I am especially thrilled to be publishing this interview with the fascinating Sam Bankman-Fried. While studying physics at MIT Sam came across effective altruism – this is a research field that finds the most promising solutions to the world’s pressing problems. Sam decided that he can have the biggest impact in solving these problems by making money as a trader and donating it to important causes. In 2017, he launched Alameda Research to address the liquidity problems he saw in the growing crypto industry. Two years later, his team is trading an average of $1B a day across crypto markets, they are among the top 3 most profitable traders on BitMex, having realized profits exceeding $50M. The team has already donated $4M to save human lives, prevent animal suffering, and research AI. I believe Sam, Alameda, and their new exchange FTX are just getting started. I loved this talk and I hope you love it as well.
- The 3 reasons that made Sam start Alameda
- Mistakes Sam made while building Alameda
- Building a world-class team
- The Products of Alameda
- Quantitative trading. Arbitraging. And the challenges involved.
- Automated OTC desk.
- Liquidity providing for exchanges.
- How Alameda became one of the top 3 best performing traders on BitMex, generating $50+ million in profits.
- The FTX derivatives exchange.
- Solving the hundred million $ clawback problem
- Clawbacks explained in detail.
- Creating crypto index futures.
- Creating new crypto futures.
- Leveraged tokens (leveraged exposure without worrying about margins and liquidations).
- FTX’s ideal customer.
- Maximizing crypto profits with leveraged tokens.
- How was FTX’s impressive trading volumes built so fast?
- Alameda’s approach fo measuring fake volumes in real-time.
- The FTT token, its utilities, and the weekly burn of FTT.
- Sam’s personal time allocation.
- What Alameda & FTX will look like in 1 year?
- Sam’s Twitter – https://twitter.com/SBF_Alameda
- FTX website – https://ftx.com/#a=1456191/ (use this link to get 5% off trading fees)
- Alameda’s website – https://www.alameda-research.com/
- FTX Twitter – https://twitter.com/FTX_Official
If you enjoyed this episode, you will love this 2nd interview with Sam:
Full Episode Transcript
(00:01) George Manolov:
This is the Borderless Crypto Podcast.
(00:14) George Manolov:
Hello everyone. I’m your host George Manolov, and in the series; I bring you exceptional entrepreneurs, investors, hustlers, and thought leaders from the cryptocurrency and fintech space.
Hey, Borderless’ friends. I’m especially thrilled to be publishing this interview with the fascinating Sam Bankman-Fried. Whilst studying physics at MIT, Sam came across effective altruism, this is a research field that finds the most promising solutions to the world’s pressing problems. Sam decided that he can have the biggest impact in solving these problems by making money as a trader and donating it to important causes.
In 2017, he launched Alameda Research to address the liquidity problems he saw in the growing crypto industry. Two years later, his team is trading an average of $1 billion a day across all crypto markets, they’re among the top three most profitable traders on BitMEX, having realized profits exceeding $50 million. The team has already donated $4 million to save human lives, prevent animal suffering, and Research AI. I believe that Sam, Alameda and their new exchange FTX are just getting started. I love this talk and I hope you love it as well.
Please note that anything that I or my guests say in these talks is for informational purposes only and should not be treated as investment advice. Although during daylight, I’m part of the team behind the crypto lending platform Nexo, at night when I do these talks, I share only my personal thoughts, which in no way represent Nexo’s opinions.
Now, it’s time to get the podcast started in three, two, one.
(02:15) George Manolov:
Sam, the projects that you’ve been building have been of massive scale, both Alameda and FTX are impressive projects and I can’t wait to dive deeper into those. But before that, could you please take the time to share, in brief, what is your background before building those companies?
(02:41) Sam Bankman:
Yeah, absolutely. My background is in quantitative trading. I went to MIT for college, I was a physics major there. Afterwards, I went to work at Jane Street Capital, which is a Wall Street quantitative trading firm. I traded international ETFs there for a few years, those are basically US-listed ETFs that hold foreign companies. I did that for about 3 years, then left in 2017 and started up Alameda Research cryptocurrency quantitative trading firm.
(03:13) George Manolov:
All right. And how did you decide to move from the traditional finance space to the cryptocurrency industry?
(03:21) Sam Bankman:
I think that there are a few things going on there. One of them is that there just seems to be a pretty big opportunity. Looking at the ratio of customer excitement volume going on in crypto and the crypto industry, and the amount of professional liquidity being provided, we basically saw a ton of demand spiking up in cryptocurrencies, but not that much liquidity and infrastructure for it, this leads to large arbitrages and lots of good trading to do because customers are demanding so much liquidity and there just was not enough being provided.
On the outside looking inside, there looked to be giant arbitrages all over the place. It’s always a little bit difficult to tell exactly how real those are; whether those do really represent giant trading opportunities or weather that is just fake data. As it turns out, there absolutely was some fake data there, but there is also a lot of real data and the real data was enough for there to in fact be a lot of really exciting trading opportunities.
I think that part of me really wanted to try building my own thing. In the end, where I’m coming from, with most of this, is figuring out what the ways that I could have as much impact as possible on the world, and it seemed that figuring out how to donate as much money as possible was a pretty plausible path at that. The interesting thing that happens when you’re looking at things from a donation perspective is that the scale of impact that you have is pretty large, and so is the scale of resources that you’re combining with there.
If you think about the amount of money that the world is deploying to fight a lot of things, like neglected tropical diseases, is in the billions and in some cases the trillions. That means that it makes sense to be a lot less risk averse and that rather than just trying to find a way to have a comfortable lifestyle, there really is an advantage to trying to go all out. So, that was another thing that drew me to trying to build up my own company from scratch.
(05:52) George Manolov:
Got It. That makes a lot of sense, and I think it’s quite a different motivation from that a lot of other people have when they do the type of work that you’re doing with Alameda and FTX. The other question I have is: how’d you decide to make the jump from working as an employee, essentially a trader in an established trading company like Jane Street, to building your own company where you’re the ultimate boss, you are the manager, you’re the entrepreneur. How was this transition for you?
(06:21) Sam Bankman:
I think a part of the decision to start a business-like Alameda was that it was quantitative trading, and it was something that I had experience with. There is a lot of trial and error; there are a lot of things that I didn’t know when I started Alameda, and a lot of mistakes I made that I had to learn the hard way. Just to give some flavor of them, if you ask me if I should have read more books on management before I started, my answer would maybe I should have. But one thing is how do you hire, how’d you decide who to hire, how’d you decide how to manage people? I had some ideas and I think I did some things well, but there are definitely some things that I really fucked up. One example of that is: a lot of my hiring philosophy was centered around finding people who seemed really smart, hardworking, capable people, which obviously is great, but I think I didn’t place nearly enough emphasis on also them being people who just fit within the specific culture of the company I was building, and thinking that someone can be a brilliant hardworking person with experience in the right industry who could do a ton in the right place, but maybe this just isn’t the place for them for random reasons that have to do with whether the coworkers, the working environment and things like that are the right match for them. There are a ton of other things like this that I just didn’t really understand when I started Alameda, it wasn’t until I’d screwed some of them up that I struck home that there are things I had to be way more careful about.
(08:01) George Manolov:
Sam, where did you source the talent for Alameda? Where did you get the people from? Was it people that had traditional financial backgrounds like you or was it people who went directly into crypto?
(08:16) Sam Bankman:
On the trading side, it was people with a similar background to me, it was people who had been working in traditional finance, and in quantitative trading at Wall Street firms. On the developer side, it was basically friends of mine or friends of friends who were working at Google and Facebook. While on the ops and business development side, it’s a broader range of backgrounds, it is more of a mixed bag; some of them were friends of mine from before, but some of them were people who we met in the crypto industry working for other firms who then joined up with us.
(09:01) George Manolov:
Now maybe it’s time for you to give us a little bit of overview of what Alameda is about. You first launched Alameda in late 2017, what did you start with? What was your initial product and then what have been the other services and products that you have been providing?
(09:19) Sam Bankman:
At its core, Alameda is a quantitative trading firm. At the very beginning, that primarily meant doing a lot of arbitrage, it’s mostly just pure arbitrage, especially in late 2017 when markets were extremely inefficient, Coinbase and Bitstamp would be trading Bitcoin at different prices, and that was the easiest arbitrage you can find, so we started out doing a lot of that. We did a lot of cross-country arbitrage as well; we basically started up entities in a bunch of different jurisdictions and got bank accounts on them so that we could do fiat-based crypto arbitrages across different countries and exchanges, in addition to coin to coin trading.
Over time server trading has become a bit more nuanced and complicated as markets have gotten more efficient, you’ll no longer see a 5% arbitrage between Bitstamp and Coinbase, that doesn’t happen anymore. Instead what you might see is, either a much smaller arbitrage or a more complicated trade where you’ve got a Bitcoin-USD market and you’ve got a Bitcoin-Tether market on another exchange and they’re trading at different prices, but of course it’s not really a pure arbitrage because you’re picking up Tether in the middle of that. Then it comes down to understanding what is USDT worth, is it a dollar bill? It’s basically a dollar, but not exactly, it’s a little more complicated than that. So, over time our trading has increased the move to things which are not quite pure arbitrages to understanding what that means and how to serve algorithmic quantitative trading in an environment with a lot of things that are not quite what they appear to be.
The Korean Won on the Korean Exchange is not quite a normal Korean Won, since you can’t really get it out of the country. Tether is not quite a US dollar. Dai is not quite a US dollar. Even a Bitcoin on Binance and a Bitcoin on Huobi are not exactly the same product; you have to move it between the exchanges to turn them into each other and then hit withdrawal limits, blockchain delays or things like that. The crypto ecosystem is really complex, there’s a lot going on and there’s a lot of things that make products not quite what they appear to be, so a lot of what we try to do is to understand what is this, how do we trade given that, how do we price a Tether on OKEx right now, and what is that worth.
(11:48) George Manolov:
How do you do that in practice? I imagine that you build some sophisticated algorithms to be pricing up the value of particular assets at any point.
(12:01) Sam Bankman:
Yup, we’ve got a lot of algorithms. We have tens of thousands of parameters all contributing to the ultimate pricing of these objects, and on top of that we have humans constantly looking at markets and saying: I understand that our algorithms are saying that a Tether is currently worth 99.8 cents, but it’s trading in the markets for 97 cents, so that’s pretty weird, our Algorithms just think that we should be buying. When things like that happen, we have quantitative traders watching our trade 24/7 trying to understand whether something just changed about the world? Sometimes it’s yeah; The NYAG just announced that it’s investigating Bitfinex and that’s why Tether’s price went down, sometimes someone just sold a lot of it, other times where we honestly don’t know what’s going on. We want to be a little bit nervous about that, maybe someone else knows something we don’t, and knowing what you don’t know is quite important. Then, we have to decide how we’re going to synthesize together all that information and ultimately decide what trading we should do.
That’s a fine example, but there are a lot of other examples. Often, we’re thinking about how comfortable do we feel accumulating more Bitcoins on this exchange right now given our withdrawal limits; like we already have 2 days’ worth of withdrawal limits of Bitcoins, so how expensive is it to be locking up the marginal piece of capital for another day? Because that’s what’s going to happen if we get more Bitcoins there; it’s going to be 3 days before you can get them out. Maybe we charge ourselves 25 BITS for that, or maybe we think it’s actually fine to have more Bitcoins there, but that’s another decision that we’re making in real time and it’s really situationally dependent, and depending on how all of these variables shake out, we’ll end up making a decision about what trade do we actually want to do here.
(13:52) George Manolov:
Okay. As I understand your very first large initial product is finding arbitrages and deciding what to buy and when to buy. When you started doing this, did you look for such arbitrages and do you continue looking for such arbitrage for your own capital or do you manage other people’s capital?
(14:11) Sam Bankman:
We don’t have outside investors, so it’s entirely internal capital plus lines of credit, and that’s a somewhat intentional decision because it massively simplifies a lot of things about our corporate structure and our regulatory regimes, and makes us a lot nimbler and gives us freedom to manage our trading the way that we want.
(14:33) George Manolov:
Okay. So, do I understand correctly that this is your first product but then you do other stuff, right? You have an OTC desk and you’re doing other services.
(14:43) Sam Bankman:
Yup, that’s right. Over the years, we expanded our business. For the first year or so, this was basically all we did, but then late 2018 or so, we’ve grown out our business quite a bit. And as you said, one of the first things that we did is that we opened an OTC desk, this was a pretty natural product for us to build because we already had the liquidity and quantitative trading, which meant that we are already able to offer pretty aggressive pricing to customers because we could match off against our internal on exchange liquidity. So, we opened up our OTC desk late last year and from the beginning we decided that we wanted to make it as automated as possible. What we did was that we automated a few systems, this is a platform where the first one was at Alamedaotc.com, where you can go, create an account, deposit and withdrawal funds, and then you can get automated 24/7 OTC quotes there; you just click one button to get a quote, click another button if you want to do a trade against the price we show, and then you can withdraw and serve your side from the platform. So, that was the first product that we built and we’ve expanded that a bit over the years.
But at its core, we see it as a combination of us offering the liquidity that we’ve built up OTC to customers combined with building technology to make it really clean and simple; you don’t have to argue over chat about when exactly someone typed something or anything like that.
(16:17) George Manolov:
Do other companies provide such a service?
(16:19) Sam Bankman:
There’ve been a few others that have done it, not very many do. As far as we know, we have by far the widest ranging one. So, while other people do have some OTC capabilities and some of it is automated, we generally see them offering way fewer products, they usually don’t have a wallet and settlement system built in. We think that we have served the most complete RFQ offering in crypto.
(16:45) George Manolov:
Right. Because the first thought I at least have that this is something manual, you’re talking to a real person.
(16:50) Sam Bankman:
Exactly, and that is the norm in crypto and we’re frankly a little bit surprised that it’s the norm because over the last few years, the rest of the financial world has been transitioning to OTC being a lot more automated. One of the things that happened is that a lot of the OTC & ETF trading moved to automated platforms, with the Bloomberg’s RFQ being maybe the most prominent of those.
(17:16) George Manolov:
So, is it correct to say that you can only automate OTC deals up to a certain size, right?
(17:25) Sam Bankman:
That’s right. We can only automate it up to a certain size, but as it turns out this size turns out to be bigger than most tradespeople want to do; you can do a million dollar Bitcoin trade against us, that’s automated, and as it turns out, most people actually don’t want to trade more than that much Bitcoin at once anyway. At the end, I think we’ve found that this gets the vast majority of it basically. There is more work to go in automating larger trades, but this gets us to the point where +90% of the trades that we do can be automated.
(17:59) George Manolov:
So, you have the OTC desk, you have the arbitrage work that you’re doing, is there anything else part of Alameda?
(18:07) Sam Bankman:
That’s the nexus of Alameda. We also do a lot of liquidity bridging for exchanges and other things like that. Not Shocking, given the rest of what we do, but that is most of what Alameda proper is doing.
(18:21) George Manolov:
Okay. When I talk to my colleagues researching Alameda before this chat, they mentioned that you are one of the top 3 traders that is profiting on the BitMEX exchange. When I go to the leaderboard of BitMEX, it currently says that you have a profit of about 7,000 Bitcoins.
(18:42) Sam Bankman:
Yup. We’ve got a few accounts there; I think that’s about what you get if you add those up.
(18:47) George Manolov:
How the hell do you get to such enormous profits?
(18:50) Sam Bankman:
There are a bunch of things that go into that. One thing that I would say here is that we’ve put a combination of things together; we do a lot of volume, we’re trading about a billion dollars a day now, a lot of different trading strategies put together, and a fairly sizable capital base. We’ve just grown out what we’re doing to be one of the larger players in crypto, if you think about the amount that you can make off that, it’s fairly sizeable if you really do everything as well as you can, if you try and maximize all of the trading that you’re doing and the size of the footprint that you have in the ecosystem.
(19:33) George Manolov:
How many people are behind all of this success? How many people have been working on building these algorithms and following all the trading that’s going on?
(19:42) Sam Bankman:
We have about 25 employees.
(19:45) George Manolov:
That’s pretty impressive. So, this 7,000 Bitcoins in profit, this is for what period of time?
(19:51) Sam Bankman:
It’s a little complicated since we switched accounts. We’ve been operating for about 2 years. Those accounts are just over the last year or so. We actually have another account up there, which is not publicly identified as ours, it’s a deprecated one, and was one of the ones that you are seeing at the beginning, but putting all those together, it’s over the last 2 years or so.
(20:12) George Manolov:
All right. So, 25 people generating at least $50 million in profit for a couple of years.
(20:17) Sam Bankman:
(20:17) George Manolov:
Okay. Pretty impressive. Keep up the good work.
(20:21) Sam Bankman:
(20:22) George Manolov:
With that in mind, I would like to transition to something new. After Alameda, you went on to build the FTX derivative exchange.
(20:33) Sam Bankman:
(20:33) George Manolov:
So, what is it about? How is it different from other exchanges that are out there in the crypto space? If you could provide a general overview.
(20:45) Sam Bankman:
FTX is a crypto derivative exchange, launched about 4 months ago. It has quarterly in perpetual futures on about 15 underlyings. In terms of what makes it different than other offerings in this space, it depends on what you compare it to, different offerings have different strengths and weaknesses, but compared to some of the prominent ones, one of the things worth mentioning the most about it is that there has been a lot of really high profile failures of liquidity in derivatives in crypto. You can see that from a lot of different things, I think maybe the starkest is the clawback; there have been hundreds of millions of dollars of clawbacks.
(21:27) George Manolov:
Could you please explain what a clawback is?
(21:31) Sam Bankman:
Let’s say that there’s some guy “Jim” and he’s got count on a derivative exchange in crypto and he puts on a few futures position, maybe he deposits $1 million and uses that to put it on a $20 million long position in Bitcoin futures on 20x leverage. Now Bitcoin starts going down, so it’s down 1%, 2%, 3%, if Bitcoin goes down 5%, he will have lost $1 million on his position, and that’s the same amount that he started with. So, 5% is the amount that Jim can lose before he’s bankrupt on the exchange. One thing which is different about crypto from the rest of finance is that when people are trading Apple stock on the New York Stock Exchange, NYSE doesn’t have to worry about people going bankrupt because there’s lots of financial infrastructure sitting in the middle of the earth that deals with risk management. NYSE is just the matching engine and clearing of the shares.
While in crypto there’s nothing sitting in between, there is no clearing from something like that. So, if Jim goes bankrupt, in fact goes more than bankrupt; let’s say Bitcoin moves down 10%, so Jim now has a negative $1 million of net value on this exchange. You can’t really reclaim any money from Jim, that money is just lost. Jim just says: Great, I’m never sending any more money to this exchange, he runs away and you’re stuck with this negative $1 million and someone’s got to fill that. In general, what happens is that the other customers fill that; they say: all right, we’ve got to take $1 million from someone, we know that other customers made at least $2 million because Jim lost $2 million, and he has to have lost it to someone, so let’s just take it from the people who made money. And there is what’s called the clawback.
Half of those profits are clawed back, because Jim had lost $2 million, but he only actually had $1 million deposited there and the other million dollars he just can’t pay up. So, the exchange has to take $1 million back from the customers, and this means that customers get screwed over if anyone on an exchange ever goes beyond bankrupt.
(23:43) George Manolov:
How is it possible? Is it because, like you said, the infrastructure is not good enough?
(23:48) Sam Bankman:
Right. In order to stop this from happening, what would have to happen basically is that the exchange would somehow have to close down Jim’s position before he goes bankrupt. It would have to make sure that by the time he’s out of money, he still doesn’t have a $20 million position open. But this isn’t necessarily easy, it has to unwind $20 million worth of Bitcoin futures. Part of the worry is maybe it has a lot of impact doing that, maybe it drives the market down by 10% just trying to de-risk Jim’s position, and in fact that alone is lethal. So, in trying to solve their risk problem, they actually dig themselves into a hole.
That’s one problem that comes up, but another is from that they just do a really bad job of it. Maybe it’s not that they could not theoretically have dealt with this risk, it’s that they just didn’t, they waited until Bitcoin was already down 4% before even starting to liquidate him. If you do that, you’re basically giving yourself no room, then somehow you have to offload this $20 million position before Bitcoin even moves one more percent. Some of this happens because it’s legitimately hard, but a lot of this happens because exchanges were negligent; they waited way too long before starting to liquidate accounts, they didn’t have good algorithms in place to liquidate them, and they didn’t have any backup metrics.
One thing that you have to worry about is what sort of liquidity just dries up on the exchange rate when you’re trying to liquidate someone, what do you do then, are you just fucked? The answer on most exchanges is yes, you are, you’re just screwed, there’s nothing you can do. With that in mind, a lot of exchanges did quite a bad job at this, but even ignoring that, it’s a hard problem. There’s no obvious answer to this and combining those two together, there were massive issues and massive clawbacks in 2018.
That’s one of the central problems with cryptocurrency derivatives that we really hated at Alameda. Alameda paid about $20 million over the course of 2018 to settle other people’s bankruptcies on crypto derivatives exchanges because we’re on the other side of a lot of those positions, so we were the ones that they clawed back millions of dollars from now and again. So, we saw it, up close and in person, how big these problems were and that really the industry needed to find a way to solve them.
(26:21) George Manolov:
Which exchanges which have those problems?
(26:24) Sam Bankman:
A lot of different exchanges have had some problems with this. OKEx was the one that had the biggest problem with this. They were, and still are, often the biggest derivatives exchange globally by volume, and people put on pretty big positions there. They’ve reformed their risk metrics recently, so it’s a less big problem, but they were quite bad in 2018, they do things like not to start liquidating someone until they’re 1% away from bankruptcy and had a $200 million position on, at which point you’re basically screwed. They really did not make things easy for themselves. OKEx was probably the biggest offender in 2018 over this, although a lot of different exchanges had their moment in the sun, so to speak.
So, one of the bigger impetuses for us starting up FTX was that we thought that we can do a better job at risk management. One big thing that we thought about when building FTX was what products do people want? Obviously, Bitcoin futures. Sure, people want that, but there’s a lot more that people want out of their exchanges, and there’s a lot of products that proved really popular in traditional finance that crypto just didn’t have yet. One obvious example of that was index futures. One thing that’s absolutely ubiquitous in most of the world is futures that expire to an index of things. While in crypto, the equivalent of that is, let’s say that you wanted exposure to Altcoins in general, you just thought that Altcoins were going to go up, so you wanted to get leveraged long, or they’re going to go down, so you want to get leveraged short, or you want to hedge your company’s exposure to the Altcoin industry, there’s no easy way to do that right now. So, another thing that we did was launching index futures; we have an Altcoin index future, a midcap index future, and a shitcoin index future representing different tranches of liquidity of coins to make it really easy to get exposure to, on a macro level, whatever it is your company is concerned with. So, that was another thing that we built.
Another thing was that there is a bunch of popular high-profile coins that didn’t have derivatives right now. BNB was an obvious example of that, so rather than listing derivatives on exactly the same products as everyone else, we took a step back and said: what do people want futures on? And why don’t we make futures on those? That had a lot of overlap, but it meant that things like BNB or LEO were often left out. Another cool example here is Tether. As we talked about earlier, a Tether does often has a premium, sometimes positive, sometimes negative, that’s a thing which is pretty important to crypto, particularly important if you’re, for instance, someone who wants to trade on Binance, but look around and say: I want to trade on their BTC-USDT market, but I’m worried Tether is going to crash, so I’m worried about storing Tether there. So, we launched USDT/USD futures, which allows people to buy Tether on Binance and then short sell Tether futures on FTX against that to hedge your exposure to the Tether premium. Obviously, you can also use it to express your opinion on how Tether is going to do, whether you think that Tether is going to crash, you can pull the short position there, while if you think it’s going to recover, you hold the long position. This seemed like an obvious product that the space wanted which hadn’t been built yet.
(30:12) George Manolov:
These types of products that you are explaining to me sound more like it would be institutional and professional traders who would be looking for such type of services. So, is FTX more geared towards professional type of audience?
(30:28) Sam Bankman:
That’s an interesting question. The line between institutional and retail is weird and blurred in crypto. For instance, a lot of firms that launched a token, such as the companies behind EUS or XRP, are those institutional traders or retail traders? How about an OTC desk that needs to hedge their exposure? How about just some guy who bought 500 Bitcoins when they were $10 each and is now a whale in crypto and started his own family office trading crypto. Actually, a lot of crypto volume comes from things that are neither pure retail volume nor a massive multinational quant trading firm, and because of that you get a lot of definitional issues of what is institutional versus retail volume. We want to build an exchange that works for everyone, but because of what you’re saying, one of our most natural audiences is the mid-sized trader, such as the sophisticated retail trader, the small institutional player, or the large firm that isn’t at its core a trading firm, but needs to hedge something.
(31:48) George Manolov:
When it comes to retail, you also have this product, which is very interesting, which I am not sure if other exchanges offer or if you can trade elsewhere, which is called leveraged tokens. Could you just say, in a few words, what leveraged tokens are and how can one utilize them?
(32:10) Sam Bankman:
Absolutely. Leveraged tokens are cool products; they’re ERC-20 tokens, but they have leveraged price movement. For instance, you can look at ETH BEAR which is an ERC-20 token whose price action is -3 times whatever ETH price action is. You could ask how do you do that, and the basic answer is that there’s a futures exchange behind this with FTX and ETH BEAR gets 3x short ETH futures. So, it’s tokenizing a position on FTX futures, and this allows you to trade ETH BEAR like a spot token; You could just buy it for dollars, Tethers or whatever and hold it, you don’t have to manage margin, collateral, liquidations or anything like that that serves ETH BEAR, the token does that for you and you get + or – 3x price movement. So, you get your leverage exposure without having to deal with margin at all, and the token will automatically rebalance to reinvest gains and to avoid liquidations.
(33:18) George Manolov:
Okay. So, this is, I imagine, your product offering to retail audience.
(33:23) Sam Bankman:
One of the most natural audiences for this is exactly that. Let’s say that you’re someone who wants to get 3x short ETH, but you basically don’t have to worry about getting liquidated or anything like that, you could just buy ETH BEAR and let it manage the margin for you. The other cool thing about these tokens, in addition to the fact that it manages margin for you and that you don’t even have to bother dealing with all these wallets, collateral and things like that, is that it’s an ERC-20 token, so it can be moved around the Blockchain, and in particular it can be listed on other venues.
Spot exchanges can list leveraged tokens, they don’t need to find any way to implement margin, leverage, clawbacks, liquidations and things like that themselves, but their customers can still get leveraged exposure. So, it allows any exchange in the world to get access to FTX’s leveraged exposure, and we’ve already seen a few exchanges do this; CBG has listed them, and by the time this is public, the listing would have happened on GOPAX which is a Korean exchange, and we’re in talks with a lot of other exchanges about this. So, it gives us an opportunity to spread FTX leverage around to a lot of venues, and it gives a lot of venues an easy way to offer their customers leverage.
(34:41) George Manolov:
You say that these leveraged tokens are basically a safer way for people to get leverage because people don’t have to keep an eye on their margin or liquidation. If I’m convinced that over the next year or two, Bitcoin is going to go to $20,000 and I can either by spot Bitcoin right now at the current price, or I can buy BO token, which is essentially a leveraged 3x Bitcoin token ,and whenever Bitcoin reaches 20 k, could be in one month or 12, I’ve tripled my profit. Is that correct?
(35:21) Sam Bankman:
Mostly. No matter what you do, you’re going to run into the issue of what happens if Bitcoin goes down a lot before it goes up. What happens if it eventually gets 20 k, but it goes down to 5 k before that, if you just buy futures answers, you’ll get liquidated, then by the time it gets to 20 k you won’t have a position anymore. With leverage tokens, instead what will happen is that it’ll basically deleverage the position as Bitcoin goes down to avoid liquidation, and then put the position back on when Bitcoin goes back up. So, over long periods the leveraged token won’t match exactly 3 times the return of the underlying because of these rebalances. In particular, it’ll do better if markets are exhibiting momentum, and worse if markets are exhibiting mean reversion.
The reason for that is that you’ll end up rebalancing in the direction of markets. So, if it goes straight up to 20,000, you’ll actually do better than 3x, and if it goes down and up, down and up, down and up, down and up, then up, it’ll do worse than 3x. But either way, it will be better than having gotten liquidated when Bitcoin went down to 8 k, and then making nothing when it goes back up to 20 k. It’s still a little bit complicated of what exactly will happen, I don’t want to portray it as a perfect solution for everything, but it has a lot of advantages over putting on a margin position and then not managing your collateral and risking liquidation.
In the end, the perfect thing to do depends on what you want to be doing, and on whether your goal here is to be what you want to do when markets move. Do you want to be reinvesting your profits? Do you want to be deleveraging to avoid liquidation? And things like that. I don’t necessarily want to make a statement about what the perfect thing to do is, but I think that there are a lot of advantages in the way that leveraged tokens handle things, and in particular they solve a lot of the problems of how you manage margin in a world where there are 24/7 markets, you don’t want to be staring at things 24/7.
(37:23) George Manolov:
With that in mind, I want to talk about FTX’s size; in terms of volumes, you’re already generating impressive amounts of volumes although the exchange is just 4 months old. Who is generating this amount of volume? A lot of people will suspect that since the exchange is so young, how could that happen? What would you reply to that?
(37:49) Sam Bankman:
First of all, it’s not fake volume. You can compare the trading that happens on it to the trading that you see on exchanges with fake volume, the fake volume exchanges, by and large, will obviously have fake trades, while FTX’s trades will look reasonable; there’ll be trades that happen at the bid and off for the right sizes. Another reason is that we’re doing a weekly buying burn of a third of the fees generated, and from that you just back out that this is real volume with real fees being paid on it, these aren’t washed trades with zero fee accounts or anything like that. That serves another way as well; we make it public, the extent to which the trade is real.
Beyond that, why is it plausible that we’ve gotten this much volume this quickly? Well, we didn’t start from square one. Exactly. First of all, we already had an OTC desk, we knew a lot of people through that and that gave us a really natural base of people to start as the original customers, some of them actually trade in quite a large size, it doesn’t take that many customers to build up a decent size sometimes if you have a product that they really want. As an example of that, Alameda trades $1 billion a day and it does put in perspective that like one party can actually trade a fairly large amount.
Another thing that I’ll say is that, we’re offering a product that there is a lot of demand for it, there are people who really wanted to hedge about $50 million a day of trading if they had the right product for it, and we really tried to design that product. In addition to all of that, we’ve been pretty aggressively building out our user base, not just among certain institutional clients, we also have a growing Chinese community with about 5,000 people in WeChat groups, we have English language communities, we’re expanding into other countries as well, and really trying to build the product that works for as many people as possible.
One last thing to say is that even from day one, this is one of the most liquid products in crypto and I think that is really a big part of what helped it succeed. And where’d that come from? Well, it came because it has the backing of Alameda Research, which meant that before anyone had ever done a single trade on it, it probably already had the most liquid order book in Crypto and that’s a pretty powerful place to start from.
(40:18) George Manolov:
All right. As you were talking here about the liquidity, you have done your own research on fake volumes and you have your own approach to this, based on what happened with Bitwise, the report that they published, and the fact that since then there’ve been a lot of companies that have started different initiatives to identify and measure fake volumes, what is your approach to this and how is it different from other people?
(40:44) Sam Bankman:
Our approach and our numbers are quite different than what CoinMarketCap got, and they’re also quite different from what Bitwise got, we think that our numbers are right and theirs are wrong. The reason for this is that a lot of people seem to have taken one of the extremes; they seem to have either thrown the baby out with the bathwater like Bitwise did, who are claiming that nothing but the US exchanges is real and everything else is fake, and anyone who’s traded a lot of crypto just knows that this isn’t true. If you’ve traded on Binance, OKEx, Huobi, or a lot of these Asian platforms, there’s a ton of fake volume, but there’s also a ton of real creation volume. Bitwise claimed that the total volume that was traded on some of these platforms was less than Alameda’s daily volume on them, so that was a pretty clear sign that they were a bit off there.
That’s one extreme, then you have CoinMarketCap on the other hand who have just basically given up on identifying real volume. You look at CoinMarketCap, and 15 of the top 20 exchanges are fake. I don’t think it’s easy to figure out what’s real and what’s fake in crypto, nothing is easy to get exactly right, but you can at least try. So, we put together a lot of different metrics to try and determine what was real and what was fake, but at its heart, there are a few things that actually get you a decent fraction of the way there.
A lot of these are pretty simple things. One example which I mentioned earlier is, are the trades physically possible? You might not think this matters a lot, but as it turns out it does. A lot of exchanges will fake trades that could not physically have happened. One example of this is that they’ll print a trade on their order book where the trade happens at the mid-price, but trades can’t print out mid; someone’s taking, someone’s providing and whoever’s taking, presumably they’re either lifting the offer or they’re hitting the bid, one of those too. So, the trade has to happen at one of those, it can’t print at the mid-price.
Another example is that they’ll print a trade 10 times the size of the orders on the order book. There’s no offer that could’ve been lifting, those are too big. There’s a bunch of things like that where they just print impossible trades. There’s another exchange that just reprinted old Binance’s trades 5 seconds later. So, the core of what we did was just filtering out impossible trades, that’s one easy way to tell if an exchange is fake, The next thing that we did outside of this is that we looked at Alameda trades on a lot of these platforms, we actually just know of how much volume is going up on some of these because we trade on them, and we can see if we get actual real fills if we put out a ton of orders and no one ever lifts them, but when CoinMarketCap claims a $1 billion a day of trading on that market, that’s a good sign that this volume is not real.
We’ve put together a bunch of metrics like that. There are some tough cases in the middle, but we think that by and large, we correctly identified which exchanges are of real volume and which aren’t. We have a daily updating; a list of exchanges by real volume at “ftx.com/volume-monitor”. One more thing I’ll mention that we did is that we include derivatives volume. For weird historical reasons, neither CoinMarketCap nor Bitwise bothered including derivatives, but derivatives are more than half of all trading volume in crypto, and that’s not great. So, in addition to chronicling the spot volume, we also chronicle the derivatives volume and.
On the volume monitor, you can also see a report that we wrote, probably the most interesting thing there is looking at the appendix where we just took a bunch of screenshots of the ways that 50 different exchanges faked volume. So, you can browse through those and just see a lot of really bad attempts at pretending that there’s a real volume going up.
(44:54) George Manolov:
So, you have automated this whole fake volume identification process.
(44:59) Sam Bankman:
(44:59) George Manolov:
So, you’ve done it once and now it repeats this process every single minute or day.
(45:04) Sam Bankman:
Yup, that’s right.
(45:06) George Manolov:
Super impressive. Definitely interesting, and I recommend everyone who’s interested in finding the real volumes to look at “ftx.com/volume-monitor”. Sam, you also mentioned that there is a so-called buy and burn mechanism that you have regarding your token because you also have an exchange token that’s launched recently on FTX. Is this token similar to other exchange tokens like the BNB or how is it different? And what is the token for?
(45:37) Sam Bankman:
It has a lot of similarities with them; it’s called the FTT, it’s the exchange token of FTX, and like BNB & LEO, you get fee discounts on FTX if you hold FTT, and we also do a buy and burn of a fraction of fees generated on FTX. Some things that are somewhat different about it; one of them is that we don’t just burn the tokens, we actually do buy them back and burn them in the market, we do this every Monday, we’re going to start our next one in 7 hours or so. In addition to the fee rebates and the buy and burn, we also give tighter OTC spreads on the FTX OTC desk for people who hold FTT. Another thing is that we have an insurance fund, as most derivatives exchanges do, and we also buy and burn 10% of the net additions to the insurance fund, we do that, first; as a way of displaying our confidence at the entrance and that we’ll be growing over time and not shrinking, and second; to pass some of that back to the people supporting the exchange. Over time we’ll be rolling out more and more functionality for FTT. One last thing I should mention is that you can also use FTT as collateral for your futures positions on FTX.
(46:47) George Manolov:
When you’re saying there’s a buy and burn mechanism, does this suggest that similarly to BNB there is a limited supply of FTT?
(46:56) Sam Bankman:
Yep, that’s right. There are 350 million tokens, none others will ever be made. In fact, there’s less now because the buy and burns have started. So now, there are 349 million FTT and that number will continue to go down over time as each week we burn FTT equal to a third of all the fees that the platform generates.
(47:19) George Manolov:
All right. When I look at the FTT token on Etherscan, I was honestly surprised that at this point it says that there’s a little less than 100 tokens’ holders. So, what’s going on? Why is that the case?
(47:30) Sam Bankman:
The primary thing going on there is that exchange is centralized; all of the tokens of their holders is usually in one cold wallet. So, the way you should read this is that there is some number of thousands of people who hold FTT on FTX, some number of thousands of people on Huobi and BitMax, but all those tokens are going to be stored in one address for each of those exchanges. What that does is that some number of hundreds of addresses are represented, you see the number of users who have chosen to withdraw their FTT from their exchange to their own personal cold wallet.
One reason that this is going to be even less than for other tokens is because of the benefits that you get from holding FTT on FTX; the fee rebates and things like that. You’re going to see a lot of people actually holding their FTT on FTX, rather than withdrawing to their own cold wallet, because they want the few rebates. So, the vast majority of all tokens are held by users on FTX. I don’t know the exact number; I’m guessing it’s about 5,000 FTX users that own FTT.
(48:40) George Manolov:
That makes total sense. All right, Sam, as we get to the close of this talk, I just wanted to ask you, how do your days look like? Are you 100%, in all the time, on Alameda and FTX or do you also have days where you’re not really actively working on those products?
(48:57) Sam Bankman:
It’s a good question. It’s a lot of work, I basically spend all my time in the office, and where my time is spent depends a fair bit on exactly what’s going on, so on some days I spend all of my time planning out what products we’re going to be listing next on FTX, some days I’ll spend a lot of my time talking to other exchanges about listing FTT or leveraged tokens, on some days I’ll spend all of my time overseeing Alameda operation, while some days I’ll be performing some business related tasks such as doing interviews. It’s pretty much work nonstop and there’s always more things to do. We’ve built a lot of things in the last couple of years, but I don’t think that means that there’s less left for us to do, it means there’s more because we’ve just opened up even more areas of business for us to grow.
(49:56) George Manolov:
Absolutely. There’s been a video that went quite popular on YouTube where it shows how you’re live handling a 7,000 Bitcoin sell wall on Binance, are you, from time to time, involved in operations when it comes to trading and engaging in such events?
(50:15) Sam Bankman:
If one trading gets really busy, I’ll often jump on, pulling myself away from the day to day overseeing of it. But when things are busy, I’ll absolutely jump on and help out. In general, I try to divide my time as makes sense between our varied series of business depending on what’s the busiest then. When all of a sudden someone decides to sell an enormous number of Bitcoins on Binance, I drop whatever I’m doing and focus on that for a bit.
(50:46) George Manolov:
All right. My last question is how do you see Alameda and FTX a year from now? I would like to ask this question in regards to a few years or 5 years, but I think crypto is evolving so fast and so quickly that it’s absolutely impossible to think long-term.
(51:04) Sam Bankman:
On the FTX side, I hopefully see it becoming the dominant crypto derivative spot forum, building out more and more features, more and more products, and really trying to become the platform that people want to use by catering to as many different people as we can, and putting together all the needs of all the different traders in crypto. How can we tell if we got there? Obviously, there are a lot of pieces to this, but internally, we’re using ETH volume of at least about a billion a day as a metric measuring if we get to where we were trying to go. On the Alameda side, hopefully continuing to grow Alameda’s footprint, continuing to provide more liquidity and just continuing to refine our trading to get better and better over time to build out a more and more fully-fledged OTC desk, and to adapt to changing market environments.
(52:05) George Manolov:
All right. Thanks Sam. Really appreciate your time.
(52:08) Sam Bankman:
(52:12) George Manolov:
Hey my friend, you can find and Follow Sam on his Twitter handle “@SBF_Alameda”. Also, you can go ahead and register on “ftx.com” to be trading derivatives, futures and leveraged tokens, or you can save 5% on your trading fees on FTX if you sign up via the referral link, which you will find in the description of this podcast episode.
A couple of things before you take off. First, if you enjoyed this episode, be generous and share it with your best friends. They deserve it, they will appreciate it and they will be grateful to you.
Second, I want to draw your attention to Nexo, without which this episode wouldn’t be possible. If you’re like me and you believe the long-term value of crypto will grow, then Nexo might be a useful tool for you. The company allows you to get instant access to cash in 45 fiat currencies without selling your Bitcoin, Ether or other crypto assets. In case you don’t need a cash loan, Nexo can also guarantee you a passive income of 8% per year on your stablecoins, Euros, US dollars and British pounds. What’s more, interest is paid on a daily basis, meaning you can add or withdraw funds to your interest earning account at any time. You should know that the level of automation, convenience, and flexibility that Nexo provides are vastly superior to what any other similar services offer. This is the reason why I myself joined the Nexo team. Having said that, none of this is investment advice, I just genuinely believe these services might be useful for you. You can explore more at “Nexo.io”.
Finally, if you have any suggestions, feedback, or anything you want to tell me, you can find me on Twitter at “@BorderlessBTC” or on my website “Borderlesscrypto.com”. Thanks for tuning in and I’ll see you again soon.