard skills in mathematics, statistics, and programming, alongside problem-solving ability, a willingness to learn, and teamwork. For aspiring fund founders, the core advice is to first define a compelling value proposition and gain substantial industry experience, recognizing that success is a continuous evolution rather than a fixed destination.

Suggested questions

9 ready-made prompts