This post was last updated on July 11th, 2019 at 06:14 pm
Part 1 Description
Today’s conversation is with Kevin Zhou, head trader at Galois Capital. Galois Capital is a high-volume US-based OTC desk. Some of Kevin’s previous positions include Head of Trading at Kraken, and prior to that, Head of Trading at Buttercoin, one of the very first crypto asset companies backed by Y Combinator. Kevin has been trading crypto since 2011.
In this conversation, you’ll get a behind the scenes look into how OTC desks operate and make money. The world of OTC trading can be opaque. Although the term OTC desk is often used, there’s much less talk about what happens behind the scenes at regulated high-volume desks. I’ve tried for some time now to get OTC desk data on our platform and it’s been difficult.
Given the reach of this podcast, I’m guessing that I could have done this interview with just about any of the top desks, but I decided to do this with Kevin because he was, frankly, the one willing to share the most information, and because he’s brilliant when it comes to breaking down esoteric financial processes.
Part 2 Description
Continuing from the first part of the conversation, in this episode, you’ll learn more about how OTC desks grow their businesses and add customers, how OTC desks differentiate themselves from one another, and what sales operations, regulations, and pricing look like for an OTC desk.
The Nomics API offers squeaky clean and normalized primary source trade data offered through fast and modern endpoints. Instead of having to integrate with several exchange APIs of varying quality, you can get everything through one screaming fast fire hose. If you found that you or your developer have to spend too much time cleaning up and maintaining datasets, instead of identifying opportunities, or if you’re tired of interpolated data and want raw primary source trades delivered simply and consistently with top-notch support in SLAs, then check us out here.
Topics Discussed In These Episodes
- Galois Capital’s origin story
- What an OTC desk is and what kind of problems they solve
- What differentiates OTC desks from other exchanges
- Whether the desk itself is the counter-party to every trade
- How fees and pricing work on OTC desks
- The origin of the term “over-the-counter”
- What an OTC desk trade looks like
- Leaning down and leaning up
- How much time before a quote expires
- Who is present while an agreement is being made and what it looks like
- How traders are compensated and incentivized
- The fail-safes used at Galois to ensure that the other party is authorized to make trades
- Settlement speeds
- The ideal hire for Galois Capital
- The origin and history of crypto OTC desks
- Business models that are paired with OTC desks
- How OTC desks match up with traditional exchanges
- Strategies for minimizing information leakage around a large liquidation
- How OTC desks grow a book of business
- Who has the responsibility for growing the customer base
- Why OTC desks are reluctant to share data
- Different kinds of OTC desks
- Why some claiming to be OTC desks are really not
- What clients are looking for when they look for OTC desks
- How Galois can differentiate itself from other well-known OTC desks
- Hedging costs
- What sales operations look like for OTC desks
- Some differences between Asian and European OTC desks and American OTC desks
- How OTC desks are regulated in the US
- How Galois Capital is registered and domiciled
- How OTC desks are going to evolve in the future
- The ad hoc mechanics around these price feeds that OTC desks are sometimes providing to privileged counterparties
- What makes a counterparty toxic
Links Relevant To These Episodes
- Cryptoinvestor Weekly Newsletter
- Clay Collins
- Kevin Zhou
- Galois Capital
- Y Combinator
- Genesis Trading
- Digital Currency Group
- Cumberland Mining
Part 1 Transcript
Clay: Welcome to Flippening, the first and original podcast for full time professional and institutional crypto investors. I’m your host Clay Collins. Each week, we discuss the cryptocurrency economy, new investment strategies for maximizing returns and storage from the frontlines of financial disruption. Go to flippening.com to join our newsletter for cryptocurrency investors and find out just why this podcast is called Flippening.
Female: Clay Collins is the CEO Nomics. All opinions expressed by Clay and podcasts guests are solely their own opinion and [00:00:30] do not reflect the opinion of Nomics or any other company. This podcast is for informational and entertainment purposes only and should not be relied upon as a basis for investment decisions.
Clay: Hey, this is Clay. We’re just about to get into this conversation but before we do that, I wanted to provide two announcements regarding nomics.com, a company that produces and funds this podcast. The first announcement is that we just hired Maya, the new editor for the blog. If you haven’t been to blog.nomics.com for a while, go ahead and check it out. We began a project there [00:01:00] where we’re posting case studies of folks using the Nomics API.
The first case study is with Bletchley Indexes who recently switched from crypto compare to the Nomics API for their market data and for pricing their indices. Anyway, if you’re doing exciting things with our API that you think the world should know about, then we’d love to hear from you. Reach out to us via contactnomics.com to let us know what you’re working on and why it would make a great case study for our blog.
Announcement number two is that we just launched our exchanges index page at [00:01:30] nomics.com. You can check it out by going to nomics.com/exchanges. That page shows the number of trades executed across a number of exchanges, global exchange volume which is a chart that to my knowledge nobody else has and much more. Check it out at nomics.com/exchanges when you have a chance. Now to our regularly scheduled program.
Today’s conversation is with Kevin Zhou head trader at Galois Capital which is a high volume US based OTC desk. [00:02:00] He was most recently head of trading a Kraken, previous to that he was head of trading at Buttercoin which is one of the very first crypto asset companies backed by why Y Combinator. Kevin has been trading crypto assets since 2011. He also has various fancy academic degrees. This conversation is a behind the scenes look at OTC desks, how they operate and how they make money.
The term OTC desk is often thrown around but there’s very little talk about what happens behind the scenes at regulated high volume desks. In fact, the world of OTC trading is pretty opaque. [00:02:30] I know this because for some time now at Nomics, I’ve been trying to get OTC desk data on our platform and it’s been difficult. Given the reach of this podcast, I’m guessing that I could have done this interview with just about any of the top desks but I decided to do this with Kevin because he was frankly the one willing to share the most information and because he’s brilliant when it comes to breaking down esoteric financial processes.
This podcast is broken into seven chapters. In chapter one, we discuss what OTC desks fundamentally are, the problems they solve, why they exist. [00:03:00] In chapter two, we walk through an OTC desk trade from both the point of view of an OTC does client and the OTC desk. This is one of the most fascinating parts of this entire interview and Kevin is surprisingly transparent during this segment. In chapter three, we put on our historian hats and discuss the history of OTC desks and OTC trading and the origins of this business model. In chapter four, we discuss the components and attributes of OTC desks. In this chapter, we peel back the curtain on the core components [00:03:30] of OTC trading.
In chapter five, we discuss the kind of OTC desk that can’t exist, the different types of institutions they create them and their various motivations. In chapter six, we discuss control and regulation both in the US and globally. And in chapter seven, our final chapter, we discuss what the future holds for OTC desk crypto trading and the business model. I should stop for a moment to tell you that this episode is made possible by the support of the Nomics API which my company makes. Here’s a little bit about that.
[00:04:00] Do you need an enterprise grade crypto market data API for your fund, smart contract or app? Nomics is an API first crypto asset market data company. Our product is a developer tool for funds, family offices, enterprise and FinTech developers. Our API enables programmatic access to clean, normalize and gapless primary source trade data across a number of cryptocurrency exchanges instead of having to integrate with multiple exchange API’s of varying quality, you can get everything through one screaming fast fire hose. [00:04:30] If you found that you or your developer have to spend too much time cleaning up and maintaining data sets instead of identifying opportunities, or if you’re tired of interpolated data and want raw primary source trades delivered simply and consistently with top notch support and SLAs then check us out.
One notable difference between the Nomics API and competition is that the Nomics product has no rate limits. If you’re an enterprise or found that requires air rate up time and accuracy guarantees along with email and [00:05:00] phone support, then consider using the Nomics crypto market data API. Go to nomicsapi.com to learn more. Okay let’s get this started. Here without further adieu is part one of our conversation with Kevin Zhou, former head of trading at Kraken and current head of trading and founder at Galois capital.
Kevin thanks for joining us today, can you tell us a little bit about the origin story of Galois capital and what compelled you [00:05:30] to start the company.
Kevin: With Galois Capital, it was sort of an idea that was brewing between me my cofounder his name is Mael. We went to school together back in the day studying financials engineering, getting our masters at Berkeley. We always kind of wanted to start a company together and we never found really the right timing. After school, I went directly into crypto and he went into data science but another small startup. I worked in crypto first at Buttercoin for two years. [00:06:00] It’s a very small exchange. I think it was the second company to be backed by YC that was a crypto company, the first being coin based.
I worked there for about two years running their OTC desk. Afterwards when things weren’t going so well this is sort of like after 2013, there was a huge air cycle and it was just a very brutal winter I think for all of us and many companies went belly up, including Buttercoin. That didn’t pan out very well but through that experience, it helped me get a job at Kraken. [00:06:30] I ran the trading desk for Kraken for another two years. And then finally, this is around end of Q1 2017, so last year. We felt like the timing was right and I felt like I’ve been trading in crypto for these two exchanges running their desks for two years each and then even before that for myself personally for another two years.
At that point I thought it’s about time to go straight off on my own. [00:07:00] I reached out to Mael and for him it was getting to that point too. He’s built a lot of experience building out the architecture and sort of backend systems and chain living algorithms for his data science company, his data science startup and we thought it was about time to go set up a shop. So that’s sort of the genesis of Galois Capital. Me and him, we both studied math and undergrad so I guess in some ways to pay homage to some of the great thinkers of the past. [00:07:30] We wanted to name the company after a mathematician.
There were a couple of names that were kind of being thrown around like […] or Julia, Galois. In the end, we settled with Galois. We started the company and I also think from another perspective, the name of the company is the first thing that anybody ever knows about a company. So we wanted to sort of attract the right talent and we’re both quants by background and we definitely think that crypto was [00:08:00] moving more towards algo trading. Now the game only gets harder, more electronic and faster.
So we really wanted to set up for the future there and we wanted to attract people who have an appreciation for very high standards of rigor as you find in math especially when you’re doing roofs. Part of it is just functioning as a light recruiting tool and we have we have been approached by some folks who really like the name. So I guess it’s worked to [00:08:30] some degree and that’s ultimately the genesis of the company.
Clay: With that background established, let’s jump to chapter one of our exploration of crypto OTC desks. Can you tell us a little bit about what an OTC desk is in sort of maybe the most primitive terms possible and what problems they solve regardless of whether or not they’re crypto, OTC desk or other kinds of OTC desks.
Kevin: I think to start, [00:09:00] OTC stands for Over the Counter. What it basically is, is that sometimes some kind of investor or trader, they need to move a really large size or a really big block trade in a particular asset or in their position. Normally when you would do that on the screens or in the lit venues or basically in the order books, you’re going to subject yourself to a lot of slippage. I’ll explain that a little bit. Slippage is this idea [00:09:30] that if you buy $1 of Bitcoin, you’re probably just going to hit the best offer, the lowest offer. If you buy you know $100 million worth of Bitcoin, the best offer is not enough. You have to go down the order book, you have to start lifting more and more offers at less and less favorable prices.
So ultimately when you get that order executed, your blended average price on that execution is much worse than if you were to just trade $1, [00:10:00] if you’re just buying $1 worth of Bitcoin. So the idea of slippage is that the larger the size that you’re doing, the worst the average blended price you’re going to get is on that block.
Since people trading these large blocks face this problem, OTC desk are there to service this issue and hopefully reduce cost, reduce slippage for this trader or investor. So the idea is that when you trade with an OTC desk, the desk functions in some way is like a dealer. [00:10:30] What I mean by that is a customer would come up to us and they would say, “Hey, I want to buy 1000 Bitcoin, what’s the price.” or they would say, “Let me see quotes for 1000 Bitcoin, let me see 2-way quotes.” so that means they want to see both the bit end and offer 1000 Bitcoin.
So then we would offer them some price, let’s say the price is at $100 and we say 99 bid 101 offer. Effectively, they’re able to get their entire block done at that [00:11:00] price and as long as that price is better than the blended average price they would get slipping through the order books, then the OTC desk would have saved them money. Now there’s a question here why would the OTC desk do this. The OTC desk, they’re selling to them at 101 buying from them at 99 but then they have all this inventory now. They’re either too long or too short in their own inventory.
Part of the job of the OTC desk for their own [00:11:30] balance sheet is to figure out how to get rid of that inventory whether it’s long or short. And to the extent that the counterparty to the desk as the desk itself—let’s put it this way, let’s say that the desk itself has better execution than the counterparty it trades with, then there’s some sort of net value which is shared between both. The counterparty gets better slippage, gets a more favorable price and the OTC desk take some margin.
So [00:12:00] it’s sort of a win-win situation and it’s effectively execution arbitrage. One group executes better than another and there are some savings in between and then some split first person, some split second. So that’s sort of the idea behind OTC desk. OTC desk specializes in managing inventory and getting execution done. We can get into either side of those I think the first is [00:12:30] if you think about it, there’s a lot of different ways of managing inventory, you don’t actually have to go to the market themselves right, the direct markets, the lit venues or the exchanges.
Clay: Hey, this is Clay cutting in here from the editor’s booth to explain what lit venues are. Lit venues are the opposite of dark pools or dark liquidity. Whereas dark trading venues do not display prices at which participants are willing to trade, lit pools do show these various bids and offers and different assets. Primary exchanges operate in such a way that available [00:13:00] liquidity is displayed at all times and form the bulk of lit pools available to traders. Exchanges operate in such a way that available liquidity is displayed at all times and form the bulk of the lit pools available to traders. Dark pools or dark venues obviously don’t display this order book data.
Kevin: You know first what you could do is you could use multiple exchanges. You have access to all the different exchanges in the world. You have some smart routing on your systems and anytime you need to hedge off risk [00:13:30] then you sort of pro route to distribute the order out to multiple exchanges. And then maybe have some kind of scheduling. Maybe you just don’t get rid of it all at once. Rather, you slowly massage into the market. You want to minimize market impact as much as possible.
You do you carry some risks, some inventory risk during that time but maybe your market impact, your slippage is going to be better because you do it a little bit slowly and less aggressively into the market over time. So those are just some things that you can do just [00:14:00] on the exchanges and lit venues to know hedge your exposure. But there’s a couple other ways to manage inventory. One way is that, there’s a whole bunch of other OTC desks and sometimes you can share risk with them.
So let’s say the counterparty buys 1000 Bitcoin from you, maybe you can say, “Well, I’m going to buy 500 Bitcoin from another OTC desk.” The counterparty pays you, you pay the other OTC desk, [00:14:30] you have little bit less risk on your books and overall, if there was another OTC desk, you could also fit the other 500 to them or you can just work the other 500 yourself on the market. Definitely sharing flow with other OTC desks is a way to get rid of it.
The last way to do it is to warehouse the risk and wait for matching counterflow. Counterflow is the most valuable I think with OTC desk. Let’s say Alice comes in and buys 100 Bitcoin [00:15:00] from us, and then Bob comes in half an hour later and he sells us 100 Bitcoin. Well then we’ve just naturally balanced ourselves. We carried some inventory risk for half an hour. In doing so, we avoided the need to have any hedging cost whatsoever. we make some profit on the spread with Alice and we make some profit on the spread with Bob rather than making profit trading with Alice and then losing that or [00:15:30] at least a part of that in the hedging process and getting rid of exposure. So those are the three main ways I would say of managing inventory and there’s a lot of new ones that goes into each one.
Clay: I think that’s super fascinating. Let’s talk a little bit about some more of the basics here. So what are some of the things that differentiates OTC desk from other exchanges? You explained how it solves the slippage problem where on most exchanges the more you buy, the more the price goes up [00:16:00] and the more you sell, the more the price goes down. So it solves that problem. It’s also the case that with OTC desks that the desk itself is the counterparty to every trade, is that more less true?
Kevin: Yeah, that’s exactly right. At least that’s the case for most desks. I should say that there are a few exceptions and it really comes down to that there’s two models for OTC desk. There’s the agency model and then there’s the dealership model. What I’ve been describing [00:16:30] is the dealership model and that is the OTC desk takes principal risk and extends their own balance sheet to facilitate liquidity for the counterparty.
The agency model is more this thing where Clay you come to me and you say, “I want to sell 100 Bitcoin. Here’s my 100 Bitcoin, get rid of it at as best as you can and I’ll pay you a commission for It.” there’s a lot of thought around which is sort of better. The way I think about it is purely from a game theoretic standpoint. If an [00:17:00] agency basis work with me and you’re going to pay me a commission no matter how well I do and I don’t have much of an incentive to execute very well for you.
If I execute badly for you and there’s a massive slippage then I still get paid. If I execute very well for you, I still get paid the same amount. So there’s no real incentive there. This is very different I think from the dealership model because in that model, me and you we agree to a price and then it’s my problem. [00:17:30] I have a huge incentive to make sure that I’m doing execution well for myself because it’s become my problem right. At least for you, you have some guarantee on the price.
You know that the market’s trading at 100 but you’re paying 101. So you know exactly what you’re paying. You’re paying more than what the market is currently at the spot price, it’s less than the slippage you would have incurred just going through GDAX and then there’s no conflicts of interest between me as the desk and you as the counterparty. I strongly [00:18:00] prefer the dealership model but I think there are a few that run on an agency basis to.
Clay: That’s helpful, that distinction. are there also generally fees involved or is it the case that because of your expertise in trade execution the agreed upon price is the agreed upon price and you’re pricing in whatever profit you need to make or upside you need to make off the trade into quote.
Kevin: Yeah, [00:18:30] that’s exactly right. For the dealership model, there are no fees it’s just everything is sort of baked into the spread itself and the spread is the combination of two things. One is our own hedging costs and two is our profit. So certain desks, they want to take a little bit more on one side or the other, it really depends. I would say overall, things are very competitive and profit margins have gone squeezed quite a bit. I would say in the last six months, margins have gone down [00:19:00] by a factor of two. The space is definitely getting very competitive but I think that’s good for the counter parties and it’s good for the consumer.
Clay: Do you know the origin of the term over the counter? I kind of scratch my head sometimes because it’s just sort of an odd way to describe the business model.
Kevin: It definitely came from the traditional markets. I’m guessing there was some counter and the parties would just negotiate over some kind of counter, maybe in the pits or something like that.
Clay: Let’s transition [00:19:30] here to chapter two to walk through what an OTC desk trade would look like maybe starting from the point of view of someone who’s calling you up to maybe they want to sell or they want to buy some Bitcoin. Let’s take this in sort of many micro steps as possible. Maybe I heard about you from someone, someone recommended you, I reach out to you via email and say, “Hey, I’m interested in making [00:20:00] this kind of trade.” how do these trades typically get initiated and for someone who’s never done business with you before, what would the next steps be?
Kevin: I think the first thing is, we like to hop on a call with them and just make sure that everybody setting expectations understanding what both side’s obligations are and sort of what to expect. That initial call will gather some information such as if they’re okay trading in US dollar, [00:20:30] if they want fiat, or if they want some kind of other currency. We can’t do RMB and we can’t do Yen, we can only do US dollar. So we want to make sure they were able to trade with them in the first place.
The second thing obviously is on the crypto side. We do trade a lot of different assets but we don’t trade all of them and we certainly don’t trade on listed assets and liquid assets. At least the ones with no main net, there’s no way we can trade them. So we don’t trade in SAX or any of those sort of pre-launched ICOs. After we [00:21:00] get through all of that and let’s say that we can do trades together, then we’ll put them through our onboarding process and there’s really two steps so that, the first is to collect some KYC information from them.
So we are a US company and we have to abide by the KYC rules. We collect them, very common things if it’s an entity, we’ll collect things articles in the corporation, operating agreements, passports or driver’s licenses from the directors of the company [00:21:30] and authorized traders. Very standard stuff, if it’s an individual, just personally identifying information. Once we have all of that, we run some checks on it, everything looks good. Then there’s an OTC agreement to sign.
The reason that there needs to be this kind of agreement is because one of the main differences for how OTC desk work is that everything is post trade settled and what I mean by that is that we agree on a trade and we [00:22:00] book it and that’s final and then we enter into settlement. That’s different from the exchanges where you would need to deposit your dollars first before you’re able to make trades with it to buy a Bitcoin. Because of that, there’s some extra risk for us. As a desk, we don’t know really who the other site is I mean, we know them to some degree based on information. We don’t know if they’re credit worthy, if they’re going to honor their trades if they’re ever going to renege.
What if their business is just really terrible they enter into default [00:22:30] and it wasn’t even their intention to renege but it is a terrible business so they enter into default. Because of those considerations, there’s a lot of counterparty risk involved and want to make sure that legally both sides are obligated to their leg of trade and that’s what that OTC agreement is. It’s to ensure that once the is trade consummated and what it means for a trade to be consummated then both sides are legally bound to those terms.
Once we’re sorted with all of that, we can turn around paperwork pretty quickly, I would say that we’ve kind of streamlined that process pretty well [00:23:00] and it usually takes us less than 24 hours from receiving all the information from the counterparty to being able to turn it over get our first trade done. One of the things that we also ask that the counterparty to provide is preferred communications channel.
We’re very flexible, we’re comfortable doing signal, telegram, WhatsApp, WeChat, Skype even Google Hangouts, I don’t recommend it. I don’t even really recommend Skype. But we’re very comfortable with however they like, [00:23:30] we’re open to doing it with them. And then for whatever channel they prefer, we’ll open a channel, I’ll bring in all my traders, they’ll bring in their people that are authorized to trade and then they can request a quote from us at any time. That can be a 2-way quote or it can be 1-way quote. I’ll say as a reason for why people ask for 2-way quotes, there’s really two reasons.
The first is not to reveal any information [00:24:00] about what direction they actually want to go. So I think that there’s some value there and the second is just to see how tight sort of the market make or the OTC desk really is. With a single quote, you know that it’s some ways away from spot but you don’t have any reference to how wide the actual spread is.
Maybe the OTC desk is leaning, maybe they’re leaning down on their quotes, maybe they’re leaning up on their quotes based on their own inventory. So you don’t really have a sense for how competitive they are [00:24:30] with maybe other OTC desks. So seeing a two-sided quote would be good for that. So they’ll either ask us for a one sided-quote or two-sided quote. We’ll give them a quote, we’ll see what we can do for them there.
Clay: And just for clarification, the leaning down or leaning up that’s let’s say you purchased a bunch of Bitcoin and you’re holding on to some inventory that you acquired relatively cheaply and you’d still have a whole lot of upside if you sold it for a little bit [00:25:00] below the current spot price, that would be leaning down.
Kevin: Yeah. One way you can lean down is you can lean it not so heavily that it’s crossing the spot price. The other is that you’re leaning so heavily that you’re crossing. I’ll just give you an example, let’s say like when I’m neutral I’m 99-101, 99 bid 101 offer. Leading down could be like 98.5-100.5. That could still be considered leaning down. But if I have [00:25:30] very severe inventory imbalances, then yeah, I could go like 96-99, I could do that. I think it all depends on how each desk manages their inventory, just how badly misaligned their inventory is from their intended position of their inventory. A lot of that is art and some of it is science.
Clay: Contracts submitted and quote is requested, how long is the expiration date on the quote usually? Is it 24 hours or is it custom for each quote, [00:26:00] or is there sort of a standard expiration date for these quotes?
Kevin: Usually the quote is good for about 30 seconds but it depends on the party. You know what’s funny is, you laughed but I actually think 30 seconds is very generous. I think it’s actually very long to be honest. There are places that offer 10-second quotes. I think it really depends if we think that the counterparty is more on the toxic side, [00:26:30] we would probably cut it off at about five to 10 seconds.
If the market’s moving very rapidly, we can always pull out quotes. So we can say, “Hey, we’re off.” so the slang is that we’re off, that means we’re off the quotes, ask for a requote whenever you want. It’s the sort of thing where if you think about it, when I give you a quote, I’m effectively outlying a very short term free option to you. I’m effectively underwriting a call a put but a really short one. Like a 5-second [00:27:00] caller put for you. The optionality has some value, maybe most of the time there’s nothing to be done. But every once in a while, some freak accident happens like I quote you at [1:00] PM exact and then at [1:00] PM and five seconds the market is just completely ripped or completely crashes.
If it’s in your favor, then you take the quote. If it’s against your favorite, then you pass on the quote. Then I’ve effectively just lost money to you for just outlying a free option for no reason. [00:27:30] I think you never want to outlay a free option for too long but obviously, you have to take into consideration that just from a human element that people need time to react and actually think about the quote and whatnot. So I think 30 seconds is pretty fair. They have two options after that.
First is that they can pass and no harm done, both sides move on with business and they can come back any time and people pass for several reasons. Usually they pass because [00:28:00] there’s either a quote that’s better from another OTC desk or there is no better quote but they think the quote is still too wide and they just don’t want to take it. That’s perfectly fine, I think we’d like to price things in a way such that if people walk away and we’re okay with it too. Because even though we probably didn’t make money there, we still didn’t take on much risk either. So there’s always some trade off there.
If we’re fairly indifferent as to whether or not our customers or counterparties are taking our orders, then that means we’re pricing things [00:28:30] pretty fairly, that’s one option. The other option is that they think the quote is good and they accept it and then they say, “Done.” and once they say the word done, that trade is just done. We will send out a trade confirmation with the trade details, information on settlement, all of that stuff.
But the trade confirmation is not when the trade is constant. When they say done, the trade is done, and there’s no going back on it. If the market moves in our favor, it moves in their favor, it doesn’t matter. [00:29:00] It doesn’t matter who ends up winning or losing after the fact, the trade is done when it’s done.
Clay: I have a whole bunch of questions around this. Can you take us to the moment of quotation. So you mentioned signal. So let’s say this is being done in a chat environment. Is it usually that you guys like, an appointment will be set to sort of work this out in real time? Obviously, if a quote is expiring in 30 seconds or even in the potentially toxic counterparty, [00:29:30] less time than that, I imagine it’s someone on your side with a whole bunch of information in front of them and someone on the other side with a whole bunch of information in front of them.
Do you have a countdown timer or are you using the expiring messages function in signal that sort of deletes? How are you actually tracking this in real time? And then when it comes to the word done, is there something in your contract that literally says that if someone says, “Done.” they cannot [00:30:00] renege on that, that it is effectively a point of no return or is that just part of the culture of OTC desk in trading and that’s just implicitly acknowledged?
Kevin: Yeah, so let me address that first. I’m actually not sure exactly that’s written in the contract but it might be actually. I do know that at the very least in the contract it says that when a trade is agreed upon and a communications channel, then it is completely binding. I think it is the culture though in OTC, you either say confirmed or you say done. [00:30:30] In some ways it’s kind of conversational so sometimes not everybody knows the protocol. Sometimes people just goes for a while, they had to use the bathroom and they missed the quote and asks you to refer for requite.
So all that stuff happens, I would say that we’re fairly strict about it because we want there to be some kind of record of somebody saying that something is done. So when someone says we give them a quote and then they say, “Okay.” I don’t know what okay means. Does okay mean that you’ve received my quote and you understand that [00:31:00] that’s the price I’m offering you or does okay mean, let’s do it. Okay is very ambiguous. That’s why I think the word done is chosen. Done is done, it’s unambiguous, the trade is either done or not done.
Clay: So in that case would you write back to them and say, “Hey, I’m not sure what that means. Can you say done?” or do you ask them to say done?
Kevin: We do actually. It’s kind of silly but we actually do. so when they say okay, we go, “So done?” and then they say, “Okay, done.” [00:31:30] we don’t do yes, we don’t do okay, we don’t do fine, we do done. Done and confirmed are the two acceptable responses to consummate a trade.
Clay: That’s the opposite of a safe word.
Kevin: Yeah, it’s the opposite of a safe word, it’s the danger word. If they get it done then it needs to be one of those words. It’s for a good reason. Most of the time there’s not going to be a problem, 99% of time everybody’s on the same page but every once in a while [00:32:00] there’s just going to be huge disconnect. One person’s going to think one thing, one person’s going to think another then you end up into getting into disputes. And then there’s one person has a profit, one person has a loss and should it then be shared or how is it resolved. It just gets messy.
So just to avoid that all together and sort of preempt all of the confusion that has happened in the past when things were ambiguous, I think the culture that’s developed around being very strict about what is done and what is not done is a good thing. What’s also [00:32:30] interesting is that a lot of times, there’s also confirmation for things not being done. When I say, “Okay, I’m 101 offered.” and then you say, “Pass.” then I say, “Okay, nothing done.” and it’s just to say that both sides have acknowledged that they’re on the same page. You passed so you’re on that page, I’m on that age too. I’ve received your pass and I’ve confirmed with you that nothing was done. It’s a courtesy to make sure that neither side is confused, [00:33:00] I would say.
Clay: I actually think this is a really interesting aspect of OTC desk culture and so, I’m going to go a little bit deeper down this rabbit hole. So done is pretty much the only word you’ll accept. If they don’t use that word, you’ll ask them to use it. What about the equivalent of the word done in other languages or do you only interact about these trades in English just to ensure that there’s no ambiguity?
Kevin: Yeah, we only interface in English unfortunately. [00:33:30] I know that there is a very vibrant Chinese OTC market but as of right now, we don’t trade any RMB. Unfortunately, I think there’s only one person in our company that have a basic grasp of speaking Chinese let alone reading Chinese. So it’ll be difficult. It’s ironic because many of us are ethnically Chinese but most of us don’t speak Chinese or Mandarin or Cantonese at all. That’s a bit unfortunate and I’m a little bit embarrassed but yeah, we only deal in English.
Clay: Let’s talk about sort of the [00:34:00] set up or what it actually looks like on your side and on the counterparty side when these things are being agreed to. Do you set up an appointment to sort of agree to a trade, who is present from your side, who’s usually present on their side in most scenarios. How does this actually look?
Kevin: There’s never any appointments. I mean I guess you could make them on an ad hoc basis but for the most part, we operate the desk 20 hours a day. So we don’t operate [2:00] AM to [00:34:30] [6:00] AM Pacific Time and we operate seven days a week. So you can just ping the desk any time basically and then they’ll give you a quote. We’re always ready to provide quotes. It’s not something that needs to be scheduled.
Clay: Obviously that means then that you can’t be present for every single one of these trades. How do you go about communicating what is acceptable and what is not acceptable throughout the org or are individuals [00:35:00] sort of responsible for and may be liable for their own performance and that’s how compensation is set up. How does this work? How do you delegate this?
Kevin: I think you really have to hire the right people. I think it’s sort of a cop-out but it’s really true I think probably for every company. I fully trust my trading team to provide quotes, to trade well, to do the hedging, to basically handle all the trading day-to-day operations. It wasn’t always the case, I think for the first [00:35:30] two months or really two and a half months, basically I did all of the trades myself. My cofounder is working on the tech and I was doing all the trades. We brought on these guys end of February this year and early March.
There’s a whole training process, so we go through Mock Trading, I explained to them my thoughts on the market, my opinion intuitions. Even if I’m wrong, it’s sort of the basis that I’ve learned how to do this and there’s sort of the processes involved. It’s extremely and operationally [00:36:00] heavy. I would say that running an OTC desk is 75% operations and accounting a reconciliation and 25% trading.
So there’s a lot to get through but I think after a while and when you see that people are sort of wrapping their heads around it and the quality of their work is starting to get more and more error free and pristine, then you sort of entrust them with greater responsibility. And then slowly and slowly I sort of let go of [00:36:30] control of the OTC desk one piece at a time, there’s actually many pieces to it. Now at this point, we completely run it autonomously. I fully trust them, they’ve actually done very well and their compensation is very heavily tied to the performance of the company.
We’re a very unique company in the sense that, we have eight people and seven of us don’t take any salary. We’re extremely aligned in the sense that if we’re all in a ship together, [00:37:00] if it sinks, then it’s really terrible for all of us, we all drown together. So everybody’s just very motivated and I never have to wonder about people’s intentions, there’s just no politicking whatsoever because we know that we’re all in it together.
Even to some extent operationally and from management perspective, it makes it a lot easier because I don’t have to make sure that people are on task or whatnot. In trading it is, you eat what you kill and I think nobody likes to starve so I don’t have to worry about motivation there. [00:37:30] I don’t think that’s common, that is how we structured our incentives. So I guess long story short, they run the entire operations that I fully trust.
Clay: No base salary, 100% bonus eating what you kill, performance payouts.
Kevin: Exactly. That actually goes for some of the developers too. I think that’s not very common at all. Even in trading operations, even improv shops, developers are usually paid a higher base [00:38:00] and they have more of a steady and safer job. It don’t have as much upside, don’t have as much downside. I think for the founding team, I don’t mind sharing the success of the company with them. I think that ultimately, it aligns incentives better.
Clay: I want to make sure I have the timing down right. So you have someone at the trading desk on the other side. Is it usually just one person? Can one person independently agree to a trade on behalf of Galois Capital [00:38:30] or do you usually kind of require two people to be there and kind of both consent before you can agree to something?
Kevin: Generally the trading team, you’re always in the same room with each other until they go home. When they are, they confer with each other. They just keep everyone on the same page but they do have the authority to make trades for the company. Each of the full traders have authority to book trades on behalf of the company.
Clay: Is there something written into a [00:39:00] contract that ties chat handles or identities to the contract holder. In other words, do someone just pops-up in signal and says, “Hey, I represent this company. We want to do this trade.” do you just take them at their word or do you reference some sort of contract and make sure that that handles authorized to act on behalf of the organization.
Kevin: Yeah, there are some legal protections in the contract but we don’t do verification in that kind of way. So I’ll explain sort of what the fail [00:39:30] safes actually are. Once the trade is actually made, we always double confirm settlement instructions through two different communications channels. One being our email and one being the communications channel that we agreed to quote in, or we give the quote and they agreed in. this is to avoid situations where for example, there is somebody who has Galois in their name, on their telegram handle pretending to be one of us then you’ll never get the confirmation. They’ll try and do a trade with you, [00:40:00] you’ll never get a confirmation. So you know that that wasn’t us.
The second reason is because there’s a lot of viruses and Trojans out there that counterparties computers can get infected by this malware where it will replace your clipboard Bitcoin address with one of their Bitcoin addresses. So they’re copying and pasting stuff or they’ll just change it on your screen. You open something up and it’ll say their address actually, because of stuff like that, we always go through two different [00:40:30] communication channels and that’s just to double confirm the settlement instructions. Once the Bitcoins are out, they’re gone.
There’s no taking back, mistakes are permanent. So we’re always very careful about how we do these confirmations. And then also you can benefit from the exploit the you’re talking about which is like, there’s some rogue trader and that sort of thing. Another thing to mention is that people’s phones do get ported space. There’s all sorts of phone porting attacks like T-Mobile or whatnot, [00:41:00] Rising or whatnot. There are certain things that are attached to the phone number, certain things that aren’t.
I think WhatsApp is tied to the phone. There’s a telegram, telegrams are tied to a number. But I don’t know, some are handled a little bit differently than others. There are provisions in the contract that say that it is the counterparty’s responsibility to secure communications channels on their side. So if somebody ports their phone and they make a trade with us, technically [00:41:30] they’re liable for that trade. These are sort of the edge cases that—and I’m glad you brought this up because it’s very important.
These are sort of the edge cases where a lot of the operations, the processes, the design of the contract, all of that and basically in drafting that thing up and in setting expectations with the counterparty, there’s a lot of uncomfortable conversations that need to be had about all the crazy edge cases and corner cases that can happen and then who’s responsible for what.
[00:42:00] I would say that before you ever sign a contract with an OTC desk, read very carefully what is expected of you, what is expected of them and then who shoulders the blame and the loss in the event of event A, in the event of event B and so on and so forth. I think if OTC desk is fair you’ll see that basically in common English, if it’s your fault, it’s your fault. If it’s my fault, it’s my fault and it’s pretty fair basically.
Clay: So there’s two channels that are [00:42:30] required for confirmation of settlement transactions but what about just agreeing to a price for a trade? Is that also required on two channels within 30 seconds or are you okay with that from just one channel?
Kevin: A single channel is fine. A single channel for that is fine.
Clay: Let’s get into time a little bit. So prices agreed upon, how long do they have to send you payments and how long do you have to give them what you promised?
Kevin: The way it works is that we will specify [00:43:00] usually in the quote itself or in the communications channel where we give them quotes how we would like settlement. By default it’s 24 hours, so if we don’t say anything, if you get it to us within 24 hours, it’s fine. The reason that a lot of times we don’t go with the default is because certain coins in assets have a lot more risk than others. If you don’t settle Bitcoin with me, the exposure that I have and I have a balance sheet on Bitcoin, it’s not that bad. But if you sell me some rank like 100 coin and [00:43:30] I want that immediately.
A lot of it is ad hoc negotiated but if there is no negotiation, then 24 hours is fine. Then once we receive the funds, technically we have 24 hours to settle back. But in practice, we settle immediately. What I mean by immediately is as fast as the physical frictions in the world will allow. So if it’s Bitcoin, we’ll send it out immediately again in the next 10 minutes or 20 minutes, there will be a confirmation, it’ll be fast. If it’s a wire, [00:44:00] then for domestic wires if it’s before the wire cut off which is [2:00] PM Pacific time, you’ll get it the same day. If it’s after the wire cut off time, you’ll get it the next day.
If it’s an international wire but still in USD, our wire cut off time is [1:00] PM Pacific, if it’s before the wire cut off, it’s one day. If it’s after the wire cut off, it’s two days. That’s pretty much what the settlement speeds look like. Obviously there’s a smaller nuances like certain chains are just a lot faster and others. But for the most part of crypto, [00:44:30] we do it instantaneously.
Clay: The process is always that they pay you and then you pay them, is that right?
Kevin: Yeah, we always take aim at first. We’re not comfortable shouldering counterparty rates for all of our counterparties. It would be too difficult. I would say that the larger the institution they are, the more we would be comfortable maybe finding some kind of other arrangement. Even then I would say the best way to go about it is first just to do a small trade. Buildup sort of [00:45:00] history of doing business. Why don’t we start with something like $50,000 and then work our way up to a couple of mil.
Clay: Great. I think we covered everything that I can think of here. Is there anything else that A, you can think of that you think we should discuss in the second chapter or B, that maybe is notable for how this works with crypto, OTC desk versus maybe more traditional OTC desks when it comes to this end-to-end process for getting a quote and [00:45:30] then executing the trade and finishing things up.
Kevin: I think for the most part, that covers about everything. That was pretty much the in depth full process. I guess there’s more to do on our side, this is from the counterparty’s perspective. But obviously for us, there’s like a whole sweeter stuff that we have to do. There’s like what to do in the accounting, the reconciliation, what to do with the hedging, all of that stuff. And daily PNL, we have NAV reports, we have a benchmark, internal benchmark to fund, [00:46:00] investor letters, all sorts of random stuff too. But not too relevant to the counterparty.
Clay: It seems like the ideal hire for you is not an adrenaline junkie, ideal junkie but a very analytical, process oriented, rational almost OCD individual when it comes to execution, is that correct?
Kevin: Yeah, it’s definitely the case but I would also say that we like to hire people who are not ultra specialized I would say [00:46:30] because there’s a lot to be done. Right now with the trading team, they handle a lot of things, they handle accounting, they handle trading, they handle a little bit on the sales side. In many ways, they handle the trading operations end-to-end and each single one of them can do all of it and 10.
So there’s redundancy built into it too. I think it’s good because eventually we will go toward specialization and as we become a bigger company, we’re going to hire specialists just to do one particular thing do extremely well. I think in some ways, [00:47:00] having some kind of sort of founding group that does know every piece of the business, maybe not super well as much as the swatch specialist but the knows how it all fits together, I think is going to be extremely important to how they coordinate between the specialist.
Also in some ways for there not to be any resentment and we always hear about this sort of middle managers and then working for them sucks and you’re working for an idiot or something like that and I think for the head chef to [00:47:30] have worked as a line cook at some point, I think that does breed loyalty. Nobody likes to do accounting, accounting reconciliation is the most painful thing. I’ve done it and my trading team has done it and the people that we bring in, they’ll understand that we’ve all gone through it. I think that helps a lot. I think that builds a more positive culture that if you’re not willing to get your hands dirty, then why should someone else. I think that is a good way to have the culture.
Clay: [00:48:00] Let’s transition to chapter three to put on our story in hats and talk a little bit about the history of OTC desk. Let’s discuss where this is coming from, the origins of this business model etc.
Kevin: I think that sounds good. I would say that I’m much more familiar about crypto OTC and its history than I am about OTC desk in general in the broader markets. I think I was probably still much too young at that time, [00:48:30] well before my time. But I’m happy to talk about in crypto sort of how it started. I think it all started probably with Genesis Trading. It’s one of the DCGs Digital Currency Groups. If I remember correctly, they ran the first OTC desk. At least the first major one. One of their traders was a guy named Bobby Cho.
Bobby at some point decided to leave Genesis and go to it itBit. [00:49:00] It was around this time and to build out their OTC desk. It was around this time that I was at Buttercoin and Josh Lim who is the former trading in treasury at Circle was sort of starting up Circle’s OTC desk. After genesis, it was pretty much Buttercoin, itBit and Circle, as the early players there. At some point I know that they brought in Dan.
He has a really I think Polish last name, I’m going to butcher it, so [00:49:30] it starts with an M. they brought him in at Circle. So Josh and Dan were basically working together running the desk, Bobby was running the desk himself at itBit, I was doing Buttercoin. And then at some point, there was some sort of changing of the guard. I know Bobby eventually left itBit and went to DRW, started an offshoot called Cumberland Mining which is a big OTC desk now. At that time it was sort of a startup and also at some point, Buttercoin [00:50:00] basically fell through and closed down as a company and I went over to Kraken and then I started up that the OTC desk over there.
After some, I forgot exactly how long but I think we’re already at maybe 2016 at this point and then at some point in 2017 Josh leaves Circle, Dan becomes head of trading over there and then he goes to UBS to go back into traditional markets to trade, I think it was either like equities [00:50:30] or like exotic options or something like that, I forget what it was. I was told when he left, because we all know each other from back in the day. In those days I would say just as a side comment, the size of the trades are much smaller than they are today.
Back in the day, I was so happy to just trade $20,000 that’s just a nice walk, we made $50, we made $100. You spend half an hour and made $50, not bad. Very different from how it is today, [00:51:00] the largest clip that I’ve ever heard of at Cross was in a single trade $75 mil. Just the size and the magnitudes of the trades have gone a lot bigger and obviously it’s gotten very lucrative. Nobody’s just picking up $50 anymore. I think there are some folks out there that won’t even take a call for less than a grand. It’s very different from what it used to be.
I told Josh when he was leaving I was joking with him and I said, “Nobody ever leaves crypto, you’re just going to be back at some point.” [00:51:30] and sure enough he’s back. So fast forward like a year to a year and a five or something. He is trader now at Galaxy’s OTC desk. So you basically see the same guys, there’s a lot of new faces now but some of the old guys are still around to do an OTC. Genesis is still operating, Cumberland is still operating, itBit maybe, I don’t even know, maybe they kind of closed down. Buttercoin closed down, Kraken is still operating. I passed on the desk to do [00:52:00] another head of trading and then there was another change of guard over there.
Circle is still trading doing well and then Galaxy is setting up their new OTC desk. So that’s sort of how it looks I would say on the western hemisphere. There’s a lot of smaller OTC desks, they’re plenty all over the place, so much of OTC. I think this is something that’s a little bit kind of a good thing about OTC is that as much about sales as it is about trading because you need to find the counterpart. Because of that, [00:52:30] you can have these smaller competitors come up and start to build their own book of business.
That’s really a nice thing because it also means with more competition at least for the counterparty it’s a nice thing. It means that prices are going to get better and better and spreads are going to get tighter and tighter. Unlike in some ways natural monopolies or natural oligopolies exchanges, it’s just very hard to start a new exchange. It’s just very hard to take market share from incumbents because liquidity effects are so strong on liquidity and trading and flow.
[00:53:00] On sales, a startup can do sales and they can build up their own book of business, they just have to hustle really hard. Some network fix is just less, that’s just a side comment I wanted to make and then outside of that, in Asia there’s a whole bunch of OTC desks. From what I’ve seen, their spreads are anywhere and in Europe too, anywhere from two to three Xs wide as in the US. I don’t know if it’s the case because there’s not enough competition or for some other reason [00:53:30] like I have some theories that they maybe they have some captive flow because a lot of these guys doing OTC desk also do ICO investment and also do market making.
So they provide sort of full on financial services for the companies or the projects that they’re incubating so maybe they’re able to charge much more than normal or maybe it’s just because there’s been such a huge crackdown on exchanges in China that everybody just needs to trade OTC and are willing to pay a bigger premium to do so because liquidity [00:54:00] otherwise is very thin. It could be any of those reasons but for whatever reason, they’re very expensive up there.
Clay: Let’s discuss a little bit motivations for starting an OTC desk. OTC seems to me like less of a core business model and more just a way of executing trades. I often see OTC desks paired with other business models that it serves some function to have an OTC desk. I’ve seen mining [00:54:30] companies have an OTC desk because it would sort of be awesome to do so. You guys are a hedge fund with an OTC desk with a strong quantitative bent. It seems very common to have an OTC desk attached to another business model. Does that seem to be true to you and what are sort of the common types of business models that are paired with at least crypto OTC desks?
Kevin: I would agree to some extent and disagree to some extent. [00:55:00] I think there are a lot of minors with OTC desk. They have a very specific problem but they do need to trade these huge blocks. So having someone internal is a good way of sort of keeping the desk on us externally. They still need to offload but then you have a guy that’s just really thinking about the prices and the sizes are so huge, the notional sums of money are so huge that it’s worth paying a guy just to do that.
For hedge funds, yeah, certainly it helps to have OTC even if you’re a long fund and you want to [00:55:30] rebalance your portfolio or you want to gain or lose more exposure in one aspect or another, then doing OTC is a good way to do it. It’s a good way to supplement alpha if you’re on the quoting side. on those points I definitely agree and I think there’s a lot of other cases where there’s OTC desks with exchanges for example, there’s OTC desk with asset managers, all of that I think makes sense.
But I also think that there are some folks that are just dedicated to doing OTC. For example with Bobby at the [00:56:00] Cumberland, it’s what they specialize in. I think that by itself if you can have enough of a market share, if you have enough flow, it is a good business in of itself and what I mean by that is it’s a business in the sense that you’ve built up a network of relationships that go to you for execution rather than anywhere else. They don’t want to build out their own execution. They’re busy doing other things or planning out their strategies maybe the macro strategies. They don’t want to worry about the micro details of execution.
[00:56:30] They’ll continue to work with you if you’ve built that sort of relationship. I think certainly, people do appreciate the white glove service. Arguably, someone could build that same software that is basically an RFQ system, a Request for Quote system. If you log in and it’s just a bot on the other side and you say, “Hey, I want to buy 100 Bitcoins.” just like with OTC and then bot spits you back with a price and you confirm or not.
Some people do use that and [00:57:00] I actually don’t mind it at all. I think it’s quite good but I think some people don’t care for it. Some people want to be able to ask questions, some people want to just hear another person on the other line, sometimes you just want to talk shop and say, “Hey, what’s going on with the market? What do you think is up with Tether? Why is it trading at 96 cents on a dollar? Is the global bank going bust? What’s going on?” part of the skill of I think OTC is just banter, you just got to be really good at bantering. I’m just joking. [00:57:30] But I think there’s some there’s some quality, there’s some human quality to it that people really do appreciate.
Clay: Yeah and the banter aspect makes sense. it seems like there’s a lot of people out there who would like to know more about what’s happening in the crypto markets and who better to ask questions about that of than someone whose livelihood is 100% dependent upon knowing what’s happening. So to the extent that they can get a little [00:58:00] free advice or to educate themselves, it’s probably worth some basis points if they’re potentially on a trade.
Kevin: Yeah and I’ll also add that a lot of rich people are very old. They don’t want to have to go to a website and login and maybe they don’t use a password manager and they forget their password and how does it work again and got to click through this and that. They just want to call up somebody and say, “Hey, I just want to buy 1000 Bitcoin.” They’re like, “Okay, I got you. I’ll get it done for you.” it’s really that kind of [00:58:00] quality of service.
You just tell someone and then it’s done, you don’t have to worry about it. I find that when you save rich people time, they’re willing to pay for it because time is extremely valuable to them. So they just don’t want to deal with the hassle, they don’t want to be thinking about it or worrying about it, they just want to know that it was done properly.
Clay: [00:59:00] That concludes part two of this conversation with Kevin Zhou from Galois Capital. In part two which comes next week, we discuss the components or attributes of OTC desks, the range of OTC desk that can exist and the different types of institutions that create them and their various motivations for creating them, control and regulation and what the future holds for OTC crypto trading and the business model. See you next week.
That’s it for this [00:59:30] week, to sign up for our free crypto investing newsletter, listen to other episodes or get the show notes from this episode, please visit flippening.com. I’ll also invite you to check out the startup that funds this podcast, Nomics spelled N-O-M-I-C-S at nomics.com. Finally, if you got value from the show, the biggest thing you can do to help us out is to leave a 5-star review with some comments and feedback on iTunes, Stitcher or wherever you listen to podcasts. Thanks for listening and see you next week.
Part 2 Transcript
Clay: Welcome to Flippening, the first and original podcast for full time professional and institutional crypto investors. I’m your host Clay Collins. Each week, we discuss the cryptocurrency economy, new investment strategies for maximizing returns and storage from the frontlines of financial disruption. Go to flippening.com to join our newsletter for cryptocurrency investors and find out just why this podcast is called Flippening.
Female: [00:00:30] Clay Collins is the CEO of Nomics. All opinions expressed by Clay and podcasts guests are solely their own opinion and do not reflect the opinion of Nomics or any other company. This podcast is for informational and entertainment purposes only and should not be relied upon as the basis for investment decisions.
Hey this is Clay. We’re just about to kick off part two of our in-depth review of OTC desk operations with Kevin Zhou from Galois Capital. But before we do that, I wanted to provide an announcement [00:01:00] regarding Nomics.com, the company the produces and funds this podcast.
The announcement is that we’re now able to provide our enterprise customers with the ability to gather 100 millisecond snapshots of almost all crypto exchanges. To our knowledge, no other crypto market data provide is able to produce data at this fidelity. If you’d like more information about this, please contact us at nomicssupport.com. Now to our regularly scheduled program.
What you are about to hear is part two of [00:01:30] my conversation with Kevin Zhou, head trader at Galois Capital, which is a high-volume US based OTC desk. He was most recently Head of Trading at Kraken. Previous to that, he was Head of Trading at Buttercoin, This conversation is a behind-the-scenes look at OTC desks, how they operate and how they make money. The term OTC desk is often thrown around but there’s very little talk about what happens behind the scenes at regulated high-volume desks.
This two-part podcast is broken down into seven chapters. In the first part of this interview, which you can hear in [00:02:00] episode 30 of this podcast, I highly recommend going and downloading that now and listening to it for some context on this conversation. But in that previous discussion, we talked about what OTC desks are fundamentally are, the problems they solve, and why they exist. We also walk thru an OTC desk trade from the point of view of the OTC desk client and the OTC desk itself. This was one of the most fascinating part of the entire interview and Kevin was surprisingly transparent during the segment.
Finally, we [00:02:30] put on our historian hats and talked about the history of OTC desk and OTC trading and the origins of this business model. In the interview that follows, which is part two of this conversation with Kevin, we discuss one: the components or attributes of OTC desks, two: the range of OTC desks that can exist and the different types of institutions that create them, three: control and regulation, and four: what the future holds for OTC desk crypto trading and the business model.
I should stop for a moment to [00:03:00] tell you that this episode is made possible by the support of the Nomics API which my company produces. Here goes.
Do you need an enterprise-grade crypto market data API for your fund, smart contract, or app? Nomics is an API-first crypto-asset market data company. Our product is a developer tool for funds, family offices, enterprises, and fintech developers.
Our API enables programmatic access to clean, normalized, and gapless primary source trade data across a number of cryptocurrency exchanges. Instead of having to integrate with [00:03:30] multiple exchange APIs of varying quality, you can get everything through one screaming fast fire hose. If you found that you or your developer have to spend too much time cleaning up and maintaining datasets, instead of identifying opportunities, or if you’re tired of interpolated data and want raw, primary source trades delivered simply and consistently with top-notch support and SLAs, then check us out. One notable difference between the Nomics API and competition, is that the Nomics product has no rate limits.
[00:04:00] If you’re an enterprise or fund that requires air rate, uptime, and accuracy guarantees along with email and phone support, then consider using the Nomics Crypto Market Data API. Go to nomicsapi.com to learn more.
Okay, let’s get this started. Here, without further ado, is part two of our conversation with Kevin Zhou, former head of trading at Kraken and current head of trading and founder at Galois Capital.
We start off with chapter four which is a discussion of the components or the attributes to the OTC desks. In this chapter, [00:04:30] we peel back the curtain on the core components of OTC trading.
The purpose of this chapter is to peel back the curtain on the core components of trading. With a traditional exchange, there’s an order book, there’s a matching process, just sorting and matching bids and asks. There’s a custody component. [00:05:00] Are there analogs to that with OTC desks or is it just a lot more ad hoc custom and it varies from desk to desk?
Kevin Zhou: I would say that it’s a bit different. There are no real order books. I would say that really the OTC desk job is to manage its own inventory and outlay quotes to its counterparties. It’s not so much that it ever matches up a buyer with a seller.
Clay Collins: Right, because it’s a one-sided market.
Kevin Zhou: It’s always one sided, right? [00:05:30] And then, to the extent that you have matching counter flow, it’s always offset by some time, right? Technically, it’s just two separate trades. They happen to be close in time and they net each other off. It really just comes down to the inventory management game again.
Kevin Zhou: I would say, though, that every once in a while, you’ll see an IOI or an Indication of Interest for some really illiquid random asset, right? And it’s just like, you know, XYZ coin, right? And it’s a huge sum of it. And then, they come to us and we’re like, “Okay, yeah. When we could take it, but we would quote you this kind of price.” [00:06:00] and they don’t like that price. And then they’re like “Hey, do you know if there could be any private buyers in your network?”
Kevin Zhou: Then you know at those points, we might go on and we might go and ask some of the funds in our network whether or not they’re interested in the other side. But I always make it clear to the counterparty that I can do that for them, and I’m happy to, and especially if they’ve done a lot of business with us. But I’m happy to anyway. But it’s just that if I do, then word is going to be out that somebody is selling, [00:06:30] or somebody is buying such a huge chunk of this illiquid asset, right? Like that information will leak, right? From just from the asking of people if they’re interested, right?
I always make it clear to them is that what they want? Are they okay with that for the potential of finding a match, right? And potentially getting a better offer. Because once it’s out there, we can’t be […]. People gossip a lot in this space. You know, rumors travel very, very, very fast.
Of course, I’ll never disclose who the party is, right? For example, if Ethereum Foundation is [00:07:00] dumping all of their Ether at us, we’re not going to say that they’re dumping all their Ether at us, right? Because that’s a terrible signal and I think it’s just bad customer service. But nevertheless, if we’re shopping around $300 million worth of Ethers, does anybody want to buy? That’s not a key note. It’s like people are going to get spooked. That’s just something to keep in mind, but we do that every once in a while, as a favor to some of our counterparties.
Clay Collins: In terms of word traveling, it’s not necessarily [00:07:30] that you’re going out and running your mouth, it’s that you need to figure out how you’re going to liquidate your position if you want to or share the burden across multiple desks, and that’s how it gets outside of your control and maybe can take on a life of its own.
Kevin Zhou: Oh, yeah, exactly. I would say that it’s like first, you could ask the counterparty “Okay, is there anybody in particular you want me to approach?” If not, [00:08:00] then I’m like—I’m part of this crypto fund working group. Something that I started Summer of last year and I think now we’re up to almost 100 hedge funds in it. That’s a good place to circulate some, but first of all, deal flow but that’s only for the long and venture style funds, but also a place to shop around some of these illiquid assets. Somebody might just be on the other side.
I tell them, “Well, I’m happy to circulate that around this group of funds and then we’ll see if there’s any interest.” [00:08:30] But I always tell them that, “Some of these guys may be holding the same direction as you right now. And the fact that you’re shopping it around, they’ll just dump it. That could be the case, you know? I just want to warn you on that.” But it’s really up to them and we really let them decide and make those business decisions.
The way that I think about it is, as an OTC desk effectively, you know how there’s always—and I don’t think it’s this analogy—but you know how there’s this analogy where it’s like, “During a [00:09:00] gold rush, do you sell shovels or do you look for gold?” I don’t we’re shovel sellers. I think like pure rent seekers like exchanges, like some of the tools that you can use online for trading. I think a lot of that, those are like shovels, right?
I think what we are, we’re like the strong guys that you hire to go dig. It’s like, “We’ll dig for you. We don’t know if there’s gold there. Maybe we don’t even think there’s gold there. But if you want us to dig there and you pay us, then we’ll go dig there for you. We just really [00:09:30] specialize in digging. We’re just get really, really good at digging. You provide us with the shovels and you tell us where to dig, we’ll dig for you.” That’s how I think of OTC does in the mining gold rush metaphor.
Clay Collins: Let’s pretend you’re the Ethereum Foundation or like maybe say an organization that is less visible. There’s fewer people watching their addresses and you want to make a big trade. You could move your funds. You could do it directly from a hardwired wallet with a decentralized exchange, [00:10:00] but the markets aren’t very deep, so there’s going to be a ton of slippage. That’s a way of not revealing your intent to liquidate.
You could go to an OTC desk and again, that doesn’t require you to move any funds out of your address when you’re agreeing to a price, but then word can potentially get out. Or you could move your funds to Binance and, of course, people would notice that a good jillion dollar in Eth has just been moved to Binance. [00:10:30] What would be your strategy to minimize information leakage around a huge liquidation like this?
Kevin Zhou: I definitely agree with your point that everybody is watching these addresses and in particular, I think even today people are watching the exchange addresses and they can connect the dots there. I would say at least OTC is a bit safer than exchange because you’ll see the funds leaving, let’s say the Ethereum Foundation, but it’s not going into Binance or any known exchange, right? It’s just going into another OTC desk account.
[00:11:00] I think that’s probably less alarming. But even then, I would say the movement itself is kind of alarming. I think one way to do it, and I’m surprised maybe people already do this, but I don’t think that many people doing it—I’m surprised why they don’t—which is that in anticipation of all of this, why not just start moving the funds randomly? Even before you want to sell. You move it out, you move it back in, ship it around, ship it back. Just get the [00:11:30] market used to the fact that it’s always moving, right?
And then once all that gets priced in, then at some point when you finally move it to sell it, well then, like you don’t spook the market and the market doesn’t front run you, right? Because if you think about it, it’s like terrible. The moment the Ethereum Foundation moves money to Binance, the price of Eth just tanks, they will just never get a great price, right?
It just almost seems like you got to keep it on this kind of like treadmill, and you just got to keep circulating it until finally you want to sell it. Just for in anticipation of selling it.
Clay Collins: [00:12:00] That kind of reminds me of how Monero hard forks every six months. Like even if they don’t need to, it’s just common practice. They need to do it, like it’s not like the whole community’s up in arms. It’s just common practice.
Kevin Zhou: Exactly. And you know, I think with the Monero thing, it’s almost like even if there’s nobody building A6, you don’t even want people to get that idea, right? It’s just like you want to desensitize them from even trying just by showing [00:12:30] you are actually upholding what you said you’re going to do, which is, you know, for better for worse—maybe it’s a bad idea—but for better for worse, you are changing the work algorithm every six months. Then people won’t even invest R&D into trying to do this stuff. It’s a deterrent, yeah.
Clay Collins: If you’re really bullish on something, you could move it, trigger a whole bunch of panic sellers and then like buy some for yourself.
Kevin Zhou: You totally could. I mean, I think the Ethereum Foundation totally could, you know? But do they really need more Ethereum? That’s the other thing, right? You know?
[00:13:00] With anything that sizable that can actually move the market and spook it, they’re probably holding so much that it’s not even thought about trying to make more money because they need to realize what they already have. You know?
Clay Collins: It sounds to be, though, because this is one-sided, rather than two-sided, a huge component of this is kind of traditional sales and marketing. Maybe what we should dive into here in chapter four is how that looks like for a [00:13:30] regulated business. Are you allowed to take out paid media? Is that even a good fit for your potential customer base? Is it more about speaking at conferences, thought leadership? How do you grow your book of business in a sort of strategic and planned way as an OTC desk?
Kevin Zhou: I would say that it’s actually quite difficult because if you think about it, if the minimum trade size is 100 grand, if you just do a very broad marketing, you’re casting such a wide net that most of those folks [00:14:00] can’t trade with you anyway. Right? It’s like, “How do you actually figure out where your customers are? Where your counterparties are?” and I think a lot of it just comes down to just networking in the space, getting intros and referrals from people that you’ve worked with. I think the sales cycles for like a cold call to a new company or a fund to user services, I mean, it’s tough, right?
But if there’s like a warm intro, it makes things infinitely easier, you know? And then even when cold calling does work, I mean, that’s like a long sales cycle. You’re going to kick tires, everybody is [00:14:30] suspicious like, “Why are you approaching me?” all that sort of thing. It’s tough, you know? I think to some level marketing works, but I think you have to market to the right demographics. I think things like going to conferences, having speaking engagements, coming on to podcasts like yours.
With SF Blockchain Week, I actually was just on a panel and the topic was, “Where’s the liquidity?” I was going to make the joke, I didn’t actually make this joke, but I really should have. I should have said, [00:15:00] “Well, it’s not in crypto. It’s like in the equity and Forex markets.” that’s what I should have said.
But anyway, I think just getting sort of the name out there, the brand out there. OTC definitely requires a lot of trust, right? At the end of the day, our counterparties pay us before we pay them, right? We need to have like a track record that is just pristine, right? That when you ask people “Oh, tell me about Galois Capital.” that they’ll say that, “They handle business in a very fair way and they’ll always settle their trades. They never [00:15:30] renege on their obligations, and they’re super-fast on their settle.” Right?
Clay Collins: There can’t be a seed of doubt in terms of counterparty […].
Kevin Zhou: There can’t be a seed of doubt. No, absolutely not. I think, you know, we’re doing the best that we can, and we’re experimenting with ways of getting more counterparties. But I agree with you, it’s not easy. It’s not straightforward at all.
Clay Collins: When you think about growth, is that not just about making profitable trades that grow the business, but in terms of new customer acquisition, when you think about your team, [00:16:00] is it everyone’s responsibility? Is it primarily your responsibility? At the end of the day, who’s responsible for bringing in new customers? And is that the responsibility of your traders or not?
Kevin Zhou: In much larger operations in traditional finance, there’s huge separation of roles, right? You have your sales guy, your execution trade, you have a whole bunch of different folks, you got your ops guys. For us, as a much smaller startup and being in a more nascent industry, and I think for the fact that I want everybody to get experience in handling all the different parts of the funnel, [00:16:30] everybody has to do everything. Our traders, they also do sales, they also bring in business.
I would say that I probably do the most of it, just because I’m probably the most public facing and in some ways, a lot of times people just want to talk to me. It’s just like, “I’m entrusting you with money. I want to talk to the guy that’s responsible.” You know. Then I go to say, “Hi. Yeah, I’m trustworthy. I’ve been doing this for a while. It’s all good.” But they do bring in some business and I think it’s been good. I think it’s the sort of thing [00:17:00] that’s not easily scalable, right? It’s like in terms of turnover on the flow, on the balance sheet, that’s somewhat scalable.
On sales, it’s just not really scalable. You just got to send more people out there because just spending on Google ads, it just isn’t going to work. Facebook ads, it’s just not going to work. You’re just casting too wide a net. You actually need people on the ground going out there and building relationships. I think that we just need to scale up by people. We just need to hire more people in order for that [00:17:30] to really play out better. I think that we have a work ahead of us for that.
But I think right now, I mean, growth isn’t bad. I mean, every month, we get more and more counterparties, and the trading becomes more consistent. The actual size of the trades have gone down recently just because it’s a bear market. If people are still trading like 200 Bitcoins at a time, now 200 Bitcoins are worth less, you know? You know, that’s fine. I mean, that part at least is just correlated with market, right? And if the market goes up, at least there is some kind of like organic growth and [00:18:00] number of counterparties and the consistency that they trade.
But I think that this kind of business is not really like a venture business, right? It’s not really like where you get exponential growth where you have some widely adopted and viral consumer product, and then you go from 10 users to 1 million users to 10 billion users overnight. It’s not like that. It’s a much slower process.
But to be fair, I would say that even though the upside in that sense is limited in terms of growth, the downsides I think are [00:18:30] also very limited in the sense that we’re profitable since day one. We’re cash flow positive and have always been. We never lose money in a week. I would say we probably […] three days go by that we lose money. They’re losing days, but even in a three days’ stretch is we don’t lose money.
It’s very stable in that sense and there’s a lot less risk. I would say it’s grindy. It’s still crypto, so everything’s still fairly fast compared to anything else, but relatively speaking, compared to like a [00:19:00] tech company in crypto, our business is more grindy.
Clay Collins: Yeah. That makes a lot of sense. We’re in the data business. We found that one of the biggest growth channels for exchanges is doing things a little bit differently than traditional exchanges have done which is to make most of their money off of data, and instead to create publicly available API’s that anyone can crawl, because those are distribution channels for the trading pairs that they have available [00:19:30] which, in a lot of ways, is the product of exchanges. OTC desks have not opened to sharing their data. Can you speak a little bit to why OTC desk are a little bit reluctant to share data?
Kevin Zhou: Ultimately, it’s because their principles in the trade and they function in a proprietary manner. Like the exchanges, they are not really trying to make money trading, right? They’re just matching up buyers and sellers, [00:20:00] they take a fee. OTC desk, they think that there’s a lot of sort of alpha or excess value in the data itself, and I agree with them. I actually think that there’s a lot of valuable data there that first of all, nobody else has, not even other OTC desk have this particular OTC desk data. There’s a lot of interesting information on there, right?
For example, I’ll just give you an example. You have labeled data on the social graph of how the funds flow between your counterparties, between them and other OTC desks, and between them [00:20:30] and the exchanges, right? Certainly, there has to be strictly greater than zero in terms of value, right? There must be at least something there. And then, I think also, they’re a little bit maybe wary about, if a lot of liquidity is crossing in these darker venues or it’s just opaque, then that means that the OTC desk get to see a little bit more about the price discovery, because they can see the lit venues, too, and they can see their own dark venue. You get a little bit more information about [00:21:00] price discovery than other people, so that probably is an advantage, too.
I think it’s the thing where, like most prop shops in Chicago are secretive because they have proprietary information or analyses or data that they don’t want everybody else to have and that’s their competitive advantage, right?
Clay Collins: Prop shops are OTC desks?
Kevin Zhou: A lot of them are starting OTC desk, yeah.
Clay Collins: What are prop shops?
Kevin Zhou: Prop shops are for proprietary in training, so you see a lot of them in Chicago. They have no investors, [00:21:30] they have no customers, they just trade their own money. They’re like a hedge fund, but without investors.
Clay Collins: I’ve definitely heard folks on panels claim that on certain months collective OTC volume eclipses exchange volume in a crypto space. One, I don’t know how you would be able to actually verify that, but does that seem truthy to you? Does that seem to be right on some months?
Kevin Zhou: No. I don’t think so. I think that if you ask the OTC desk themselves [00:22:00], nobody would say that. They would either say nothing or they would say that that’s probably not the case. That’s my opinion. That being said, I still think it’s very significant, in my opinion, as a portion of the total volume, but I don’t think it’s greater than the lit venues as of right now.
Clay Collins: Okay. Got it. That’s interesting. I understand now what you mean by lit venue. Those are sort of transparent trading venues versus dark venues, where no information is revealed.
Kevin Zhou: Yes.
Clay Collins: Okay. Got it.
Kevin Zhou: [00:22:30] Here’s another piece of game theory too. Let’s say that I revealed all my data and it’s public now, and then one of the other desks sees that one of these addresses I trade with is also one of the addresses that he trades with. Now, all of a sudden, he knows that this counterparty is shared between both of us. Well, before maybe he thought he was the only OTC desk that they went to, right? Because of that now he’s going to change how he quotes because he knows he’s more competitive with at least one other party that they have to quote tighter. [00:23:00] That the person passing isn’t passing because they think the quote is too wide, but because they’re getting a better deal somewhere else, right?
Information like that could leak too. It’s just kind of like a standoff, right? Because it’s the sort of thing where if you think about it it’s like the prisoner dilemma kind of situation where it’s actually in everybody’s interests to kind of pool all that data together and then everybody can use that data because then everybody benefits. That’s the socially global maximum. But then [00:23:30] locally, the maximum is to just never share your data, right? It’s a complex coordination game and it’s almost like a prisoner dilemma situation where the natural equilibrium is to have the inferior global result.
Clay Collins: Well, to anyone listening to this if you operate in OTC desk, my goal is to find the right granularity and interval for data to reveal and also do a ton of marketing around this. Hopefully, we can create a scenario where you’re revealing an amount that feels [00:24:00] very safe to you and the upside is, we can get you a lot of exposure and there’s a lot of things we can do to accomplish that. Hopefully, I’ll crack this code at some point. It might be a long time though.
Let’s transition to chapter five which is an exploration of the different kinds of OTC desks that exist today. The reason that I’m even bringing this up is because the other day, I was in a telegram group and I was actually looking to speak to a variety of OTC desk operators. [00:24:30] I asked just kind of generally for someone to contact me, or for folks to contact me if they ran an OTC desk. I had like 30 people reach out and tell me that they ran OTC desks. It was like crypto Pee Wee Herman. Everyone was claiming to run an OTC desk.
Could you walk me through the major categories of, I guess, legit and [00:25:00] I guess maybe there’s a gray area and then perhaps there’s categories of potentially illegitimate OTC desks that exist and how you personally carve up the space?
Kevin Zhou: I think overall there’s so many OTC desks that have been set up this year and then I think in many ways, I don’t consider any of them to actually be OTC desks. Like what you were saying, some of these guys are more like large block brokers. The way that I think about an OTC desk is [00:25:30] either you’re functioning as principal and every time you make a trade, you’re outlaying some kind of capital from your own inventory or you’re doing it on an agency basis. By that I mean if someone comes in to buy or sell a large block, you find them a seller on the other side, and then you match them up and you take a fee.
But some of these guys what they’re really doing is that they’re not really a desk they’re more like a buyer or seller’s agent, right? And there’s all these huge blocks that are [00:26:00] just floating around, I’ve never seen any of them cross, I think they’re all fake. Usually, it’s of the magnitude of 10,000 Bitcoin or 100,00 Bitcoin and people are looking for 5% discount to spot which is, first of all, very bizarre because usually the person that’s doing the seeking is the one that’s paying up, right?
If I want to go and buy a huge clip of Bitcoin, I have to be the one to pay up against spot and there’s no hope for me to get a discount there [00:26:30] unless I find somebody on the other side who’s, I guess, seeking even harder than I am. In the business, we call this being axed. It’s when you have an ax to grind, it’s when your axed to buy something or you’re axed to sell something. It means you have a very strong intention of being on one side and because of that you’re willing to pay up for it.
Usually, when the person who’s looking, who’s axed, is looking for some counter flow or for some match they’re the ones that are paying up. [00:27:00] But this isn’t always the case for these really large phantom blocks. Sometimes people have approached us and said, “Hey, if we buy 10,000 Bitcoin, can we get a discount?” And my response, initially in my head at least is, “Well, you came looking for me, so if anything, you should pay a premium.” We’ve been seeing a lot of these blocks floating around. I think I’ve personally seen it at least a hundred times and they’re all, from what I see, to be fake. And I was wondering-
Clay Collins: [00:27:30] What’s the play there?
Kevin Zhou: I was wondering why that’s the case. I came up with some initial possibilities and I think I finally settled on one. The first was that I thought that maybe there are people out there that are just looking to pump the price. It’s a very scary thing if someone says, “Hey, I want to buy 100,000 Bitcoin.” It barely even exists. There’s very few people in the world with even 100,000 Bitcoin.
A lot of the stuff that crosses in the OTC market, which is very opaque, [00:28:00] eventually ends up effecting the real market and the lit venues, sometimes in a muted way but there’s at least some effect that trickles over. The fact that there’s somebody looking for a huge block could mean, at least you could think, that the price maybe will be going up at some point. I thought maybe people were just looking to pump the price and circulating these fake blocks.
My second thought was that maybe what this is, is that the government themselves are very curious as to [00:28:30] what’s going on in these opaque markets and they’re just trying to figure out where all the dark liquidity is. They already have a pretty good grip on what’s going on in the lit venues, all these exchanges. They have some reporting requirements for KYC, AML at least in the US, and maybe this is a way for the government to poke around and see where all the dark liquidity is—so that was my second thought.
And then my third thought, which I think is the actual reason, is that I think what [00:29:00] people are looking for is that there’s a whole bunch of scammers out there and they’re looking for who holds a lot of Bitcoin. In finding who does they can eventually do phone recording attacks, right, where they call in to your service provider and port your sim card, that sort of thing. I think it’s more of a social engineering angle and people trying to get information about who holds what. That was sort of what I settled on. I think that’s really what’s going on.
Everybody that I’ve come into contact with who are circulating these large blocks, [00:29:30] sometimes they come from very reputable middle men and third parties. In the beginning, I was taken aback by just how fishy the whole thing looked given that it came from some of my closer contacts. Yeah, I mean in the end, just talking with the folks directly usually you can’t even see the buyer so usually it’s somebody’s agent, right, and you’re talking with them. Just something just feels off and I don’t know if these guys are good people that are just getting roped into this stuff, maybe they’re just a little too greedy, [00:30:00] they want a huge commission off these billion-dollar blocks or what not, or if they’re in on it or something like that.
At the height of my paranoia, I thought that what had happened is that they themselves got scammed out of their Bitcoin and now they have to do these scammer’s dirty work in order to get their bitcoin back and thus it’s like a pyramid scheme or something like that, or it’s some kind of pass it on forward kind of scheme. That’s sort of my thought. I think a lot of these OTC desks, or at least these so-called OTC desks, they’re really just a whole bunch of these agents [00:30:30] that are circulating these blocks. I’ve seen these deals where there were like 14 people in the middle and it’s just like nobody even knows who the buyer and the seller is.
Clay Collins: Wow.
Kevin Zhou: It’s just madness out there. I think it’s a very dangerous, dangerous thing to get involved with. Ever since about, maybe about half a year ago, we just stopped looking at those blocks entirely. The moment someone even mentions, “Oh, there’s 100,000 Bitcoin block.” and we just assumed it’s not real. I’ve also talked with a whole bunch of [00:31:00] other OTC desks, at least the bigger ones, and they’ve all said the same thing, that this comes across their desk all the time but nothing’s ever crossed.
I think there’s a very easy way to prove when something has crossed, and I’ve had even agents insist that blocks like this have crossed which is at the point on the chain through transactions that represent these blocks crossing. I’ve even heard from another OTC desk that somebody had actually tried to do that to them and then really what it was is it was an exchange that was doing cold storage movement. [00:31:30] They were basically lied to.
I think overall, it’s just a very dangerous thing to get involved with and at this point I just assume it’s not real.
Clay Collins: I’ve never seen blocks of 100,000 Bitcoin. I’ve seen blocks of several thousand, tens of thousands, but 100,000, that’s some real serious business. Could even be state actors looking to hack into accounts, or like you said, trying to find actually hidden pockets of wealth [00:32:00] that are hackable. Sometimes I wonder if there’s just a nameless person out there that’s talking to a whole bunch of so-called kind of dark net OTC desk operators and telling them all that they’re going to give them one side of this order if they can find the other side. Yeah, it’s hard to trace what’s really happening there but it does seem incredibly fishy.
You’re right, that doesn’t seem like a true OTC desk. It seems almost like a human order book that [00:32:30] someone’s just trying to match up two sides but really unable to independently execute anything on their own.
Kevin Zhou: Yeah, exactly. That’s exactly how I see it. Everybody’s calling themselves an OTC desk. I also think this kind of thing’s circulating, really it tarnishes the reputation of the entire OTC space regardless of if you’re doing it legitimately or not. Now, the word OTC is associated, [00:33:00] in part at least, with these huge phantom blocks that are just circulating. I think it’s just been very detrimental to the entire OTC space. I know that a lot of the bigger desks have faced very similar problems, pretty much the same problems, and nobody pays it any attention anymore.
Clay Collins: I guess the largest desks, Circle, Cumberland, Genesis, kind of have established names as an up and coming, and fast-growing OTC desk. [00:33:30] How do you differentiate against those smaller groups? It’s almost an unfair question. You’re not a nameless, faceless, anonymous entity on Telegram. But what do you point to when establishing legitimacy on a 15-minute exploratory call with a new client? Do you point out that you’re regulated? I’m sure there’s SEC filings. What’s sort of the quickest path to establishing [00:34:00] credibility with a newer client?
Kevin Zhou: I think really clients are looking for two things. The first is that ultimately, with the OTC desk they have to outlay the capital first. They do bear some counterparty risk of the desk and generally, if the desk is bigger, or more reputable, that risk is a lot lower. It’s going to be very hard pressed to find that Cumberland’s going to renege on their deals, right? There’s at least some baseline of trust there especially if folks like them [00:34:30] have been doing it for such a long time.
The other thing that people are looking for, and I think this really depends on the type of counterparty, but for the most part, most counterparties, at least to some degree, are price sensitive. There is a lot of price competition in the OTC space. Just to give you a sense of that, I would say that in the past six months spreads have gotten cut by a factor of two. They’re half of what they used to be even six months ago and a lot of that’s due to new entrance [00:35:00] and a lot of that’s just due to ongoing competition and also a bit of the bare market.
In terms of how I think we differentiate ourselves really, I think in many ways the service itself is reasonably commoditized. If you go to us, if you go to Circle, if you go to Cumberland, your experience is going to be about the same. I would like to think that we have a better service and we provide much better white glove service, but I think for the most part all of the desks that have been doing it for a while [00:35:30] customer service is a very high priority for all of us. I think we definitely treat the counterparties all very well. I think on that front you’re going to get a pretty similar experience.
I think on the price competition front, to even stay in business, to actually have a flow done through your desk, you have to be price competitive at least with most counterparties. I think maybe with some that are less price sensitive, maybe minors for example who have a mandate to sell, you don’t have to be as flexible on price, but for the most part everybody’s come down on pricing. [00:36:00] Things are definitely not as lucrative as they used to be for any OTC desk and through this competition. We charge maybe about, on a million dollars’ worth of bitcoin, maybe 35 basis points and then about half of that, or a little bit more than that goes to hedging costs, right? We’re only making about 15 basis points every million we trade in Bitcoin—that’s not very much. I think margins for other desks are similar, of the major desks at least are similar.
I think ultimately, what really differentiates the desks is not [00:36:30] anything sort of external, but more about sort of their internal operations. What I mean by that is, for example, if you have a very strong sales team and a very big one, you’re just able to find more counterparties, right? And in doing so, once you’ve built up a huge client list, then you’re able to get naturals matched up a lot more frequently which drives down further your hedging costs, which further drives down the pricing that you give to counterparties.
I definitely think that there’s some efficiencies [00:37:00] of scale. It’s not to the extent that exchanges have efficiencies of scale, liquidity does beget liquidity, but it’s not as extreme as that. It’s still partly a sales game. I think that’s really what differentiates the different OTC operations. It’s sort of how they manage their own internal processes and how they go about making sales and soliciting for new counterparties. In terms of what you’re going to get as the customer, as the counterparty, [00:37:30] it’s pretty similar at least among the major desks and the professional outfits it’s all pretty similar I would say.
Clay Collins: A couple questions came up. What are hedging costs? For those not familiar with that term? Actually, let’s break that down and then we’ll move on to the next item that popped up out of your comments.
Kevin Zhou: Hedging costs are what we bear as a desk in order to facilitate liquidity. If I’ve sold you a million dollars’ worth of Bitcoin, I have to go find some way to cover myself, to get back to a balanced inventory. [00:38:00] I can do that in a number of ways. I can source it from the lit venues, the exchanges themselves, I can source it from other OTC desks, or I can warehouse the risk and wait for matching counter flow. Those are all different ways of sort of hedging back to an initial balance before I made the trade.
Obviously, between the three the most valuable one is to actually have natural matches because then instead of having a hedging cost where you bleed away part of the spread that you’ve made you actually get paid twice. [00:38:30] You get paid on the buy leg and you get paid on the sell leg. I would say that it’s not even just doubly as good it’s maybe four times as good to have natural counter flow and to match it naturally.
That’s sort of what I mean by hedging costs. If we didn’t have matching counter flow, I would say that based on the spread that we’re charging maybe half or maybe a little bit more than half of that goes into our own cost that we bear in getting back to a balanced inventory. There’s sort of the gross margin which is what we charge [00:39:00] which is this 35 basis points, and then there’s the net margin which is more like 15 basis points, let’s say on a million dollars’ worth of Bitcoin.
That’s something that I think maybe not a lot of people understand that they should keep in mind is that the sort of ticket price that they see when they’re dealing with an OTC desk isn’t exactly how much the OTC desk is making. The OTC desk makes less than that and sort of depending on their own risk preferences, how much less, but almost always less than that.
Clay Collins: The second piece I pulled out of there was the [00:39:30] discussion around sales operations. Can you get into that a little bit more? I come from a SaaS world, a B2B enterprise sales world, and when I hear about sales operations, I think about sales development reps, outbound phone calls, one-to-one email campaigns, things like that. I have a feeling that you’re referring to something a little bit different, or are you?
Kevin Zhou: There’s some similarities there and there’s some differences. I agree in the sense that [00:40:00] OTC desk sales are very much like B2B, in the sense that if you cast a wide net and you do a whole bunch of Facebook adds or Google ads, you’re going to find a lot of folks that can’t really trade with you. The average guy who wants to just buy a thousand dollars’ worth of Bitcoin or sell two thousand dollars’ worth of Ether that’s sort of below the minimum size that’s required for an OTC desk. I think even the smallest ones are asking for minimum ticket size of 50,000, and for us, [00:40:30] it’s 100,000. I think Circles was even 250,000 at some point.
It doesn’t really make sense to do very broad marketing because most of the folks that you find won’t even be able to trade with. It has to be, in some ways, very targeted.
Now, the difference, I think, between OTC sales and B2B is that with B2B, you generally know who your customers are, and you just have to figure out a way to approach them. Cold emails, cold calls, finding some back channel with some mutual connects that you have, [00:41:00] that sort of thing. I think the real challenge with OTC sales is that you don’t really know who your counterparties, your future counterparties, your base of counterparties really are. You know their demographic, you know that ICOs that need to liquidate assets they’re a good demographic, you know that family offices and rich individuals they’re a good demographic, you know that miners are good, you know that hedge funds without execution facilities are good. But for the most part, there’s no list that you can just look up.
[00:41:30] You have to do a lot of digging, right? Maybe with the miners you kind of know who to target, but for everybody else, I mean there’s just so many people out there. There’s so many different hedge funds, so many different ICOs, so many different family offices. It’s actually kind of hard to figure out where to go and how to find these guys. That’s definitely a challenge.
From what we’ve seen we’ve made the most business within our own internal network. Sometimes you do a trade with somebody and everything goes well, they really like your service and they recommend it to [00:42:00] a fellow hedge fund, a friend and then one meeting leads to another and eventually, you build out your book of business there.
I would also say that having been in the space for a long time it definitely helps because even people that you know that you haven’t done business with before OTC, having built up that kind of trust over the years, especially in kind of Wild West space like crypto where you’ve seen that person for four or five years, and you know what they do, and you guys have gotten lunch many times, and drinks many times, [00:42:30] there’s definitely a lot of trust that builds up. I think having a good history and track record in the crypto space goes a long way, and then having an extensive network definitely helps.
Now we’ve also tried a lot of outbound sales campaigns and they’ve worked to varying degrees. You go to conferences, you’d imagine that’s a great place to meet people and that’s a great place to do business. It turns out you get some here and there, but it’s not really that great. At conferences you [00:43:00] meet so many people, everybody collects so many business cards, there’s not too much follow up. I find that one-on-one is a lot more productive than just randomly meeting people at conferences and having a five-, ten-minute chat. I think it’s all about sitting down for coffee, spending a good half an hour to an hour really getting to know each other, and maybe even the first meeting there’s nothing to be done. Maybe it’s just a prelude to eventually six months down the line doing business, right, as the people get more comfortable. [00:43:30] I think that’s all part of the cost of doing bus dev, that’s sort of what I’ve seen.
Clay Collins: It’s about being top of mind when they do actually want to execute a trade. It’s being that first phone call which might be a great deal of time…
Kevin Zhou: Yeah, definitely.
Clay Collins: …since the first conversation.
Kevin Zhou: You’re really just seeding them with the thought that you exist and you’re ready to provide this service. When they’re comfortable doing it, they can come to you at any time. From what I’ve seen, actually even within the company, there’s really two [00:44:00] major styles of doing sales. You have one kind where the sales guy is very aggressive. He targets a potential counterparty and just aggressively follows up, and just tries to bring them in, and tries to close the deal, just active pursuit, very much like hunting.
I think the other style—and this is sort of it’s more my own style—which is that I like to sow seeds. I don’t really like to aggressively try and convince people that they need our service or not. [00:44:30] I just put it out there and I talk to a lot of people. I seed them with the idea that we, as a service, exist. We’re there for them when they need us, and to come to us when they need us. I think it’s a calmer, nicer way to approach things. I’m not slamming their email, and their LinkedIn, and their phone all the time. I just kind of sow the seeds and some of your crop is just going to die and they’re not going to bear fruit and some of it is going to come back maybe two months later, three months later. That’s [00:45:00] more the two styles that I’ve seen at least on doing sales.
Clay Collins: Getting back to that original question about different types of OTC desks, obviously there’s the Telegram Anonymous OTC desk which is someone located who knows where, who’s motives are unclear, they’re not even willing to name themselves, they’re not regulated, they’re not filing paperwork, they’re not domiciled anywhere, and they’re really just brokers, right? They’re not taking a side of the trade. [00:45:30] And then there’s all the way on the other extreme, there’s you guys and other names that folks have heard of. Are there any shades of gray? Are there any other kind of major categories of OTC desk that you see existing?
Kevin Zhou: Yeah, I would definitely say so. I think the Asian OTC desks are in a class of their own. Many of them are real, but from what I’ve seen, the spreads are massive. I don’t know why that’s the case. It seems like here in the U.S., at least on the dollar crosses, [00:46:00] price competition has been very severe, and I think margins have come down a lot. For the most part, liquidity’s very cheap.
In Asia and Europe, we’ve seen spreads quoted by OTC desks anywhere from two times to four times as wide. You’re going to be able to get it done there, but I really wonder why the pricing is so bad. Is it that people over there don’t have many options? Is it that part of that spread goes into the conversion or the movement of, let’s say, onshore R&B, right? [00:46:30] Maybe that’s anywhere there’s capital controls, there’s very huge cost of getting money into and out of the country. Or is it something else? But from what I’ve seen, there are these Asian OTC desks and even in Europe too where the spreads are massive, and some of these desks, they also run a whole bunch of other operations. They also do market making, they also invest into ICOs, and they provide a full stack investment house service for everybody that they work with. In that way, they sort of have captive order flow.
If an investor [00:47:00] invests into your token sale, even if they’re a little bit more expensive, maybe it makes sense just to do your flow with them OTC, just to keep relations, that sort of thing. I think you see some of that that’s happening in Asia. I think here in the U.S., I think people are a lot more specialized. You have your VC hedge funds that always do investment. You have your OTD desks that only trade in secondaries. You have your market makers that only do market making. I think there’s less of this kind of mix of services in [00:47:30] sort of vertically integrated investment house. But everything’s done differently in different places
I don’t think it’s a bad business in particular, to have captive order flow, but I really wonder if that’s really the best way to go for some of these ICOs and if they’re not really giving up a lot to this OTC desk that happens to be their investment. If you think about it, if the investors writing a small check, and then they get all of the flow for the entire token sale, they could make back all the money that they invested, and for them than it’s just a free option. [00:48:00] They just get a whole bunch of tokens for free if they’re charging such wide spreads. Not a bad business from a business perspective, but I really wonder if all the incentives are aligned.
Clay Collins: Usually, when there’s a large spread, it’s you’re paying for something. You might not always know what it is.
Kevin Zhou: That’s right.
Clay Collins: Moving to chapter six, let’s talk a little bit about control and regulation. How are you regulated, how are most OTC desks in the U.S. regulated, and then maybe let’s extend that to [00:48:30] worldwide, what does a regulator landscape look like?
Kevin Zhou: Sure. There’s a lot of different angles you can take when it comes to regulation, and especially of the OTC space, which I think is a little bit less regulated than the exchange space. There have been desks that have taken the position that because the desk is a principal to every trade, that they’re a first party not a third party to every trade, that’s effectively like just trading, buying or selling crypto.
A lot of these desks [00:49:00] don’t believe that they need any kind of broker-dealer registration or any kind of state by state money transmission license. I think it’s an aggressive position to take but there’s sort of varying degrees of aggressiveness. On one hand, it’s aggressive to not have money transmission licenses. On the other hand, it’s arguable whether or not these things are securities. Then there’s also the point that even if these things are securities, broker-dealer registration allows you to trade—from my understanding—[00:49:30] to trade registered securities, to function as brokers or dealers for registered securities. But even if a lot of these assets are securities, most of them are not registered securities, right? They’re unregistered securities. There’s a question of whether or not broker, broker-dealers would be able to trade even these assets even with that kind of registration.
I think on the money transmission sort of angle, I would say that almost every proper OTC desk does abide by AML rules set up by FinCEN and the regulatory bodies [00:50:00] and KYC rules set out by those same bodies. But whether or not they’re registered as a money transmitter, I think everybody abides by those rules just in case. It’s not useful just to abide by regulation to sort of preempt that position. But also, because every once and a while you’re going to get a random subpoena from the FCC or the IRS or law enforcement’s going to come knocking and ask you about a lot of random transactions that you’ve done. [00:50:30] In the same way that exchanges have to keep track of what they’re customers are doing. I think you want to be in a position to be able to say, “Yeah, we traded with this person. We know exactly who they are.” rather than, “We traded with them. We have no idea who they are.” I think that’s just a really bad look.
It’s sort of my thought that even if it’s still legally unclear whether or not you need to abide by these rules, you should still do that, and that’s something that we definitely do. We collect AYC information on everybody. We’ve had [00:51:00] counter parties approach us and say, “Hey, I know a guy, and he wants to do this, but he doesn’t want to be known who he is.” Then we just always turn it away. It’s just not worth it. I think there’s sort of that angle.
There’s another thing to talk about, which is that in the U.S., if you’re a US-based entity, you’re trading derivatives for futures, you need to register with the CFTC. There’s I think two registrations that you can make. Also, I want to say—I should have said this at the beginning—which is that none of this is legal advice, [00:51:30] and I recommend that everybody consult with their own legal counsel. This is just my opinion, and it can be right or wrong, but I just thought I’d give it.
With the CFTC, you can register as a CPO or a Commodity Pool Operator or you could register as an ECP, and it’s like an eligible something rather, I forget what it stands for. I think the ECP registration is much lighter registration. It’s much easier to get. CPO’s a little bit more in depth. But you can get all of these. [00:52:00] It’s not super difficult. It’s not impossible. But generally, you need some kind of registration like that in order to trade futures or derivatives. That is the trade on CME or CBO directly. Obviously, even without I think these registrations, you can still go through a broker that has access to these things. But to trade direct, you will need some kind of registration.
Now, whether or not you’re able to trade on BitX with their unregistered sort of futures and options with or without CPO registration, very hard to say. [00:52:30] We personally don’t right now, but I guess you could make a case that it’s completely unregistered. It doesn’t matter, especially if you’re offshore. That’s an easier case to make. But I think as far as the US-based firms go, I think people are a little bit more careful. I think on average people just don’t trade. At least the ones that do, they just like VP and out, and then they just don’t talk about it. I know that that’s pretty common too. I think that’s sort of an overview on this kind of stuff.
[00:53:00] I think a lot of the stuff that OTC desks trade, it’s sort of in a gray area on whether or not it’s considered a security. I know the FCC said Bitcoin […] thrown off securities, then you have to think what about, let’s say, Litecoin, what about Ripple, what about NEO, and then you can go further and further down the list.
I think in some ways, until there’s absolute clarity, a lot of these desks have to take on some kind of risk assessment. They can be more aggressive, capture more business, or they can be more safe, and [00:53:30] capture less business. I think that’s in some ways a business decision but also obviously consulting with legal counsel is extremely important. Even then, legal counsel could give bad opinions, but I think there’s some trade off there. At least you’re somewhat covered if a large law firm says that it’s okay, but then maybe you’re going to have to pay them for that opinion letter. It’s that sort of game and dance in many ways.
I think so far, the regulators have been surprisingly, [00:54:00] I was actually a bit surprised about this, but I think they’ve actually been pretty reasonable about things. The FCC’s only really gone off to very egregious offenders and clearly bad actors in the space. I think mostly they’ve done a good job of in some ways staying hands off but, in some ways, really going after the absolute, unarguable, unambiguously bad actors.
I would like to think that if you’re intentions are good, and you try and do things the right way, that things will be okay. I think really the thing [00:54:30] the FCC really doesn’t like is when retail guys or even just random folks, random citizen, just start complaining to them about a particular business saying that, “Oh, these guys ripped me off. These guys scammed me.” This, this and that. I think as long as you do good business with people, that risk is much less. You still could be in violation, but at least that risk in particular is much less.
But yeah, definitely a very gray area right now. [00:55:00] I think we’re definitely on the lighter shade of gray, and I think most of the desks are on a wider, at least the legitimate ones are on the lighter shade of gray.
Clay Collins: Could you just walk us through how you’re regulated, what you’re registered as, where your domiciled? Are you Delaware LLC? If you could just walk us through sort of the basket of things that you’ve subjected yourself to legally to get where you are today.
Kevin Zhou: I would say that we’re a US-domiciled fund and we trade out of the fund. All of our entities are in the US. [00:55:30] We actually have a whole number of entities. This is actually getting more into how a hedge fund is structured rather than an OTC desk.
Clay Collins: Yep.
Kevin Zhou: There’s general partner entity, there’s management company entity, there might be subsidiaries under them. Then there’s also the master fund itself, which is a limited partnership. The other two entities are LLCs and they’re all based in Delaware. That’s how we’ve structured ourselves. For investment, we’ve only taken accredited investors. We’ve followed pretty much the guidelines for setting up a small hedge fund. We followed them pretty much to the T. Legal counsels on the same page. [00:56:30] Our auditors, administrators are on the same page.
We do have yearly audits. We have a third-party administrator, and we have actually multiple firms as legal counsel at this point. In terms of sort of the compliance measures that we have, we collect KYC and do AML checks on every counterparty that we interface with. That includes looking them up on politically exposed person’s list, […] fact list, all of that stuff. [00:56:30] We do file suspicious activity reports for anything that looks or seems highly suspicious. I think for the most part, we comply with everything that you would imagine an OTC desk should, at least that’s based in the US.
Now, that being said, we do have an initiative right now to sort of re-base ourselves offshore, but it’s not really for compliance reasons. I just want to sort of set expectations there. Right now, a lot of these exchanges that are offshore like OKCoin, Huobi, bitMAX, Bitfinex, [00:57:00] they won’t take on US customers as clients. We aren’t able to access a huge pool of liquidity just by being a US entity. We’re thinking about that maybe it makes sense to re-base either Cayman or BVI, which is very common. I think most hedge funds, any industry, are located in BVI and Cayman.
What’s funny is it used to be because of tax benefits. I think at this point there are no more tax benefits. I think it’s just tradition, what they’re used to. I think they have good law [00:57:30] there for hedge funds. They’ve had such a bustling industry for such a long time in the hedge fund space. They have good law now for hedge funds. I think that’s one thing that we’re thinking about doing. Who knows how long it’ll take and how smooth the process to cut over will be, but that’s initiative that we have ongoing.
There is one small consideration I think on the tax side, which is whether or not foreign investors is subject to US tax withholding or whether they fall under something called the 8-64B exemption. It’s either the 8-64B or just 8-64 exemption. [00:58:00] It has to do with how offshore investors are treated. But that’s less relevant to the OTC trading. That’s more relevant to the hedge fund investors. There’s some question about whether or not re-basing to Cayman would make that at least easier or more clear.
Clay Collins: Let’s transition to the final chapter. What do you think the future looks like for the OTC business? It strikes me that you’re set up as a fund, so you could go in a number of different directions. You have a lot of moves on the chess board, if you will. [00:58:30] What do you see happening into the future as margins continue to get squeezed or maybe the trend reverses? How do you see this business evolving over time?
Kevin Zhou: I would say that it’s probably the case that exchanges are more in a position of buying OTC desks than the other way around.
Clay Collins: Interesting.
Kevin Zhou: OTC desks are generally very profitable, but not at the extent that exchanges are profitable. Because if you think about it, there’s still a very human element to it, [00:59:00] and because of that, it’s not the easiest thing to scale. As opposed to like a matching engine, just pure code is a lot easier to scale. I think maybe the exchange, they’ll buy out some of the OTC desks. I also think that some of the OTC desks are just going to close down. They’re going to lose in this price competition and just going to get squeezed out. I also think that overall margins going to continue to go down. Maybe we’re going to have a slight reprieve if there’s another bull run very soon. I certainly think it’s still going to be very profitable when there is a bull run and that maybe [00:59:30] the margins are still only a little bit higher or just as low, but you make it up in volume. Volume goes 10X and there’s a lot more business to do. I think there’s definitely possibility there.
I think overall, I think what you’re going to see is you’re going to see a lot of infrastructure built out to sort of automate the OTC process. It’s kind of almost an oxymoron because the whole idea behind OTC is there’s a human on the other side. It’s like service, right? But at least you’re going to start seeing these dark pools where the desks themselves [01:00:00] have access to and they cross with other desks. Sort of like matching services between desks, right? I think that’s always going to be coming at some point.
And then also like broker chats. In the traditional space, there’s a bunch of broker chats. A lot of liquidity costs is off market in sort of a peer-to-peer fashion, and then people just put out their indicatives. If some guy is like, “I’m selling 100 Bitcoin at this price,” and then some guy says, “Okay. Well, I’ll take you up on that.” something [01:00:30] like that, but in sort of more of in a group format rather than just everybody’s single party to single party.
Right now, that’s the case because none of the OTC desks, nobody wants each other to know who their counter parties are. But I think at this point, all the OTC desks know each other. Having some kind of broker chat at least between themselves would be a natural way to go. Then outside of that, I think everything becomes more sort of automated, right? There’s already a few desks that are doing RFQ systems. [01:01:00] What that is instead of having a human on the other side, you ping them for a quote, give them some sort of size and it can be generally pretty small, I’ve even heard of as low as like $10, and then they’ll just spit you out a price, they’ll just quote you a price using their engine, using their algorithm. Kind of like an automated, like a robot OTC on the other side, right?
Then even beyond that, I would say there’s streaming services. What that is it’s like RFQ, but you never have to request for the quote. The quote is just continuously being streamed to you at different levels. Like a 10 Bitcoin, a 100 Bitcoin, a [01:01:30] 1000 Bitcoin. If you go up to a certain level, you just pay that price. If you go past that level, then for the first level, you pay the original price; for the second level, you pay whatever the remainder is. Something like that. It’s sort of like an automated order book with only one market maker on the other side. Much more like dark pool, like automated dark pool style or there’s only captive, right? It’s completely captive.
I think you’re starting to see services like that being put up, and then ultimately, if you think about it, as long as people have technological capabilities [01:02:00] that is a more efficient way to source liquidity, to provide liquidity, I do think that there’s always going to be a place for human interfacing in the OTC space just because a lot of times people just really appreciate that. People like hearing someone else on the other side. People like having their questions answered at hawk, right? They just come up with a random question. The person the phone can answer for them.
I also think that a lot of very wealthy individuals, they value their time so highly [01:02:30] that they’re willing to pay for you to save them time and make their experience better. I always think that there’s some value to be had doing manual OTC. But I do think that overtime, that space is going to shrink and there’s going to be, I think, bigger incumbents, and most of it is going to get automated and margins will continue to fall. That’s sort of my opinion. That’s not great for the OTC business. But I think it is great for counter parties. I think it’s great for customers because there’s better pricing for them, and a smoother and more modified service.
Clay Collins: I have a question about the continuous price quotes [01:03:00] and price feeds. I imagine that they’re not making a hard commitment that if they were to approach the desk about doing a trade right then, that it would be exactly that price. It’s probably an approximate value, and then how did they account for volume? I guess the more you want to buy or sell, the higher a low or the price gets. How is that accounted for? What are sort of the ad hoc mechanics around these price feeds that OTC desks are sometimes providing to [01:03:30] privilege counterparties?
Kevin Zhou: I would say that there’s one major streaming service that I know that exists out there, and the way they do it is that everything is immediate or cancel. They’re providing you with basically a sparse order book—it’s an order book with a few levels. You can buy a Bitcoin, 101 per Bitcoin for 10 Bitcoin; 102 for 100; 105 for 10,000, something like that, and then you’re just sliding through their book [01:04:00] as if it was an order book depending on what size you wanted to buy. But the prices do move. They update their systems very, very quickly and rapidly. They have some algorithm in the back. If you issue an order and the prices moved, they’ll just reject the order and say the price has moved, and you’re not able to execute at that price.
In some ways they give you a free option, but in some ways, they have a free option on top of their free option when you exercise on their free option. They’re not going to take any exposure if the price is moving and their interface just hasn’t caught up. I would also say that [01:04:30] this service is a fairly privileged service in the sense that not anybody can get an account with them in order to trade with them. You have to do a certain amount of volume in order to even receive this service, and you need to do even more volume to have an API from them.
At the first level, if you do enough, you can just trade through the screens, through their own interface, and then if you do even more, they’ll eventually give you an API. I would also say that the levels they show you are also very custom. [01:05:00] If you in particular always want to buy 50 Bitcoin, they’ll show you a 50 Bitcoin level because that’s something that’s more common to you rather than going from like 10 to like 80, they’ll show you that 50 line.
I would also say that the symbols that they show are also a bit different too. In the beginning, at least with us, we could only see Bitcoin and Dollar and then over time, as they got used to trading with us and they had some measure of our toxicity, they were then able to show us other symbols, other pairs, and then expand from that to more and more illiquid stuff.
Really, they have to figure out [01:05:30] who they’re trading with because in these kinds of games, in OTC and marketing making, in streaming services like this, it’s all about figuring out the toxicity of the counter party. I mean, if they’re very, very toxic, you just don’t want to trade with them. If they’re just slightly toxic, you want to be wider or you want to over hedge. You want to join them on their side after trading with them. You never go in front of them, but you can always go after them.
I would say that that’s sort of the level of sophistication that I’ve seen so far, [01:06:00] but certainly, you could take it even further. I think that’s really the advent of everything becoming more electronic. But, I mean, in some ways there’s a different demographic. I mean, that’s more to service, professional outfits and the OTC desks are more there for, in some ways, still professional but folks who are less sophisticated in execution. The more price sensitive people would prefer the streaming service even if it’s more tedious to deal with.
Clay Collins: Suffice it to say, if you do want a live price feed [01:06:30] with different levels, you’re not going to get it right off the bat. If you do get it, you’re probably signing an NDA. You’re not going to get all the information at first. The potential for more information increases over time is as they come to know and trust you. This is not a mass market thing. These are feeds that are custom-tailored to sort of value B2B relationships.
Kevin Zhou: That’s exactly right. And this is also why I’m very hesitant in saying who this is, who this party is. [01:07:00] I think, it’s important to maintain relations, and I also think it’s if they want to be more private, than they should have that privilege too.
Clay Collins: I’ve heard of a couple different desks doing this. I think I have a sense of who this might be, but I’m going to leave this off here. When you spoke to toxic counter parties, do you consider a toxic counter party someone who might be trading on privilege information a day before the hammer hits the ground on something big? If you trade with someone and take a big loss, [01:07:30] are they toxic or are they just good traders?
Kevin Zhou: Actually, being good would mean that you’re toxic too. I think it’s not mutually exclusive.
Clay Collins: Oh, interesting.
Kevin Zhou: I think toxic would mean one of three things usually, maybe there’s more. Either you’re just really good and really sophisticated. Two, you’re not showing your full size, so you’re running the desk over. Or three, you have privilege information nobody else has. I think for us we’ve run into different levels of each. I would say that the most jarring [01:08:00] example is every time Coinbase makes an announcement, there’s always some folks that come to our desk to try and buy things ahead of time, and then sometimes we take a loss on that. It’s not very fun. I think it’s reflected in the price too. Every time Coinbase makes an announcement, you see the price trail up maybe a couple hours or a day before. I mean, somethings going on. Even in the lit venues, it’s going on. Or it’s like people that are doing OTC and OTC desks are then hedging themselves causing that to happen on the lit venues. You see stuff like that.
[01:08:30] There are definitely counter parties out there that will run you over, that will give you half the size, and they’ll run you over on the lit and use when you’re chasing the same liquidity or ship off the other half of the block to another desk, and everybody’s chasing the same thing, and not letting you know that it’s not the full block. You definitely see stuff like that. Occasionally, you’ll run into folks, I guess, who are just very good at time in the market. I think it’s sometimes it’s luck, but I think at some point you do have some counter party metrics and some people just have something figured out. Should definitely [01:09:00] be careful on how you trade. At least be a little bit wider.
Clay Collins: Kevin, this is one of the most insightful interviews that’s appeared on this podcast. Thank you for being so forthcoming and for, in a lot of ways, just opening yourself up in terms of how you operate. I’ve never heard an interview or any content that goes into this degree of depth on OTC desk operations. You’ve been very generous. You’ve been very kind.
Kevin Zhou: Yeah, I really appreciate that. I really appreciate you having me on the show.
Clay Collins: That concludes part two of this conversation with Kevin Zhou of Galois Capital. I’m off to continue working on a documentary we’re creating about the range of exchanges and exchange products that exist in the crypto space. See you next week.
That’s it for this week. To sign-up for our [01:10:00] free crypto investing newsletter, listen to other episodes or get the show notes from this episode, please visit flippening.com. I also invite you to check out the startup that funds this podcast, Nomics at nomics.com. Finally, if you got value from this show, the biggest thing you can do to help us out is to leave with a five-star review with some comments and feedback on iTunes, Stitches, or wherever you listen to podcast. [01:10:30] Thanks for listening. See you next week.