0:00

If you're competitive, markets

0:02

are quite a good arena

0:05

to play in, but you have to

0:07

be at least in the top three to make money. Turn

0:10

$25,000 into $125 million essentially, it would be more.

0:15

You can segment things in different ways, but

0:17

I think what we call trading

0:20

and winter media is mostly market making. Private

0:23

side is easier and harder.

0:25

Private side is essentially access

0:27

and filter. When you get access to

0:30

the right deals and when you have access to the right deals, can

0:32

you say this is a good deal I want to invest in or

0:34

not? You need to have

0:36

at least 14 investments to see your money back.

0:39

Usually the first seven or eight, so just go down to zero.

0:42

You have three or four or five of the

0:44

business. He returns your money and will give you a bit

0:46

more. Then he's sent out one investment

0:48

every less than 10. That

0:50

actually just yields 80% of your return. This

1:01

episode is brought

1:03

to you by Gnosis. Gnosis

1:10

builds decentralized infrastructure for the

1:13

Ethereum ecosystem. With

1:15

a rich history dating back to

1:17

2015 and products like Safe, Cowswap,

1:19

or Gnosis Chain, Gnosis

1:22

combines needs driven development with

1:24

deep technical expertise. This

1:27

year marks the launch of Gnosis

1:29

Pay, the world's first decentralized payment

1:31

network. With a Gnosis

1:34

card, you can spend self custody

1:36

crypto at any Visa accepting merchant

1:38

around the world. If

1:40

you're an individual looking to live more on

1:42

chain or a business looking to white

1:44

label the stack, visit

1:46

gnosaspay.com. There

1:49

are lots of ways you can join the Gnosis

1:51

journey. Drop in the Gnosis

1:53

DAO governance form, become a

1:55

Gnosis validator with a single GNO token

1:57

and low cost hardware, or

2:00

deploy your product on the EVM

2:02

compatible and highly decentralized Gnosis chain.

2:05

Get started today at gnosis.io.

2:10

Chorus One is one of the

2:12

biggest node operators globally and help you

2:14

stake your tokens on 45 plus networks

2:16

like Ethereum, Cosmos, Celestia

2:18

and DYDX. More

2:21

than a hundred thousand delegators stake

2:23

with Chorus One including institutions like

2:25

BitGo and Ledger. Staking with

2:27

Chorus One not only gets you the highest

2:30

years but also the

2:32

most robust security practices and

2:34

infrastructure that are usually exclusive

2:36

for institutions. You

2:38

can stake directly to Chorus One's public

2:41

node from your wallet, set up a

2:43

white label node or use the recently

2:45

launched product Opus to stake up

2:47

to 8,000 ETH in a

2:49

single transaction. You can

2:51

even offer high-year staking to your

2:53

own customers using their API.

2:55

Your assets always remain in your custody

2:58

so you can have complete peace of mind.

3:00

Start staking today at Chorus.one. Welcome

3:05

to Epson and the show which talks

3:07

about technologies projects and people driving decentralization

3:09

and blockchain revolution. I'm Brian Crane and

3:11

today I'm speaking with Jo Turpin who

3:13

is the co-founder of Vintermute.

3:16

Vintermute is one of the largest

3:18

or maybe the largest market making

3:20

company in crypto and

3:24

yeah I think it's a

3:26

field I'm super excited to dive in here so thanks

3:30

so much for coming on, Joon. Thanks

3:32

Brian. Thanks for inviting us. We're

3:35

the largest spot market making

3:37

firm. We're not the largest across

3:39

derivatives yet. Not yet. Working on it as

3:42

though a spot is true. Define

3:44

most likely as well. Getting

3:46

there. You know working to be so like

3:49

we can't claim the full name yet. How

3:52

did I know about the crypto space in

3:54

general? We need to do a bit more

3:56

derivatives but we're working on this. Thanks

3:58

for the invite. Thanks

4:00

so much for coming on. We

4:03

hung out a bit in Singapore recently and it

4:05

was really great to talk and so I'm super

4:07

excited to dive

4:09

into interviews. But

4:12

maybe to start off, share

4:15

a little bit, how did you become interested

4:17

in trading and markets?

4:21

Well, okay, trading and markets are

4:23

well before crypto. So

4:26

I am to be like full disclaimer, so I'm 41

4:28

and I started to get interested in trading when I was 12. And

4:32

I was essentially reading newspapers,

4:36

so my dad would be a subscriber

4:38

to something called the Figaro in France.

4:41

And they had essentially that a

4:43

main newspaper for just typical

4:45

news, politics and the likes.

4:48

And they had like, they had like,

4:50

also what we call the salmon papers, which

4:52

are essentially like similar to the FT, sort

4:54

of the market papers. And I was just going through this and I

4:56

was just took to

4:58

compassion to read through that. One

5:01

of these kids who had

5:03

kind of 120 options of like what they wanted to do when they grow

5:05

up. So I wanted to

5:07

be you know, like, what are the doctor,

5:09

experimental physician, astrophysicist between essentially the age

5:11

of 7 and 12. And

5:14

then from 12, I really wanted to be a trader. So

5:17

I got into this discussion

5:19

with my dad, had

5:21

a stock portfolio that he would

5:23

manage. We didn't have the

5:25

internet per se, we had Minitel. So

5:28

Minitel was quite popular in France at the

5:30

time. You just log in into the

5:33

6637, it's similar to a web page, but you just

5:35

type in a number. And

5:37

essentially, we just log in into the what was the credit

5:40

union at the time, and we just we could just buy

5:42

stocks such. So I started a

5:44

study like this,

5:47

putting some orders into to buy stocks with

5:49

my dad's money when I was 12. It

5:53

wasn't a ton of money, but it was just enough to be

5:55

able to passion for it. And then

5:58

I got him to buy me like a an

6:00

options book when I was 15. So my first

6:02

option took and then basically I couldn't trade options. So basically

6:04

I found a way to trade warrants. Essentially,

6:06

you could only buy stuff. I

6:09

was right directionally, but

6:12

I realized that the

6:14

banks on the other side, essentially warrants a bit different from

6:17

options in the way that they should buy banks or

6:20

like a warrant issuer. And essentially they

6:22

were taking so much margin that I was right

6:24

in my bets, but I was giving too much margin away.

6:27

And it's probably one of the first things that led me

6:29

to think of more like market making percent. So

6:32

I kind of hired one of my, like

6:35

a little cousin who's, I think he's five

6:38

years younger than me over summer and I was going

6:40

through like sort of finding arbitrage across the page and

6:42

so on. So anyway, I started slightly quite

6:44

early on market in

6:46

general. I think the theme has always

6:48

been that I would

6:51

go through different subjects

6:53

and I would get a bit bored of it. Half

6:56

of my family is entrepreneurs, half of my family

6:58

are educators. And then one thing I really didn't

7:00

want to do at the time is just not

7:02

be a teacher as such because I thought it was just a

7:04

bit too repetitive, going through the same

7:06

lessons and teaching the same things every year. And

7:09

one aspect in market side of it is just

7:11

changes all the time. Yeah, these

7:13

economic cycles, they say every three, four years or so,

7:16

you can see that in crypto, you get like,

7:18

when do people call it more like seasons

7:20

and such. And it just

7:22

changes enough and it's noisy

7:24

enough that it forces you

7:26

to relearn things and so on. So

7:29

I found that, oh, if I want to have some

7:31

sort of career for 10, 20

7:33

plus years, it needs to be linked with markets.

7:37

So in a way that it needs to just

7:39

be refreshing enough. Yeah, that's the

7:41

long short of it. And then I ended up training

7:44

my first teachers when I was 18, got someone to write off,

7:46

sign up

7:50

for a professional investor letter for me to

7:52

trade a future in some bull's trauma when

7:54

I was 18 as an internship. Basically, I

7:56

qualified that as an internship with the... school

8:01

and then they wanted to like a

8:04

bank as the first sort of gap year and then wanted to do

8:06

trading as the first job at 23. So

8:09

when you got interested in trading

8:11

this early, you mentioned one of the things

8:13

you liked was this fluctuating,

8:16

the change. Do

8:18

you think that was the thing that kind of

8:20

caught your attention and got you to deep in

8:23

there or were there like other aspects that ended

8:25

up being so fascinating for you about

8:28

training and markets? There's

8:32

a competitive level. That's

8:34

interesting. There's a few things. I

8:37

did some self-trading before doing

8:39

trading professionally and I found

8:41

that self-trading was not always

8:43

aligned with the customers and I didn't really like

8:46

it and I found that trading was much more

8:48

aligned with, it's pretty clear

8:50

where the P&L comes from. I

8:53

found that was just a very honest way

8:55

of making money. So I quite like that

8:57

aspect. You get very

9:00

clear feedback, very tangible feedback if you're doing things right,

9:03

you basically make money, if you're doing things wrong, you

9:05

lose money. It is much more like

9:08

honest process. I think it

9:10

gave me a sense of independence as well

9:14

to the degree that you can trade some things

9:16

on your own as such. But

9:19

I think there's also not quite a lot of

9:21

things that we do today that are to be

9:23

data. We can only

9:25

do because we have the infrastructure, because we have the team in

9:27

the background. There's quite

9:29

a lot of things that have to

9:31

be quite collaborative between getting everything aligned

9:34

between the right access to capital, right access to

9:37

exchanges, right access to an

9:39

inventory, right strategies

9:42

and so on. So I think there's

9:46

also a collaborative aspect that's interesting there. I

9:49

think if you're competitive, markets

9:51

are quite a good arena

9:54

to play in because it's global pretty

9:56

much from day one. So

9:58

if you think about us when we started. in

10:00

the crypto space, it was essentially you're both from

10:02

day one. So you know if it's going to

10:04

work pretty quickly, then it's not

10:06

destroying some sort of follow with zero to

10:08

one like when it works. You

10:11

want to be in the top three definitely, ideally,

10:13

you're the first one in certain vertical

10:16

as such. So you want to

10:18

be first one in spot trading as such. But

10:22

you have to be at least in the top three to make

10:24

money. But no, we

10:27

interact on Twitter here and there and so on, and it's

10:29

more about I think people

10:32

get somehow satisfied to be doing a bit better.

10:34

But I think the realization, especially in

10:36

trading is that you need

10:38

to be doing better than the competition all the time.

10:42

And it's a bit more about you

10:44

know, there's this marginal advantages that you

10:46

need to add on to that

10:48

make it quite competitive. So

10:51

it's just, yeah,

10:53

keeps you on your toes. I

10:57

want to go into this but mainly before

10:59

that, then what's the story of

11:01

how VTmute started?

11:04

There's a convergence of two things. There's

11:07

a convergence of Yevgeny

11:10

set up the company 25th of July 2017. And he was doing

11:12

like little sort of hackathons here and there

11:17

with our first CGO R. And

11:20

then on my side, I essentially had

11:22

already built a few companies, a few

11:24

trading related companies, but I see also

11:26

some startup studio, a

11:29

bit more of a VC advisory

11:31

firm as well. And

11:33

on my side, I was going back into trading 2016-17.

11:36

And I was sitting next to

11:40

this ex-martial waste. I don't know if you know

11:42

of martial waste, but it's probably the

11:44

largest long-shore touch fund in London that

11:47

you know, a Philip Marshall, I

11:49

think it's a Philip, made

11:53

a few billions out of it. But essentially, they managed

11:55

something like 86, $90 billion of

11:57

the UM and they and

12:00

they do a lot of long short, more

12:02

in the equity space. And one

12:05

of their first employees was sitting up, was sitting not

12:07

too far from me on a trading desk.

12:09

And he was setting up his own macro fund, and he was all

12:11

over crypto in 2017. So we reviewed

12:13

a lot of ICOs together, about 56, I think.

12:15

And he wanted

12:18

to have a macro hedge fund that would

12:20

have actually a crypto sleeve, what they call

12:22

it. Essentially, they have a macro hedge fund that

12:24

have different strategies, and you have a sleeve of it.

12:26

So part of the allocation is in crypto. And

12:29

he wanted me to manage that allocation. So

12:33

it came to mind that, well, I can't do

12:35

this alone. I need a,

12:38

you know, I can trade, but I need someone else to

12:40

help with, just

12:42

automate by everything, some more with CTO, and

12:45

I need another trader because it's 24 seven.

12:47

And then essentially, just I reconnected just discussing

12:50

very openly with some ex

12:52

colleagues from trading, suggested

12:55

me to reconnect with you again. And then we

12:57

we aligned on this and I joined

12:59

in as a co founder, essentially late

13:01

2017. We put everything

13:03

on paper with with our CTO actually on in the

13:05

early 2018. That's the decision or thing

13:09

that he he had just left a

13:11

trading farm and he wanted to do something in crypto.

13:13

And I had basically the allocation we didn't

13:16

we did not end up taking money from

13:18

from that fund, actually. But I was enough to just just

13:20

for doing it like rekindle the interest for

13:23

me to get into crypto. And philosophically,

13:27

for me, it's more like I had done

13:29

both trading and VC before. And

13:31

I found that crypto was super interesting in terms of

13:34

like democratizing like liquid is one we

13:37

call like liquid venture essentially, so

13:39

very early stage bats, but that

13:42

you could get access from just with

13:44

trading very, very early on, which

13:46

I couldn't find in the private markets

13:48

anymore, like within private and public markets

13:50

in what to to

13:52

add some background to that Brian, where

13:55

to like, I had

13:58

done first non listed. investors in

14:00

2008. I was a pretty

14:02

active angel investor, so London and song between

14:04

2012 and 1516,

14:08

let's call it. I

14:12

think you still have this in web 2 where in

14:16

the past 90s or so, you'd

14:20

be a city investor at Amazon. Amazon just

14:22

lists like three years later, sometime

14:25

around 1999, you can still buy Amazon in

14:27

$1. You can actually get a lot of the

14:29

growth and you can get a lot of the upside. But

14:33

after 2009 especially, while

14:35

you get into the year by listing, you get into the

14:38

Facebook listing, there's nothing like there's not, there's not

14:40

necessarily a ton of upside left. So

14:43

everything just lists like 10, 15 years

14:45

into the life of the company. And

14:47

it's much more what we call encrypt or

14:49

extra liquidity, but it's key equity

14:51

like typical web 2, share

14:54

all those of listing companies just end

14:56

up becoming, you know, extra liquidity for

14:58

VCs or private equity groups essentially. So

15:02

I was, I found it really refreshing

15:04

that crypto was trying to solve

15:06

that problem essentially. And another

15:08

aspect that I love in crypto is

15:10

more like I've dealt with

15:12

people in the trading space and the hash space

15:14

from the EFS space and stuff. And people

15:17

can be a bit cagey and

15:19

enders anders space. And I found that

15:21

actually like people in crypto are always

15:24

quite open. And maybe you

15:26

made some scammers here and there, but

15:28

actually in general, people are quite that

15:30

pretty goodwill and are quite

15:32

a lot more interesting to interact with. So

15:36

I'd love to get a bit into

15:38

sort of trading and like, I

15:40

think a lot of people they will

15:43

have and I, you know, they will know

15:45

trade a lot of listeners will know sort

15:47

of trading in the sense of,

15:49

you know, okay, I have

15:51

some US dollars and euro

15:53

or something, I want to buy some

15:55

Bitcoin, I go on like an exchange

15:57

and I put in, you know, maybe

16:00

a market order, maybe a limit order,

16:03

and I buy some tokens,

16:05

right? Or like maybe I

16:07

do it on the exchange, I buy some shares. But

16:10

I think that's probably the extent

16:12

to which most people listening

16:15

here will understand trading. And

16:17

obviously that's something very different

16:19

from having the things you did

16:22

previously at before

16:24

crypto, at this training company

16:26

or then at Adwintmute. Can

16:29

you tell me what's

16:31

trading? When you say you're trading, what

16:33

does that mean and what does that

16:35

look like? You

16:38

can segment things in different ways, but

16:40

I think what we call

16:42

trading at Adwintmute is mostly market

16:44

making. So essentially, we're there

16:46

in the order book, so we're there in

16:49

a marketplace, we're there in the exchange, and we're

16:52

there to always be so that if someone

16:54

wants to buy or sell, there's someone on

16:56

the other side, someone like us, Mark and

16:58

Mickey, there's someone on the other side there to

17:00

fill our order. Because

17:03

let's say if you, Brian, if

17:05

we both individuals and somehow

17:07

like you want to sell Bitcoin at 70K and

17:09

I want to buy it at 60K, nothing happens.

17:13

And the market makers are there to make sure that we

17:17

always show what's called a

17:19

bid and an offer. The price

17:21

of which we can buy and the price of which we can sell. And

17:25

then we're trying to be there to always make

17:28

trading easier and the cheapest possible. So

17:32

what we call trading tends to be associated

17:34

with market making. Now most

17:37

people, to your point, most people call trading

17:39

and are trying to buy with

17:42

that $1 and hope that it gets $2. That's

17:49

often not

17:52

the greatest strategy. I think the way that most

17:54

people should be trading in crypto is

17:57

to see it as venture. different

18:01

x amount of money, either they don't

18:03

do any research and just stay in like BTC and

18:05

ETH and sort of the top four or five names

18:07

as such. We

18:09

start to see more what's

18:11

called more basket products and more index

18:13

products as such that would give you

18:16

exposure to the top 10, 20, 30

18:19

names which

18:21

is for most people is probably a better way to

18:23

do things. I mean,

18:25

how do you trade? Let's

18:28

ask this, this side that way. I

18:31

know you've done some venture investments but

18:33

do you trade anything, any

18:36

liquid tokens? I'm

18:39

very like buy and

18:41

hold and don't do much else and

18:43

you know maybe. Do

18:45

you have any leverage when you take a position or? No,

18:48

no. Okay, sounds very cute. Very

18:50

leverage. It's pretty good. I

18:53

don't have

18:55

the impression that I'm

18:58

good at predicting basically

19:00

market movements so I don't

19:03

try to do that. Yeah,

19:06

so I mean that's quite reassuring. I

19:09

have made money historically with some

19:11

cycles like in stocks or so like taking

19:14

directions but through

19:16

previous firms, not necessarily through interview. Whether

19:20

it's VCs as the venture arm is doing quite

19:22

well but it's very much

19:24

like venture investment so again to separate

19:27

like I think what you refer to as trading

19:29

is much more liquid

19:31

investments but on the private side, private

19:34

side is easier and

19:36

harder. Private side is essentially access

19:39

and filter. Can you get access to

19:41

the right deals and when you have access to the right deals, can

19:43

you say this is a good deal I want to invest in or

19:45

not? And

19:47

that's venture for you. The

19:49

problem usually of venture or in

19:52

crypto and some of the spaces is just that you're

19:56

married to the position. The good thing with it is

19:58

that you don't tend to have any leverage. So if

20:01

it goes on to zero, it goes on to zero, but you

20:03

get psychologically to

20:06

do things going on to zero. Now

20:09

you'll understand why I mentioned this for

20:11

trading is I think most people should

20:14

consider even when it's liquid crypto, they

20:17

should consider it like a venture bet. They

20:19

should consider that they're actually making an investment

20:21

that actually should be pretty much with them

20:23

down to zero. And they

20:25

also should look at the stats of actually venture

20:28

investments. So venture investments for like, I

20:30

don't have the stats for crypto, but

20:33

like for Web2, Angel, like even

20:35

10, 15 years ago, you could already get all the

20:37

stats where you needed to have at

20:39

least 14 investments to see your money back. And

20:42

usually the first seven or eight, so just go down to

20:44

zero. You have three

20:47

or four or five that basically return your money

20:49

and it'll give you a bit more, then you

20:51

tend to have one investment every less than 10.

20:54

That actually just has the yields 80% of

20:56

your return. And

20:59

that causes some challenges

21:01

when you think of how people trade,

21:04

which means that the

21:06

winners should

21:08

essentially return 10 plus times their money. And

21:12

I think it's very, very hard psychologically for

21:14

most people to think of like keeping

21:17

something that does 10X and not sell it

21:19

before it reaches 10X. Another

21:22

parallel with this is that, do

21:25

you know what valuation the Uber investors invested seed? It's

21:29

2008. So 2008, do you know

21:32

what valuation they invested? I

21:34

don't know. $3 million. So

21:39

essentially when Uber listed, people

21:42

had a 5,500X. So

21:46

whatever J-Call and Strands and Son made like

21:48

$25,000 into 125 million, essentially,

21:51

but a bit more. But

21:53

that's basically the math. So

21:56

imagine if Uber had a 5,000X, but it's

21:58

not a 5,000X. been liquid before

22:00

that. I'm pretty sure that

22:02

F-HOLT's sold some of it. It's

22:06

a bit of a challenge of thinking of investing

22:08

in crypto. You

22:11

end up with a liquid venture and then you need to

22:13

find a way to trade

22:15

maybe. If

22:19

you sell your 5,000, if you have 100X but

22:21

then it turns into a 5000X, it's

22:23

quite a big miss. It brings us

22:26

to the challenges of comparing venture and

22:28

trading. Trying to find a way to

22:30

invest and trade is actually pretty hard.

22:34

With the market making that you guys do

22:36

at the interviews, how do you make

22:38

money? We basically

22:40

are forced into a position we don't want. We

22:42

call it a position that is being normal or

22:44

short. As long as you make money if things

22:47

go up, short to make money if

22:49

things go down. We're forced into a position by

22:51

providing liquidity by just basically saying

22:54

if you want to buy Bitcoin for me. Let's

22:57

say there's a particular coin

22:59

and then you have buy

23:03

orders and you

23:05

have sell orders. That means

23:07

you have to hold both

23:09

sides. What

23:12

cost us money in the first place

23:14

is infrastructure. People to run algos

23:17

to send the orders and so on.

23:19

Coast of capital

23:22

to borrow the

23:24

coins on each

23:26

side. Even if sometimes we get

23:28

a sign and a ship borrows but

23:30

there's still commitments, there's still some things that are associated with

23:32

it. The borrows

23:35

are often from the

23:38

team or

23:40

foundation or whoever is issuing the

23:42

token. That's much less than half

23:44

of what we have in the balance sheet.

23:47

Some of the borrows are from foundations.

23:49

Some of the borrows used

23:51

to be from the lending borrowing platforms like

23:54

BlockFi and the rest. It must have been

23:56

actually going to us. We have

23:58

some borrows in DeFi. that you'll

24:00

know through like wildcat people, the

24:02

maple of the well and so on. But

24:05

essentially, yeah, a lot

24:08

of borrows come from some of our shareholders or

24:10

so, just directly from the emphasis and so on.

24:13

So we borrow directly from foundations and

24:15

other people. So we have a certain

24:17

cost of capital. And then we

24:19

make money by essentially being in the order book

24:21

and trying to capture as much as possible of

24:23

the spread. So we essentially

24:25

get paid over time to take risk

24:28

because we pushed, as I was saying earlier, we pushed

24:31

into a position because you want to buy. We don't

24:33

actually want to sell, but necessarily we want to

24:35

sell at that price. And then

24:37

we hope that we can hedge it as in

24:40

we can buy back whatever we sold to you.

24:42

We hope we can buy it back at a slightly lower

24:44

price. But let's

24:46

say I want to buy a token and

24:50

then you guys

24:52

have to sell order. So now you're buying

24:54

that token at that sell order. And then

24:59

you hope that maybe you can sell

25:01

it at a different exchange at a

25:03

better price than buy

25:06

it at a better price

25:08

than you sold it to

25:10

me or do you hope that the price moves? Market

25:14

making is actually, this summarises is

25:16

quite simple. It's quite simple to

25:18

explain, but quite difficult to implement.

25:21

Market making is just what do

25:23

I think this and the length of

25:25

this, but Bitcoin is worth that time to how

25:28

wide do I want to quote and how what sort of size

25:30

do I want to quote it? So you

25:32

just have width, quote size

25:35

and then your mid price or what I

25:37

think the theoretical value of the token is.

25:41

And then we quote around this and as

25:43

yes, there's small amounts of money that

25:45

come from arbitrage, so basically

25:47

be able to sell, you know,

25:49

near enough at the same time, sell on one

25:51

exchange, the same asset and just

25:53

buy it, buy it better than the other exchange at

25:55

a lower price. But yeah, often it's

25:58

just, it's pretty much trying that

26:00

price things as well as possible. It

26:04

turns out to be like it's much more about, it's quite

26:06

a competitive space in the way that to

26:09

be able to trade both sides, we need to be

26:11

really within the thread. So we

26:13

need to be the best offer on one side and the best bid

26:15

on the other side. And

26:17

it's very infrastructure-driven. So it means that

26:19

we need to be able to

26:21

adjust the orders safely so we

26:24

don't get taken out by other people. So

26:26

there's some infrastructure cost in that. And

26:29

obviously, you don't have the capital

26:31

constraints, there's cost of capital

26:33

because we

26:36

borrow all these assets. Yeah, long

26:38

story short, we make money by

26:41

taking your risk, we

26:44

don't necessarily want to be short this Bitcoin, but we

26:46

want to buy Bitcoin. And

26:49

now if you have

26:51

different companies that compete,

26:55

different market makers that compete, you

26:57

mentioned you need to be the top or top three

26:59

in stuff to make money, but what differentiates

27:02

the top from the

27:05

other players? So

27:08

that's an interesting discussion that I had even in

27:10

touch with when I was at

27:12

my first venture in trading in 2011.

27:17

And I actually did it in 2014 and I was

27:19

looking at trying some other

27:21

top five firms and just trying to learn from

27:24

other people. Basically, my first venture was the one

27:26

senior trader there and I felt I didn't really learn

27:29

enough. And I think a lot of it is

27:31

actually much more about cultural differentiation. I think

27:33

if you can attract them, I'd tell it because

27:36

people are very honest and they can actually just

27:39

bail. But I think there's an aspect of people

27:41

being aligned and being there and being traders for

27:44

the right reasons. Often people

27:46

don't actually stay traders just for

27:48

money. It

27:50

can be quite rewarding financially, but it's actually

27:52

often people just get, if you

27:54

are just interested in money, they just get the first bonus and they

27:56

sort of leave and they do something else that's a lot less stressful.

28:00

So either people can like math,

28:02

they like whatever problem solving, there's other

28:04

aspects that make it interesting. The

28:08

math comes in

28:10

because when you are

28:12

basically around determining this or the midpoint,

28:14

the width and things like that, that's

28:16

when you use some kind of data

28:20

analysis. Yeah, yeah,

28:22

you try to predict as well as

28:24

possible. It's never

28:26

an exact science, which is why

28:28

infrastructure comes into play quite

28:31

heavily, where it becomes

28:33

a more exact science because you end up trading

28:35

very high frequency. I think some people know

28:38

the term of GFT is sort of high frequency trading

28:40

as a thought. And you end

28:42

up trading, you end up making money every

28:45

day at least at the start and ideally

28:47

every hour because

28:50

you trade very, very high frequencies here. So I

28:52

think we trade four, five, six million times a

28:54

day these days. It

28:57

becomes very granular. So

29:01

it also becomes very infrastructure dependent.

29:04

So it means you

29:06

need to essentially 50

29:08

plus people to

29:10

go and cover, just

29:13

make sure that everything is running smoothly. And

29:15

especially in a crypto space between CFI,

29:17

DeFi, between the OTC side of

29:20

the business, between an API offering, you

29:22

end up crossing quite a few things. Getting

29:26

into this a little bit, like if

29:29

you compare market making in

29:32

a traditional market versus crypto,

29:35

what are the biggest things that

29:37

you need to be good at specifically

29:40

to do well in crypto markets? So

29:44

I can tell you what I think is

29:46

a good differentiation for the business is that

29:48

we build a very good brand of reputation

29:51

by being honest people and not

29:53

looking for the short term optimal

29:56

that we can really wish for for a long time.

29:58

So I think that was quite a bit of a challenge. quite key. So

30:02

maybe you can explain why that matters because you

30:04

think if all if you just have the orders

30:06

in the exchange and you know it just takes

30:08

the best order I don't even know who's on

30:10

the other side so on

30:12

that side it sort of shouldn't matter right? So

30:15

as soon as you try to see your face

30:17

counterparties we would call counterparties they're not really customers

30:19

as such they're just because we

30:22

tried as we would call principles we try to

30:24

own capital. So

30:26

essentially you want to be

30:29

reliable like when you show a price basically

30:31

you know you're there for

30:33

certain price versus in size for an extreme amount

30:35

of time. So reliability is

30:38

being more predictable is worth something. So

30:41

market making can be done I think people in crypto

30:43

space are used to hear the term

30:46

towards foundations but it's also towards

30:49

exchanges. So if

30:51

you're reliable towards exchanges do that when

30:53

things move around but so on then

30:56

the exchange is

30:58

just the room essentially and then you need

31:01

the exchange needs people like us to be in

31:03

the room to trade with some other many of

31:05

the users and if

31:07

you're reliable as a market maker on an exchange and

31:09

you know you're there when things move and so on

31:12

essentially their users will have better execution will

31:14

have a better experience in general and

31:17

that's quite valuable to an

31:19

exchange and then

31:21

just build a good relationship with that small akin

31:24

to a partnership

31:26

with an exchange than anything else.

31:29

And then do the exchanges also

31:31

pay directly for the liquidity you

31:33

provide? Very nice and ones do

31:35

but like usually they don't. You

31:38

know you want to be in a situation where they're big

31:40

enough for the big enough about

31:42

the opportunities they show on us. Yeah

31:45

they're way

31:47

there to take risk with our money and take

31:50

risk you know providing liquidity as such but

31:52

then the exchange should be

31:55

there to bring users on

31:57

board and do the onboarding into

32:00

space essentially. One thing I

32:02

didn't do

32:04

is how we build the business is that, and

32:06

especially the difference between crypto and

32:10

the tradfire space, is

32:12

that there's a whole default

32:14

element that we actually focused on

32:17

early on, especially from 2019. And

32:20

we found that it was quite a good differentiation to

32:22

the business, but also it's something that

32:24

we could only do because we were

32:26

much longer term focus. So

32:29

essentially we took VC money and we

32:32

very much pitched ourselves and we

32:34

still see ourselves as a tech

32:36

company first before being

32:38

a trading company. So we essentially took

32:41

VC money which was just quite different from

32:43

having investors from a hedge fund space or

32:45

trading space. Which means that we

32:47

didn't have actually people on our back

32:49

asking every day what the results were

32:52

and we had more people, much more patient

32:54

just researching us build infrastructure

32:57

first. So 2019 we started on

33:00

our on DYDX for example in

33:02

around August 2019 and that

33:04

enabled the rest of the business to actually be

33:07

more robust around essentially on chain transfers

33:10

to the point that we're

33:13

probably driving Manson users like crazy

33:15

at some point thinking that they're

33:17

following smart money but they're

33:19

just seeing transfers going back and forth. So

33:22

if you have tens of thousands of transfers,

33:24

I don't know what sort of signal

33:26

they can get from it but probably not much from

33:28

us. So I think there's a big

33:31

differentiation of being able to trade on chain is

33:33

quite a differentiation from the trade fight. What's

33:36

the biggest difference about trading on

33:38

chain versus on centrist exchanges? There's

33:42

aspects around obviously just direct integration with

33:44

the chain but there's aspects

33:46

of managing your own custody,

33:49

thinking of potential sandwich

33:52

attacks as such, thinking of

33:54

the risk of actually not really being failed

33:58

as well if the cost So just,

34:00

you know, pushing for a transaction and

34:02

not always knowing exactly when the block

34:04

is going to be, you know, confirmed.

34:08

There's quite different level of uncertainty in terms of,

34:10

you know, what you traded and what you didn't

34:12

trade between, you know, DeFi

34:15

and essentially CFI and

34:17

like the traditional finance. So

34:19

these things are quite, is quite different.

34:21

On the other side, once you've made the effort to

34:24

integrate and to build a certain amount of

34:26

DeFi knowledge, then there's,

34:30

you know, if it's a barrier to entry

34:32

for you, it's a barrier to entry for

34:34

the people as well. So yeah, this

34:36

is, well, it becomes, it becomes a

34:39

competitive advantage. There's

34:41

some aspects that we

34:44

realized later as we were starting

34:46

to invest. So

34:48

we started to invest in

34:50

different projects from late 2020,

34:53

essentially from the summer 2020 when I

34:56

think Ewes was probably DYDX for that. I

34:59

don't know if it was a series A or series B at the time

35:01

already. But essentially

35:04

because of the operational experience

35:06

and trading experience and dealing with

35:08

these, these DeFi protocols,

35:11

we had a good amount of insights on

35:14

just doing a technical due diligence

35:16

for investing. So

35:18

we ended up just, we

35:21

ended up dealing like different relationships with VC

35:23

firms, for example, which is here coming to

35:25

us and like, not only from a

35:28

potential trading perspective to help, you know, build

35:30

some other volumes for them, but also in

35:32

terms of just getting feedback on due diligence

35:34

on just how robust was

35:36

a certain protocol, for example. Well,

35:39

what do you feel like were the key things that

35:42

allowed to interview to get, you know, become

35:44

so big? I mean, it's a

35:46

very competitive space, right? Full alignment

35:48

of blood. Full alignment. So

35:50

full alignment of blood, where you want to

35:53

get into position. There's a bit of an

35:55

execution play in terms of, because

35:57

We had a really good team to start with in terms of

35:59

the experience. Riyadh and sunset. Add

36:02

express crazy many. If. Any built

36:04

since it. Is the first. Company.

36:07

Divini of a belt but it he adam. He.

36:09

Built the each year form of the of a

36:11

large market maker Oats and a tough I mock

36:13

America and Europe. Said

36:15

you him could express any Jess. And

36:18

I'd go to stress and non the often

36:20

trading but also just raising raising funding As

36:22

such. An. Office to deal with squat pica

36:24

didn't have as getting for skeleton of and for such

36:26

as we had a. We. Had some risking

36:29

that. I'm. One.

36:32

Factor that was interesting is because we ended up

36:34

closing the first round of funding. And.

36:37

with quad out of difficulty. In.

36:39

The middle of two thousand and eighteen Crypto

36:41

winter. We ended up in a space

36:43

that was our seat. Client. He. And

36:46

then we ended up with dealing with

36:48

Freddie into O O G Groups and

36:50

Crypto Lived Oshima. Comments on. Who.

36:52

Are quite heavy jarvis around and. Them.

36:55

In our and winters visited upon is very small.

36:58

Say. A young have like if she

37:00

first that survives that an Nsx to club

37:02

has trip into this is pretty healthy interaction

37:04

does that. she resilient enough reserves is quite

37:06

a happy that I think she. Guess you

37:09

got started down the same time you about

37:11

six in our seventies. All this I'm up

37:13

Yeah yeah of course on he shouted a

37:15

dad yeah basically said of twenty aging. I.

37:19

Think it's excellent for that like it's of the

37:21

the the base because you just have a focus

37:23

on building. In. Our the distracted with

37:25

dislike Like we are not as and

37:27

say distracting. I'm but it's

37:29

fun. Hims is more more than happy to

37:32

see like me, eyes and song as as.

37:35

For. The price to. The. Prices to the marketing

37:37

for us as into it. But. In general

37:39

is kind of. Caesar. To

37:41

saw think having this pretty good answer even and what

37:43

to and bubble on and so on. And

37:45

says of people billing, successful company starting and

37:47

like in crosses his. Ah,

37:50

So. That. That

37:52

was helpful. I think

37:55

it was really helpful that. two

37:57

things there was really helpful the giving out of

37:59

really the idea of like being

38:02

tech fast and being like very much

38:04

product driven and being in your everything

38:06

being automated and we

38:08

never really had trader roles we've

38:11

always had dev slash quan slash

38:14

trader roles so anyone basically

38:16

just code and everything gets

38:18

into into play

38:20

very very very very

38:23

soon very very quickly you

38:25

mean so this this basically means that like

38:29

it was always to focus on like you know

38:31

building a system and building algorithms as

38:33

opposed to because I mean

38:35

I guess it's always algorithm is market

38:37

making no I mean I guess that

38:39

people not doing this manually nobody you'd

38:41

be surprised there's a lot of trans5 firms who

38:44

who have very separate roles between people

38:47

who there's quan roles

38:49

to be able to people who build the

38:51

algorithm they actually traders who do like little

38:53

manual trading actually so

38:56

you'd be surprised that we start even in crypto with

38:58

staff competitors who are actually trading a

39:00

lot more manually than you would think so

39:04

I think there was there was a strong tech focus there

39:07

and the tech the motor is

39:09

things being reliable and scalable essentially

39:13

now on my side I had

39:15

good experience with you know raising

39:17

raising equity and raising debt so

39:20

we had the capital that access capital

39:22

sorted and we were

39:24

all very aligned in the vision

39:27

of having like

39:29

in terms of the culture I think essentially it was

39:32

more startup like when we started then

39:34

now we had to find them we

39:36

had to compensate people with a

39:39

hybrid of like bonus and shares over

39:41

time because they were just getting pushed a bit

39:43

too too much by essentially hedge funds and in

39:47

anyone who wanted to enter the space because

39:49

the shares were too like too

39:51

much of a long-term thing and

39:54

yeah so it's still a still a

39:56

balancing act but I think we've got

39:58

something quite quite

40:00

nice nowadays. I think

40:02

you'll end up with something like a

40:05

program where people have few shares but

40:07

they get some bonus once a

40:09

year or so and then they can reinvest that bonus as they

40:13

choose and we

40:15

start to need to have some secondary answer here.

40:18

We invest a bonus in buying

40:23

shares or in lending

40:26

into the company or something.

40:28

Yeah, essentially so. We

40:30

very much want to have

40:32

some VC shareholding as such but we really

40:34

want to be as employee owned as possible.

40:39

It makes a ton of

40:41

sense for companies that trade

40:43

their own capital essentially because you want

40:46

best alignment possible between people who trade

40:49

and people who own

40:52

the capital. These things we got

40:54

right. Some of the

40:56

things were difficult to fully align on at this

40:58

time because some of the things are much longer

41:00

term in terms of building the

41:02

brand, building the beauty side of the business

41:05

versus building the day-to-day algorithm

41:08

trading side of the business. I

41:10

think we clicked around late 19 at

41:13

the age of 2020 when

41:16

we had the first COVID hits and March

41:18

and the first big dips in the

41:21

Bitcoin at the time. Well,

41:25

actually we saw all the algos actually caught robust

41:27

and we started to make real

41:30

money but essentially we started to, we

41:32

were building the rocket for two years and

41:34

we started to really put fuel in the rocket two

41:37

and a half years in or so.

41:40

So that took some time. What

41:42

was that because that the algorithms weren't good

41:44

enough? Some

41:46

of it was because of this. Some

41:49

of it was because we didn't, we had not,

41:53

we are not rich enough of a critical mass between

41:55

what we could show, exchanges that what

41:57

we could become because essentially exchanges, need

42:01

to build a certain volume to actually get into

42:03

better and better fee schedules. That

42:06

was a challenge. And then

42:08

simply, we had not built

42:10

enough in terms of getting cheap

42:14

enough access to debt as well.

42:17

So we were using equity, equity

42:20

fundraising to hire

42:22

and build a tech. But

42:24

actually for trading itself, we needed to borrow.

42:27

And getting, slowly getting access to inventory

42:29

was a, it becomes

42:31

a bit binary, basically. Nowadays

42:33

we're one of the two to three

42:36

firms to get essentially first steps on access

42:38

to inventory because we're

42:41

quite reliable and because we turn

42:43

to these things quite quietly. When you say access

42:45

to inventory, what do you mean? Your

42:48

ability to borrow tokens from

42:50

other teams or the people in general.

42:54

So inventory says we trade what we call

42:57

market neutral. So we don't make many things

42:59

go up and down. But it means

43:01

we need to, we can't buy inventory all

43:03

the time. We can't buy

43:05

the tokens themselves. Because if we buy the tokens, then

43:07

we're sensitive to things going down. But

43:10

often we can, we can quite a lot larger

43:12

size by simply borrowing the

43:14

tokens and trade around it. Yeah.

43:17

Yeah. I mean, I

43:19

guess you can see also here

43:21

the sort of advantages of size,

43:23

no? Because like, yeah, you trade more

43:25

and then because you're trading so much,

43:27

I guess the exchange trading fees matter

43:30

a lot and

43:32

then makes it hard for someone

43:34

to compete. Yeah.

43:38

2019, we still had the down year because

43:40

we, 80 plus percent

43:42

of the money we were making trading was

43:44

going into fees. Wow.

43:47

But it was just, and something

43:49

like a rule of thumb that kind of works

43:51

for me, like for proof sums, I

43:53

don't know exactly why we are nowadays, but

43:56

it was more like if a side of

43:58

the money we make, if you make like. one

44:00

dollar on a trade and you spend 30 cents

44:03

in fees, you can run

44:05

a business. Because

44:07

you start out financing costs behind the staff to pay

44:09

your people. If you start out

44:12

money, you have to start out to pay bonuses

44:15

and taxes and stuff. So

44:17

one third is maybe

44:20

moderately where you can

44:22

start scaling. But 2019 was

44:24

essentially 80% of the money going into

44:27

exchange fees. In

44:31

terms of the interviews company culture,

44:33

you mentioned something. You mentioned this

44:35

thing of trying to be a

44:37

start-up mentality, having a sense

44:40

of ownership and also this long-term

44:42

focus. Do you feel like

44:44

there are other things about the interviews culture or

44:47

maybe the way you've hired people that have been

44:49

crucial? There's a bit of

44:51

– so for further

44:54

background, I think I mentioned – maybe

44:57

I didn't make it clear. Evgenia

44:59

and I used to work for a company called Optiva

45:03

in Amsterdam. So the quite large

45:05

firm is still very much an option firm

45:07

in Tratfire. And it's

45:09

very much – the

45:11

Dutch culture is 100% imprinted at that

45:13

company. And how

45:15

brutally honest Dutch people are. And

45:18

then essentially it's part of

45:20

the culture. There's also a big

45:23

differentiation where we took the best

45:26

from it in the way that it's a completely

45:28

company-wide bonus pool. And

45:30

we took the same thing, which is – what

45:33

I noticed is that a lot of training firms

45:35

are structured very differently. A lot of training firms

45:37

have little siloed teams and no one

45:39

shares information. But actually when you're in a firm

45:41

and training firm that actually has a company-wide

45:43

bonus pool, people do behave

45:45

a lot more like shareholders or

45:48

such or at least think that if

45:51

they share knowledge with – the

45:53

senior people are sharing knowledge with junior people. It just

45:56

– there's a

45:58

lot more of that dynamic when people are having a good time. happy

46:00

to teach other people. And

46:02

that's basically some percentage of the profits go

46:04

into this bonus pool. And then how do

46:07

you determine who gets how much

46:09

of that bonus pool? There's

46:13

still something that we work

46:15

even to perfect. But

46:19

there's this philosophy that even people like with

46:21

us, there's even people like in legal, so

46:23

like, oh, can't say everyone is

46:25

a P&L contributor to a degree. We

46:27

have internal recruiters and so on. And external

46:29

recruiters, but they said it will also assume

46:31

that everyone helps recruiting. So I think there's

46:33

a few functions like this where obviously that

46:39

we want people to feel that owner

46:42

is as much as possible. And

46:45

I think that that has

46:47

worked well. I think I mean,

46:49

a general advice for how

46:52

many people do you have nowadays? The

46:54

chorus one? Chorus one, like 60 ish,

46:56

65, something like that. And

47:00

when did you, assuming at

47:02

the start, you were always involved with

47:04

hiring, when did you stop being involved

47:07

with hiring? Oh, I still am. I'm

47:09

still interviewing. Yeah, I'm still interviewing basically

47:11

everybody. My real firm is like

47:13

the first 20 people or so.

47:16

It's really, really key to

47:18

get right because then you

47:20

imprint the culture of the company and basically on the

47:22

first maybe five, 10, 20 people. And

47:25

then they actually do. Then there's actually

47:27

people you hired who help you scale

47:30

and not learn. They scale your training. Yeah,

47:33

I know I'm still, for me,

47:35

I see like interviewing

47:38

is like just really crucial and

47:40

I think I'm pretty good at

47:42

it. So I, I

47:44

focus quite a lot on it. And at the

47:46

same time, I feel we haven't figured

47:49

out. Well, or

47:51

like, I like one of the

47:53

things I still really want to do. And I

47:55

don't think we're there. It's

47:57

sort of to figure out how to.

48:00

formalize, systematize like the injuries I do

48:03

so that I feel like okay I

48:05

can have like other people do it.

48:08

I mean there was just there's a few people

48:10

and last year I was out for like a

48:12

few months or like a month and a half

48:14

because of this like health issue so then there

48:17

was a few people we hired that I didn't

48:19

interview but I've interviewed except for like three or

48:21

four people I've interviewed everybody. That's

48:23

interesting. I got less

48:25

and less involved with it. I don't know I

48:27

mean to check with the beginning I'm

48:29

pretty sure he doesn't get involved with everyone but

48:32

I think we very structured maybe we've got too

48:34

many rounds of interviews as like maybe now. How

48:36

many rounds do you guys have? Like

48:39

for a BD person, Subban 5. They

48:42

go through like a use case and so on. For

48:45

technical interviews for like

48:49

some of the infrastructure guys I'm

48:51

not too

48:53

sure but I think I mean

48:55

beyond that I think cultural fit is still really

48:58

to keep off of it. I think

49:00

if they yeah this fit in the culture and

49:02

it's kind of the challenge or the question now

49:05

that we're expanding across Asia is just

49:07

maintaining that culture. If they see

49:09

that they fit in the culture like how do you how

49:12

do you what questions you ask or how do you

49:14

assess that cultural fit? Depends

49:20

on the role. Often for like

49:23

traders who live from John Fyte to

49:25

crypto. We know we don't

49:27

have some excolates we know people who know them we

49:29

can get a lot of third-party

49:31

validation. I think there's

49:34

a there's a there's a red line around like

49:36

for everyone's like honesty that's quite key.

49:40

Long-term focus. Yeah

49:42

people being entrepreneurial being quite

49:44

self-reliant to be honest like

49:46

quite I like to say

49:48

self-repelled. You just don't really need

49:50

to just push that much. We have a lot

49:53

of people in the company that I can just go and pick at

49:55

3-4 a.m. like I need to tell them good to bad. So

49:58

my head of trading is like this. a few

50:00

people I just think there's a

50:02

strain of work, work, the whole

50:05

is in but so that's

50:07

I like to think is well placed at F but

50:09

then you need to force people to be in holidays

50:11

you know though quite a few people in the beauty

50:13

team in Singapore like this. Yeah

50:16

I think there's this this this three elements like

50:18

this and like sort of the work ethics it's

50:20

very cheap. I'm thinking long-term

50:22

enough and being

50:24

much more mission driven than mercenaries and

50:26

so on. The challenge is

50:29

I think hiring like the interview process

50:31

is often not sufficient so I think

50:33

it's just like you know the

50:35

first one two three first month

50:37

is often key to make sure

50:40

that we didn't make a mistake but um

50:43

we don't. I

50:45

think we've done a good enough job

50:47

to be honest like surprisingly enough and

50:49

the hiring because we don't like

50:53

we don't that many we don't have really

50:56

people even leaving that much or we

51:00

did hire fairly slowly think

51:02

about it like founded in you

51:05

know July 2017 we had the first time in

51:07

September we officially

51:10

October 2018 and

51:12

then finished a year of a lot of

51:14

basically five people total finished

51:17

2020 barely 10 or 12 finished 2021 we did all of

51:19

2021 with between 20 and 25 people

51:26

and now they really started to hire

51:28

more because we we've done much more

51:30

into product company from 2022 and I

51:33

think we finished with 50 people 2022 or so

51:35

slowly really really slowly hiring but thanks

51:38

to that we never we never had a we

51:40

never had any rounds of layoffs as well. Yeah

51:44

yeah that's very similar with us. One

51:46

of the things that I

51:50

noticed across one in terms of hiring that

51:52

I feel like has been one of the

51:54

most reliable predictors is

51:56

whether people are really passionate about

51:58

crypto. And

52:00

it happened a bunch of times that we hired

52:02

people who were, you know, they seemed like a

52:04

good fit in other regards. So like, you know,

52:07

we really needed someone for the position, but didn't

52:09

really seem to care much about crypto. And

52:11

then none of these have

52:13

worked out. They all ended up leaving, basically.

52:17

Yeah, we had this with a, it was more 2022 or

52:19

such. Like,

52:22

I think people, some people are a bit too optimistic. I

52:25

think maybe that's informal from 21. And

52:28

then it's sort of like start and then just come in and

52:30

then winter starts or something like this. And

52:33

we have people who are sometimes tempted by a tried fire

52:36

as well. And it's true that

52:38

22, maybe 23 was actually quite busy in

52:40

effects and commodities. It's like,

52:43

that's fine. Like, there's a few people leaving, but

52:45

it's far from being neural. Like, usually people stay

52:47

and, you know, understand, understand

52:50

the cycles and whatever. There's

52:52

pros and cons to any side of the cycle, to be honest.

52:55

It's a bit of a filter as well. And

52:58

I'm with you that people only have some

53:00

passion or bind to the vision of the

53:03

space, at least in terms of either democratization

53:05

or like, be a bit like,

53:07

slightly libertarian, whatever, you know, they need to buy to

53:09

the vision of the space at least. They don't have

53:11

to be as like

53:14

crypto believers as, you know, if you're not. Well,

53:17

we're quite strong. But I went to the crypto

53:19

thinking that basically sort of crowdfunding had a bit

53:21

failed. I was really into crowdfunding and I said

53:23

to Phil and meanwhile, in

53:25

terms of democratizing, investing and so on. I thought

53:27

that was quite interesting. And they said that, you

53:30

know, so many rules around being

53:32

a sophisticated investor and so on. Especially in the US, like if

53:34

you try crowdfunding in the US, there's just a bit

53:37

of a way. It's

53:39

not really crowdfunding. You need 5 million on

53:41

your bank account outside of your

53:44

house to qualify to be able to invest

53:46

and so on. I

53:48

mean, different doctors. I want to qualify. I was

53:50

just thinking about what I said before. I want

53:52

to qualify it a little bit. I think it's

53:55

passionate or at least make curious about it.

53:57

I think we've also had people who worked

53:59

out to work. like I'm kind of

54:01

skeptical about crypto but I'm also very interested

54:03

in it and like I'm trying to understand

54:05

it and like that I

54:07

think is also so something

54:09

the indifference is probably more anything of

54:12

like that is like okay if people

54:14

have feel like it's just

54:16

a sense of indifference about it and then

54:19

they didn't work. Yeah, that's

54:21

a bit worrying but that is probably

54:23

it's probably a massive right like the

54:25

best. We've

54:28

had people who like depending on

54:30

the roles we've had good

54:32

success with people who actually came from trot-fire

54:35

who are used to like too much structure

54:37

to a degree like the banks

54:39

or so and they were frustrated of like not

54:41

being able to do enough and

54:43

we've had good success with them because they

54:45

you know some of them had traded crypto

54:47

like PA or so and

54:50

and it is a balance because we actually we

54:53

want people to understand regulation we understand

54:57

like a not as like a like a

54:59

good long-term framework. So

55:02

sometimes people who are just purely crypto native are not

55:04

a good fit they don't understand how you know markets

55:07

should work and so on and how it works long-term.

55:10

But being a leader by this and it's

55:13

good if people have gone through you know

55:15

both both well both yeah both

55:17

have some experience of crypto some experience with trot-fire

55:19

I think I think it just ends up yielding

55:21

the best trot-fars. We have pretty

55:23

good experience hiring straight from university as well but

55:25

for more technical roles you see it's

55:31

it's yes it's it depends

55:33

on the roles. So

55:35

you mentioned building right and

55:37

and I guess one of the

55:39

things that you

55:42

know that you guys have been building that has you

55:44

know become pretty significant is

55:46

R-Sync which is

55:50

is it the I guess the largest block

55:52

below on Ethereum or certainly one of their lead

55:54

like. 25 is one of three

55:56

launch I think I think this is a

55:59

very even. I just said it. Yeah.

56:02

I think it is 25% of block building. I

56:05

don't know the latest numbers, but I think you

56:08

should know that better than me. Yeah. Well,

56:11

I know it's foreign for large, but I don't know exactly right

56:13

now. So

56:18

why did you guys decide to

56:20

do that? And what's sort of

56:22

the connection between the block builder

56:25

and the market making? I

56:28

think it's a good representation

56:31

of the fact that we tech fast,

56:33

that we very much build this fast, and

56:36

an analogy that I like to use, and you

56:38

probably have me say this for you guys, but

56:41

I'm not necessarily a fan of Google, but Google is a

56:43

good example. If

56:45

you think of Google, it's a tech firm. It's

56:47

a tech firm fast, and they happen to

56:49

make money from advertising, essentially. And

56:51

then we want to be, and

56:53

we are tech fast, and we happen to make money from trading.

56:56

But essentially, what we do day to day is

56:59

tech building. So there's only,

57:02

now there's about 10 people more like

57:04

me who are just more externally facing

57:06

and do the more commercial aspects

57:09

of the business, but most other

57:11

people are basically product buildings, infrastructure building,

57:13

and so on. So our

57:15

thing comes into play as a subsidiary,

57:17

as such, to help with block building.

57:20

And it's just making sure that we're part

57:22

of the block confirmation, as such. We

57:25

are quite cautious of not being

57:27

too vertically integrated. So I think the good

57:30

example of that is on the like

58:00

a system in general, you

58:02

need to have some infrastructure

58:04

built to be able to participate in trading. And

58:07

so block building basically means, right, that

58:10

it's an entity, right, that basically gets

58:13

transactions either from the mempool or that

58:15

people directly send to this block builder

58:17

and then, you know, basically put it

58:19

together in a block and then, you

58:21

know, send that to, you know, relayers

58:23

and validators and the one that sort

58:25

of has the highest monetary

58:28

value that implied

58:31

that it's the one that goes in there.

58:35

And sort of the

58:37

attractiveness here is, for

58:39

example, that, you know, you

58:41

can sort of construct your own blocks and

58:43

put in all of the arbitrage, all of

58:45

the transactions, the on-chain transactions that you guys

58:48

want to do and I guess you can

58:50

also prevent being sandwiched in

58:52

that block. Is that sort of the main? Because

58:55

you guys would still, would

58:57

you still submit transactions to

58:59

other block builders or like all of the

59:02

interview transactions that you guys do? That's a

59:04

good question. You need to ask

59:06

one of my default guys to get the

59:08

full answer. But I'm assuming we do because

59:10

we, you know, we would still

59:13

look if another block builder is

59:15

performing better, the trading team could

59:17

independently just push this to

59:19

a better block builder. Like

59:22

many in the space, we used flashbots in the

59:24

past and others. I'm

59:27

sure it works a

59:29

bit like an option, but it needs to stay

59:31

competitive enough to a degree

59:33

there's enough competition in-house to

59:35

make things more efficient. Because

59:38

if we don't keep things efficient enough, even

59:42

if it's between different teams basically

59:44

in-house, then some of the competition

59:46

will take it away. And we want to mean if

59:48

we only favor our own blocks, then

59:50

at some point it becomes inefficient. It's

59:53

a good question. And

59:55

I'm assuming, yeah, I'm assuming we'll use other blocks, but

59:57

it's true that in terms of defending

59:59

our essential. files against Senator Shataksev,

1:00:01

you want to have some control

1:00:03

there as well. So

1:00:06

I'm curious also a bit about your

1:00:08

thoughts on sort of the long

1:00:10

term evolution both with crypto and

1:00:14

what it means for a company like venture

1:00:16

mute and maybe like some of the topics

1:00:18

I'm curious your opinion about. One

1:00:21

is, do you think

1:00:23

that in the future, sort

1:00:26

of like most assets will be blockchain

1:00:28

based and that's how

1:00:30

markets work? Do you think

1:00:32

that decentralized exchanges are going to overtake

1:00:34

centralized exchanges and you know those will

1:00:36

be the largest exchanges and marketplace in

1:00:39

the future or like how where do

1:00:41

you see things going? So

1:00:43

it's funny because I've had different answers

1:00:46

to this question though that yes, it's

1:00:48

one thing is I'm

1:00:51

one of these guys who is

1:00:53

cautious with like you know people like

1:00:55

Larry Fink for example that's saying everything

1:00:57

will be tokenized. I think a

1:00:59

lot of things will be tokenized but I think for

1:01:02

some aspects like

1:01:04

around our ability for example so we're all

1:01:06

what I said so like around sort of

1:01:08

like tokenize the equity and stuff. It's

1:01:11

been tried before and some of

1:01:13

it if it doesn't bring a sufficient like

1:01:16

factor of improvement it's quite

1:01:18

difficult to distribute and just to

1:01:20

get adoption. So

1:01:23

I think it will happen but it won't happen

1:01:25

all at the same time. There's

1:01:27

a few things in the crypto space that I'm still

1:01:30

really eager to see like so I'm in Asia now

1:01:32

for six months because we trade derivatives

1:01:34

from Singapore from Singapore office and I think

1:01:36

there's a big thing that's still not happening

1:01:39

if you compare the crypto

1:01:41

space with TradFi. There's still very

1:01:43

very little volumes in options for example. So

1:01:46

essentially options in TradFi trade about 25%

1:01:48

of the total volumes and

1:01:50

in crypto is like it's a

1:01:52

tiny percentage it's just single or very low

1:01:54

single digits and

1:01:56

I think there's a few factors to explain

1:01:58

that but I think it's just. like with a bit more

1:02:02

that actually positively is digitalization of

1:02:04

the space. There

1:02:06

will be more people around the table and actually

1:02:08

we'll get more derivatives there. More

1:02:10

derivatives we did a bit of price discovery

1:02:12

down to the line tokens, down

1:02:14

to the base spot token as

1:02:16

well. That's the

1:02:19

second trend, more derivatives, more tokenization

1:02:21

for sure. I think it

1:02:23

was 25th of April 2022, I gave a

1:02:25

talk in Davos about DeFi flipping C5. I

1:02:28

was quite happy to discuss the topic. You

1:02:31

have such a memory for dates, whenever you were

1:02:33

mentioning dates you were like exactly ok it was

1:02:36

incorporated on the 5th of July 2007.

1:02:42

It used to be as more as the

1:02:45

grid strike, the very few

1:02:47

things I'm good at, just remembering people and

1:02:49

remembering dates. It's

1:02:53

funny because I was at this foundation because

1:02:55

they invited me to speak about it. One

1:02:59

inch of a little booth and

1:03:01

so on had a pretty nice

1:03:03

discussion there. We

1:03:06

were just right next to WEF

1:03:08

and they were all WEF guys

1:03:10

that basically all talk about oh 90% of

1:03:12

the money is DeFi, it's like criminal money but

1:03:14

we all know it's rubbish. But

1:03:17

I was happy to talk about potentially what we needed

1:03:20

for DeFi to flip C5. It

1:03:24

was interesting because there was like April 22 as

1:03:26

I mentioned, obviously we had FTX

1:03:29

in November. My

1:03:31

first reaction when FTX went down was like

1:03:33

oh people will realize that the thread is

1:03:35

because of centralization and that people will just

1:03:37

go for DeFi. The

1:03:40

problem is I had, I was talking about

1:03:42

this with a founder at the time,

1:03:44

was that actually

1:03:46

C5 is very good for adoption.

1:03:48

C5 basically is centralizing strangers regardless

1:03:51

of risk of fraud and centralization

1:03:53

and so on. It's very

1:03:55

much they're actually onboard the new users. When

1:03:59

people get sufficiently familiar with the

1:04:01

tech and so on and get sufficiently incentivized

1:04:03

to basically do custody

1:04:05

themselves, and they're essentially confident

1:04:07

enough and like the tech to actually

1:04:10

send, you know, send funds on

1:04:12

chain. Then

1:04:14

they really are part of DeFi. But

1:04:17

at the time, it was about

1:04:19

so turn two is about 400

1:04:22

million total users across the crypto

1:04:24

space. And you add about 6 million

1:04:28

four and a half whatever four, let's call it, I

1:04:30

think sorry, it was 300 million total

1:04:32

users in the crypto space and there was

1:04:34

about 400, four and a half million DeFi

1:04:37

users per se. And now it's

1:04:40

double across the board, but I see the percentage of

1:04:42

DeFi users stays the same. It's a

1:04:44

bit of like there's a bit of a failure in

1:04:46

terms of like DeFi in the way that we

1:04:49

did have like contraction around DeFi summer because there

1:04:52

was like zero yield everywhere or negative yields in

1:04:54

most of the places and basically DeFi was able

1:04:56

to to attract some users through

1:04:58

through having some yield. But

1:05:01

essentially, there was a whole phase afterwards

1:05:03

where well, TradFi actually

1:05:05

get it up as of getting yield again.

1:05:08

And, and I think the DeFi space

1:05:10

and essentially crypto was just bit stroking on the

1:05:12

on the on the yield front. So not not

1:05:14

to be attractive that I

1:05:16

think there's still a lot of things to solve in terms of UX

1:05:19

UI, getting that leverage and

1:05:21

DeFi, that some sort of, you know, credit

1:05:24

scoring in the meantime, it's a lot of

1:05:26

people are building pretty cool things to

1:05:28

help solve these things. But I think we still

1:05:30

end up with this sort of bottleneck from DeFi

1:05:34

to really flip. See, fire is actually still

1:05:36

quite challenging because if you see

1:05:39

how people are boarded into the crypto space,

1:05:41

don't ever see first. But if you think

1:05:43

like, you know, 20 years or

1:05:45

10, like a long term

1:05:47

future, so I think DeFi

1:05:49

work can can be more much more

1:05:51

more prevalent once you get like

1:05:54

once you get robots everywhere once you get

1:05:56

like proper Internet of Things, because

1:05:58

then whenever you're fridge

1:06:02

buying your food automatically based on what

1:06:04

you've eaten the day before and stuff,

1:06:06

your fridge is not scattered like sending funds

1:06:08

online. Well, your fridge is not scattered in

1:06:11

the custody of stuff like probably this, maybe

1:06:14

you just have your house wallet or whatever

1:06:16

you want to phrase it. Yeah, you can

1:06:18

have your fridge sandwich to have the heating

1:06:20

system. It's like disproportionate

1:06:23

like hacking risk as well. But

1:06:26

I think it's like the machine to machine trade. I

1:06:28

also feel like this in 20 years time. I think

1:06:30

it's pretty normal. But I

1:06:34

think that's probably precious

1:06:36

D5 first. And

1:06:39

it's surprisingly if you talk to institutions, a lot of

1:06:41

them are like there's a few even

1:06:44

three years ago at this side

1:06:47

chat, this little charter house,

1:06:49

like chat after panel

1:06:52

at a London conference. And

1:06:54

they were like traditional finance, they

1:06:56

have like these guys who call like rheumus custodians who

1:06:58

hold all the people's money. And so

1:07:00

they just top 10 out there, that's one of the

1:07:02

JP Morgan's one of them, but like BNY, there's

1:07:05

a few Australians there as well. And

1:07:07

there's a few of these custodians who are

1:07:10

completely ready to think of a world where

1:07:12

D5 is first. But basically

1:07:14

they want to be the wallet. So

1:07:17

they basically that thing is like, okay, people can trade,

1:07:19

trade actually everywhere, but

1:07:22

they want to be they want to secure

1:07:24

the funds essentially. So there's a future there.

1:07:27

I don't know how much we have to pay them, but I

1:07:31

think I think I think that can work. Dreds

1:07:34

is just not happening yet.

1:07:36

It's not happening yet. And

1:07:39

maybe one more topic when it comes to sort of, you

1:07:41

know, the future. What

1:07:43

about AI? Like, what do you think

1:07:45

is the impact of AI on like

1:07:48

winter mute? Very

1:07:50

little in terms of pricing and delivering most of

1:07:52

what we do is already said here to make

1:07:54

it. I think there's

1:07:56

a really annoying aspect that I'm trying

1:07:58

to give myself. head around is

1:08:01

on the commercial side, is around risk

1:08:03

of fraud or people impersonating and so

1:08:06

on. It's more and more

1:08:08

possible that tech is there for

1:08:10

you to think that you could have had the

1:08:12

podcast with me but you could have had the

1:08:14

podcast with someone else sitting

1:08:16

in you know from the 50th cent

1:08:18

army and actually just using Gemini, using

1:08:20

some AI tool to basically

1:08:27

replicate how I move, how

1:08:29

I speak and so on and thinking you have

1:08:32

a commercial call and

1:08:34

then convince you to

1:08:36

send me money for a deal

1:08:39

and so on. Yeah I think we're

1:08:41

not too far away from having to go

1:08:43

back to like physical locations and having different

1:08:45

offices in different countries and then just having

1:08:48

people to just check in that

1:08:51

we have this but actually blockchain solves some of

1:08:53

this because you think of that you know for

1:08:57

us to have you know either it's linked

1:08:59

with an address that's like we resumptile it

1:09:01

or not. We have a few

1:09:03

points of validation that we can

1:09:05

check and we can we can function trustlessly

1:09:09

but it's yeah that's a bit that's

1:09:12

a bit of a concern because there's

1:09:14

disasters about life in AI or so

1:09:16

on like in tech in general where

1:09:18

we favor convenience or the privacy or

1:09:20

where many other things you know of

1:09:22

a security and I think

1:09:24

it's a bit of the rest as well so I

1:09:27

don't know how to see AI

1:09:29

affecting your business as well. I

1:09:32

mean for us it's a pretty significant focus

1:09:34

I mean we have like a team that's

1:09:37

like focused on basically building sort of like

1:09:40

well I think the obvious thing is

1:09:43

that I think it could just improve

1:09:46

everyone's productivity by like maybe you

1:09:49

know maybe 20% maybe 50% maybe 100% by a

1:09:51

lot but then I think to do that you

1:09:56

know it's not like easy I don't think you would

1:09:58

get that just from like you think like 20pt

1:10:01

or something but you actually have to like

1:10:03

integrate it into how the entire

1:10:05

organization works so that is like a

1:10:07

focus for us. I think

1:10:09

that is the sort of thing where it can

1:10:12

become like something that's you know very compounding where

1:10:14

okay maybe this year will

1:10:16

help us like 5% right

1:10:18

but maybe next year it will be 20% then maybe

1:10:21

the year afterwards it will be like 50% and

1:10:25

then I think these

1:10:27

effects can just become like massive over

1:10:29

time. AI

1:10:31

is like for me AI is like so I

1:10:33

used to invest in air businesses in morning work

1:10:35

too and you can sort

1:10:38

of put them in like optimization automation or

1:10:40

discovery and

1:10:42

like and initiate so by the sound of

1:10:44

it what you think is it's more optimization

1:10:47

this day is still like a 10-20% you

1:10:49

know a marginal improvement on your

1:10:51

site but I think

1:10:53

if you just if you just get generally inside

1:10:56

like if you get the discovery aspect about that's

1:10:58

actually quite useful. What do you mean with discovery?

1:11:01

Yeah discovery is like think

1:11:04

of an AI that can think of

1:11:06

like you know alpha

1:11:08

go is probably an

1:11:11

example of this like you know they just think

1:11:13

pretty differently or if

1:11:15

you apply AI to like biology and

1:11:17

just finding that discovering like new treatments

1:11:19

or so. I

1:11:21

think that will probably emerge right but

1:11:23

I think the first thing is like

1:11:26

kind of like I mean

1:11:28

for example one thing is like in

1:11:32

terms of a sort of thing we

1:11:34

would like to have so right now you know

1:11:36

we run all this infrastructure for you know

1:11:38

lots of different blockchains and then you know

1:11:41

we have all these engineers on call and

1:11:43

then the you know systems so like alert

1:11:45

and then something happens and someone has to

1:11:47

look at it and then like respond to

1:11:49

it. So for example can you

1:11:51

build a system that like knows

1:11:54

how all of the issues

1:11:56

are right and all the ways they've been

1:11:58

resolved and if there is an alert,

1:12:00

it can like basically, well

1:12:02

autonomously resolve it. Or

1:12:05

maybe at least suggest to the engineer that comes

1:12:07

up, hey this is a lot probably you should

1:12:09

do this right. And

1:12:11

then maybe they can check but maybe some points not

1:12:13

necessary to check, you can just go ahead and do

1:12:15

it. So something like that

1:12:17

is for example something we want, yeah we

1:12:20

want to build but it's not easy. This

1:12:22

is going to take probably going to take

1:12:24

two years or something. What

1:12:28

is it? I think it's 83

1:12:30

with neural networks, 2006 was the

1:12:33

postman, how do you call it? Postman machines and stuff.

1:12:36

And then like it took over,

1:12:38

I think a lot of progress been made in the eye

1:12:40

just by like doing

1:12:42

just playing around with an optimization curve. I

1:12:44

think people were quite surprised with this, how

1:12:46

far we went. And

1:12:48

I don't know,

1:12:50

I played around with chat

1:12:53

DPT and so on recently but it's like

1:12:57

quite interesting because essentially anything

1:12:59

that deals with words used to be considered

1:13:02

very very difficult. Like you

1:13:04

could before you could build some trees,

1:13:06

you could build some things and actually just do

1:13:08

like chatbots, you could build them like in a

1:13:10

very narrow sort of lexicon,

1:13:12

so a sort of like very

1:13:14

narrow context and now the fact that we can have other

1:13:16

limbs is quite interesting. So I

1:13:19

don't know. I think it's often

1:13:21

the thing that like we sort of over estimate

1:13:24

what we can get in two years and we

1:13:26

massively underestimate the changes in ten years time. So

1:13:29

we'll see where we're all there. I'm

1:13:33

just curious about the last, maybe last thing. So

1:13:35

I understand like you know like a

1:13:37

lot of languages and you learn a lot of languages,

1:13:39

I'm curious like how do you

1:13:42

learn languages and how do you maintain the

1:13:44

languages that you learn? I

1:13:47

learned them fast and I maintained them very

1:13:49

difficult. It was very difficult for me to

1:13:51

maintain but I'm still maintaining

1:13:54

about five and

1:13:57

what I don't like 24-7 I used

1:13:59

to just have a just watch movies

1:14:02

or read or trying to read in a language

1:14:04

or try to... I

1:14:06

love Singapore for like Asian languages because you

1:14:08

can maintain Chinese, I can maintain

1:14:11

some Japanese, I can learn Korean slowly

1:14:13

with the limited time I have. There's

1:14:15

two to three big dimensions, there's

1:14:18

something in learning that you

1:14:20

call schemas. It's sort of like the

1:14:23

scaffold things of like how you learn. So

1:14:27

essentially let's say if

1:14:29

you're French, let's say this you have similar

1:14:31

schemas to like Spanish and Italian and so

1:14:33

on, so basically like same sort

1:14:35

of framework and I say it's quite easy for you

1:14:37

to learn all the Latin languages but

1:14:40

once you just like you're European and you learn

1:14:42

Asian language like it takes a while to get

1:14:44

there. If you've got good memory

1:14:46

for new words it's useful but

1:14:49

there's still a dimension why even

1:14:51

you understanding like rough

1:14:54

grammar is not sufficient to

1:14:56

really just you know like to

1:14:58

put words and speak perfectly. I

1:15:01

think a good way to do things is just

1:15:03

consider yourself a child and basically don't

1:15:05

be fussed about making mistakes. Like

1:15:07

I just half joke

1:15:09

about torturing people, not learning a new

1:15:12

language but basically torturing people, I'm

1:15:14

gonna listen to you like basically you're gonna

1:15:16

make mistakes and so on and I think

1:15:18

people are especially when you learn a new language

1:15:20

as an adult people are too fussed about making mistakes

1:15:22

and I think it's a bit it's a good lesson

1:15:24

for learning in general. Like if the

1:15:26

price language is where it's in the school. When

1:15:28

you learn a language you focus on one at

1:15:30

a time or if you sometimes learn multiple at

1:15:32

the same time? Just one at a

1:15:34

time usually. One

1:15:37

at a time is much easier. The

1:15:39

thing is to get to a level where you should

1:15:41

be able to have a list of

1:15:43

conversation or at least watch a read

1:15:46

like this kind of material or

1:15:48

so like simple things because you

1:15:50

want to be able to maintain it and

1:15:52

then you can only maintain it if you can have a

1:15:55

mix of active and passive learning. So

1:15:57

active learning is really you literally like

1:16:00

learning new words and so on and learning new

1:16:02

sentences and practicing. And passive is

1:16:04

basically just you watch a

1:16:06

Kdrama show or so, it's

1:16:09

much more passive but at least what's quite

1:16:11

key is get your ears to hear

1:16:14

all the sounds. And

1:16:17

I found this difficult for the Korean for example.

1:16:20

Life for Chinese is really difficult to get around

1:16:22

all the tones. I mean in biscuit tones

1:16:24

and they just generally get around

1:16:26

that. Many people

1:16:29

use soft tones in Chinese as well so it's more like

1:16:32

you just need to end up having to

1:16:34

learn to say like the

1:16:36

same thing in three different ways and just make sure you can

1:16:38

be understood. Then just

1:16:40

by practicing and be a bit shameless

1:16:42

about practicing and just I

1:16:45

think this is the biggest barriers for adults actually.

1:16:48

But in general, I find this

1:16:50

fascinating in terms of language learning because

1:16:53

it's usually like 50% of the language

1:16:55

is about gives you good insights on

1:16:57

the people's culture because this

1:16:59

is about who speaks to who and

1:17:01

how they articulate things and so on and so on.

1:17:04

It's just a mash little. It's

1:17:06

just a GTA exercise but it

1:17:09

was my pastang for a long time. What's

1:17:12

your favourite language or what's your task to

1:17:14

ask very tricky questions? I

1:17:18

mean in college I studied

1:17:20

quite a bit of Chinese and

1:17:22

then I spent some time in China too

1:17:24

and I spoke pretty well at the time.

1:17:28

This was a long time ago. This was like 2006. I

1:17:31

was kind of like a two

1:17:33

hour dinner someone and talking

1:17:35

Chinese and I could like you know. Then

1:17:38

I did not use it at all and

1:17:41

I really enjoyed learning Chinese. It

1:17:44

just found it fun. Like more fun than learning

1:17:46

other languages. And then I

1:17:48

didn't practice at all and I forgot

1:17:50

kind of everything until

1:17:53

like when I was in A

1:17:55

actually when I was in Singapore like in

1:17:57

January I started studying again a little And

1:18:00

so I've been studying, you know, every day for

1:18:02

about, you know, seven, six, seven

1:18:04

weeks. And,

1:18:07

you know, it's coming back kind of

1:18:09

to some extent. Yes, having

1:18:11

Chinese is the one I enjoyed learning the

1:18:13

most in terms

1:18:15

of studying the language. What

1:18:18

I found difficult for you for Chinese and Japanese

1:18:20

was that if you just

1:18:22

really in it and you practice every day, you can

1:18:25

get to really, really good level just reading and writing

1:18:27

as well. But I find that

1:18:29

the spoken and

1:18:32

understanding people speaking in a

1:18:34

general spoken level is not

1:18:36

too difficult to maintain because you can put

1:18:38

shows and stuff inside. But I found that

1:18:41

maintaining reading and writing at

1:18:43

a proper level gets very difficult with, you know,

1:18:45

all the hanzo or the kanji to

1:18:47

maintain. That's something that needs to

1:18:49

be a lot more, a lot more active on my

1:18:52

side. Yeah, reading, I

1:18:54

never managed to get to like, I never

1:18:56

could get to the level where I could

1:18:58

like read a newspaper or something and like

1:19:00

understand a lot. I can send you some

1:19:03

books. There's

1:19:07

a book by someone

1:19:09

called Richardson, he

1:19:13

actually started with a Japanese book. The

1:19:16

legend is that he was in the 70s in

1:19:18

Japan and basically he land like all the 1914

1:19:20

whatever official kanji list in necessary month.

1:19:22

And he used

1:19:26

like a memory palace kind of method. So

1:19:29

he would just deconstruct the root of the kanji and

1:19:31

basically just build around it and build a little story

1:19:33

around it. And it's

1:19:35

just called Remembering Kanji. Actually, there's

1:19:37

a hanzo, there's a Chinese equivalent

1:19:41

that's been same writer, it's

1:19:43

also Richardson. I used

1:19:46

this in 2009 or so. It's

1:19:49

all the positives and the downsides

1:19:51

of memory palace and when you build a story

1:19:53

around remembering something, it's kind of

1:19:55

slow. But you

1:19:57

actually remember it. So it's kind of slow to recall.

1:20:01

like finding the whole story around, oh there's

1:20:03

a ladder, oh there's something, there's a moon,

1:20:05

there's a sun there, there's something and then

1:20:07

you can rebuild the kanji or the hadza

1:20:10

but it's kind of so so it's

1:20:13

useful in like a first kind of

1:20:15

practice and then you kind of forget

1:20:17

about the story and you just like it goes into

1:20:19

the long-term memory and just once you read it you

1:20:21

just you just read the right signs as it is

1:20:24

but if you have a bit of a barrier to remember them

1:20:26

it's quite it's quite useful. I'll send

1:20:29

you the reference. Cool, well

1:20:31

thanks so much Yo, it's really fun

1:20:33

daddy on, we enjoyed our conversation. Thank

1:20:36

you same bit same bit. Thanks

1:20:39

so much for listening for tuning in and we'll be back

1:20:41

next week. Thank

1:20:44

you for joining us on this week's episode. We

1:20:46

release new episodes every week. You

1:20:48

can find and subscribe to the show

1:20:50

on iTunes, Spotify, YouTube, SoundCloud or wherever

1:20:52

you listen to podcasts and if

1:20:54

you have a Google Home or Alexa device you can tell

1:20:57

it to listen to the latest episode of the epicenter podcast.

1:20:59

Go to epicenter.tv slash subscribe for a full list

1:21:02

of places where you can watch and listen and

1:21:04

while you're there be sure to sign up for the newsletter

1:21:06

so you get new episodes in your inbox as they're released.

1:21:10

If you want to interact with us, guests or other

1:21:12

podcast listeners you can follow us on Twitter and

1:21:14

please leave us a review on iTunes. It helps me

1:21:16

behind the show and we're always happy to read them. Well

1:21:19

thanks so much and we look forward to being back next week.

1:21:24

Thank you.