Crypto is Open Source on Steroids with Michael Feng, Co-Founder of Hummingbot

Speaker A: This is Tim Dalhouse and you're listening to the Dao Talks podcast. In each episode you hear me talk to builders and makers of Web3. Together we'll be exploring our multi chain future, share personal stories and discuss how we're investing, experimenting and failing with the startups that define the future. So join me on this journey of discovery as I chat with these leaders, change makers and misfits about tech, life, the universe and everything. Today I'm talking to Michael Feng, co founder and CEO of Hummingbot. Michael is a former Wall street trader that fell in love with Web3, created Hummingbot, got shut down by the SEC and pivoted into open source platform. We're going to be talking about the current market, development of regulations and the future of the multi chain universe. Michael, good morning, good evening and good night. No matter where you are. Where are you these days?

Speaker B: I'm in Sunnyvale, California, right next to Apple.

Speaker A: Well great. I um, can remember this vividly actually. My older son who's not 10 was uh, born at uh, Stanford. So we lived pretty much around the corner there. Ten years ago was start of our long journey. I want to talk with you about decentralization daos, especially the future of the multi chain Web three. Let's start there. What does Hummingbot do? Give it to us in a nutshell,

Speaker B: we're kind of interesting project uh so we actually started about six years ago as a quant hedge fund. Back then we're trying to build an on chain like hedge fund protocol. Long story short, we kind of got into hot water with the SEC inadvertently so we had to close down our fund. But then we open sourced our stack which uh, became Hummingbot. So we didn't know what else to do so we open sourced it. But after a few years we actually did kind of find um, a way to sustain a business and build a business model around it. So where we landed is actually with two entities. One, one is a token foundation that basically kind of administers this open source project, works with the community and primarily interfaces with developers getting them to build and maintain various parts of the code base. And then we have a sister company, Coinalpha, my co founder still run that basically power the ecosystem. They uh, build integrations for companies who use Hummingbot. They uh, also provide market making services for projects who kind of want that with Hummingbot. And also they operate a platform called Miner which kind of lets individuals earn rewards by running Hummingbot. The model we kind of use is like Red Hat Linux versus like Red Hat versus Linux where you have this open source project, but also a company that was instrumental in commercializing it, making it more useful for enterprise customers and enhancing adoption. And I think hopefully over time will be measured by how durable this ecosystem is, not by how high in price our token is. So our goal is really to build like the Linux of quant trading yourself.

Speaker A: You come from a background in financial industry I think from what I know and give me a bit of your background and when to think about this idea of like a hedge uh, fund and uh, you know on blockchain. I'm really interested in this moment of inception, the initial ideas and drivers.

Speaker B: Yeah, the thing with me like I'm like a huge finance nerd. I was a kid, my dad worked on Wall street but I was like into computers and like playing computer games. And so I read this book called Liar's Poker. This is like more than 20 years ago about how people created these like structured residential mortgage backed securities and you know, all this other stuff. And you know most people found it super boring but uh, I just got sucked down the rabbit hole for some reason. I joined this group out of college on Wall street in uh, 2001. My first day was a week before the 9 11, uh, so I also lived in that area. So it was just a total unusual entry point into my career. Ended up was I created These derivatives called CDOs like collateralized debt obligations. And um, my claim to fame is that I structured a bunch of the ones that were backed by subprime mortgages and unfortunately had a hand in blowing up the economy in 2008. So after that honestly I was actually really depressed because I had actually accomplished what I wanted to do in life which is be financially successful banking and all that kind of stuff. But I just felt like I had done all this harm to the world by costing a lot of people their jobs, their homes, you know, due to these like securities I created. So I kind of had this like come to Jesus moment where I was like, you know, what's the point if I'm not building something valuable? This is 2010. At that point I was like, I quit my job, I got a master's degree from Stanford in engineering. Then I've basically been in barrier ever since, like trying to start companies and you know, stuff like that. Ah, in both like the normal Web2 space and then eventually found my way to Web3. And I think it's always in search of like trying to actually add value by building stuff instead of trying to like just you know, make money for

Speaker A: the help and Then specifically, at some point, what you said earlier, okay, yeah, built this hedge fund on chain. What was it that you saw very specifically, or I mean, I imagine you explored the space and you saw, hey, there's a lot of stuff happening. First of all, what year was that?

Speaker B: This is 2017. But really the motivation was because I spent so long finance that I kind of understood what it took to create, like, securities. So the process of creating your security, I mean, it's like a pain in the ass because getting lawyers, like write all this stuff, then you have to make this thing tradable. You get traders to do it. And so all that took lots of people, millions of dollars, typically, and just lots of time. And that compared to creating a smart contract that anyone can just push a button and just launch onto the blockchain is like night and day. So, I mean, to me, that's what innovation looks like, where you can take things that were used to be expensive and inefficient and distill them down to what their essence is and then make that digital and accessible to everyone. So when I say about finance nerve, what I mean is when I look at the history of finance from the time that people used to exchange shelves with each other to the invention of a joint stock company in the Amsterdam, the 1600s, to the invention of the credit markets, to the change from hard money to the Bretton woods, eventually to the fiat financial system we have today, that was the first couple thousand years. And now we finally are taking these systems, putting them in digital form and making them more decentralized and accessible to more people. And you're kind of, you're already seeing this because, like, you know, right now it's like there's all these virtual economies being created in both crypto, but also in the fiat world with like, you know, V bucks, you know, as Fortnite Eve Onlines, coin. So I think it's kind of like what we're really doing here is we're taking like, the way the society allocates resources and which is inventing new ways to create money and to exchange these digital assets with one another. So to me, it's like when I look at the history of finance, crypto is just another evolution of that technology of finance is improving.

Speaker A: And when you said vbucks, I assume it's, uh, your son or daughter also asking you for your credit card to buy vbucks. I have this, uh, every week as well.

Speaker B: For her, it's a Roblox one, but similar. Right? These are all just digital currencies we're all using today.

Speaker A: Very interesting. And so when you said, hey, you're going to create this hedge fund on chain in 2017, this has obviously evolved in five years. We likely have a totally different perspective. But where did you see like initially the biggest opportunity? How did you want to tackle it? And where did you get into trouble with the SEC at that point?

Speaker B: Oh yeah. First what happened was, you know, we really wanted to attack the parts of the hedge fund system that were the worst and um, the biggest pain in the ass. And that's typically setting one up and like doing all the legal work and like the accounting work and hiring the custodian administrator needed to actually set up the hedge fund. I mean this is kind of why you have to raise like $10 million before you even get started, because you have to pay these idiots, right? So we were like, why do you need all that? Just launch a smart contract and get your friends to like deposit some Ethan smart contract, like, what's the big deal? And so, uh, obviously that was pretty naive, but that's basically what we did. We actually created a smart contract. It's open source, it's called fund protocol. It's a coin off of GitHub. We didn't see as a protocol that we want other people to use and adopt as a way to create hedge funds. The way it worked is it's smart contract. We set the ETH and then we had the rules of the hedge fund, like the fees would be calcul, how performance calculated everything in the smart contract. But the way it worked, and this is kind of why it was differentiated from other kind of protocols at the time, like Melonport, was that we basically assumed that the manager is going to be trading on exchange. So there was an Oracle layer that would basically pull the balance from a Coinbase or Gemini account, you know, use an API and use that uh, daily nav at that nav price. And then attached to that, we had a strategy that my CTO Martin created that use machine learning. They try to predict the price of ETH and Bitcoin in the next 15 to 30 minutes and then kind of automatically trade it on that basis. So we ended up running that fund for about a year. It uh, actually did pretty well, like made 20% of net of fees. In a period where ETH and Bitcoin lost like 10%. We had like a probably 30, like 20, 30 investors who are all like individuals and mostly our friends. Um, but we talked about this protocol a lot and then sec one day we got A letter saying that they thought we had done general solicitation, which basically meant that, you know, we had a website and a contact form and potentially some mom and pop investor could have, you know, found it and invest, even though we had like a very strict accreditation process. But anyway, my lesson learned is here is it more like if you can just build technology, as in like just build technology, build software. Because software and technology is not regulated, right. When you start actually building funds, exchanges, banks and stuff, that actually holds money for people and does all that stuff. And that's where the regulations come in. And so I'm glad we went through that because what I like to do in life is build stuff. What I don't like to do is deal with lawyers. And so that's kind of informed my choice personally of working on open source as opposed to working at a hedge fund or an exchange.

Speaker A: What's this passion that you now have for building stuff and building it openly? Something that you had already before or something that you really discovered in this Web3 community?

Speaker B: I would say I've always had it, but I think uh, I found like my home in the web 3. I'm like the kid that I built my own like Monopoly game when I was in high school. Like, you know, it's kind of in past scout I was like, I always like to build games and stuff like that. But I think even when I was doing CDO is like my cell models were like, you know, to me it was like a game of like kind of gamifying an agency, rating agency criteria. What I love doing is build like games and tools that allow people to exploit games, I'll call it that way. So in a previous life maybe I would have been like arbitraging or been a craftsman or something like that, but I feel like that is what I like doing. But overall I think it's a great fit, I think for what crypto really is. Because crypto to me is an evolution of the open source movement that Silicon Valley really started like 20, 30 years ago. And uh, that's why every single protocol is open source. And I think how well they attract developers through their open source tooling, all the ecosystem around it, the community, it's really open source on steroids. And so that's why I think there's actually a very strong evolution by open source.

Speaker A: Yeah, like crypto somewhat being the business model for open source. Right. I put it on steroids obviously, very clearly. And you said that love to build stuff and where regulation kicks in just to open that can of worms is then when you handle other people's money and not on building stuff. Except for what we had with Tornado Cash, I assume, which set a big present. This is obviously a topic we could spend a series of podcasts on, but how do you see regulation evolving specifically? You know, I mean, you've been in this for a long time. You're very passionate about it. You've been on both sides. You've been in financially, you've tried to build something with the managing other people's money on chain and then moved into the actual software side. What do you see and what do you believe will happen in terms of regulation in the US from your vantage point? Right.

Speaker B: So I think it's hard for me to predict what happens. But let me try to walk you through how I think about, like the players in the game and kind of what their incentives are and how it might shake out. So I think what we're seeing here is kind of like what I call the use of religious analogy, the rise of Christianity during the Holy Roman Empire. Because for a long time you had Rome, which controlled all the religious scenes. And it was like, I'm sure they have all these different gods. And then finally there's one religion comes along that says, hey, there's only one God. And of course the Romans react negatively, right? And so for a few hundred years, Christians were persecuted heavily by the Romans. But eventually what happened was the idea of Christianity is really. And that's why it's different from all the previous religions before that was that it was very much about the people. It was a very decentralized religion compared to previous ones, where it's like, hey, you know, it's like only the emperor was the gaga. So that's why I think in the decentralized, ultimately it appeals to the people. And so eventually what happened was this one Roman emperor was like, hey, if I want to win the hearts of mindful people, I gotta become Christian, right? And so Christianity ended up invading Rome despite the fact that, uh, at the very beginning it was much less powerful. And that's because there were two different things. Rome was an institution in one place in time. Whereas Christianity was a technology, it was a religious technology. It was superior to other religions before that because it was more decentralized and was kind of more open. So Christianity carved the path for Islam and for kind of like more decentralized visions that exist today. So anyway, that parable is just a way of saying, I think when you think of the US or even the SEC versus technologies, like blockchain it's actually two different things. And so I think when blockchain crypto, whatever happens with regulation in the short term is not going to change what happens in the long term. Which happens in the long term is the decentralizing technology behind crypto is, is going to eat the financial infrastructure globally, us, China, whatever. And we'll have a more decentralized, maybe hybrid systems and then other ones in the future. But what we're seeing right now is the SEC clearly believes that pretty much every crypto being created today should fall under their purview, as in those securities. Whether they can force that, we'll see. But honestly, uh, I don't really care. And I don't care because I've specifically put myself in position where I create software that people use to access different exchanges regardless of where they are. And it's up to them to comply with their local laws and jurisdictions. And it's because I went through an experience where I felt like I was forced to like, deal with this bullshit. And I never want to do that again.

Speaker A: Very good analogy. Okay, so, you know, talking about this technology, religious technology, and projecting forward as it, you know, is here to stay, it's inevitable, so to speak. Right. Let's talk a little bit about perspective on how you look at this. Different networks and chains. Right. I would imagine 2017, when you started, you likely looked primarily at Bitcoin and then opened your perspectives onto Ethereum. Um, first of all, where would you put yourself today between Bitcoin maximalists, Ethereum maximalists? Give me insight into your head here.

Speaker B: Got it, got it. So I'm neither. If it matches anything, it's probably crypto. When I say crypto, what I define is I would say to me, there's a very specific difference between these crypto systems and the previous quote, like Fiat or other systems before that, whether it's EVE online or whether it's dollar or whatever. And that's, to me, it's the difference between what I call analog versus digital financial networks. So to me, an analog network is where the transition happens, but there's no broadcast of that thing across the entire network. So what I call a digital financial system is one where whenever a transition happens in that system, it's propagated across the entire network. So state zero to state one. Everything flips in that one state. So that's what we call blocks. But the need to have something like blocks, where you're basically changing state from one network state to another, is actually necessary for all networks. If that were available in let's say the Fiat system when the 2008 crisis happened, that would've been great because we would finally have uh, figured out where all that Lehman counterparty risk was and then people, you wouldn't have all this bullshit afterwards. But that's hard because in order for that to happen, everyone kind of has to agree to follow that rule. Because the reason the Fiat system doesn't have that property is because anyone can pay anyone, right? There's no record of that payment in some global database. So I think if you think about this digital versus analog dividend, I think of Bitcoin as the first true exemplar of a true digital financial system where you have this every transaction is reflected across the network kind of property. However, you can't do anything with the Bitcoin. Why? Because there's no programmability into it. It's like Bitcoin only works if we're in a village and we have shells and we're just exchanging shells. It doesn't scale because the modern financial system has things like credit derivatives, things that require some type of logical programmability, just like finance. But you start with the shells, you start with a really basic exchange type economies, and then you move into the more complex economies, uh, that are only possible with programmability. So to me, Ethereum is the next example. Okay, now you have another economy chain created, but of course it's slow, right? So if these trade offs now you have these roll ups and everything else. But this is like the same thing that happens in the traditional finance world as well. For instance, when a company launches, they can list their stock in the New York Stock Exchange or they can list their stock on the Hong Kong Stock Exchange. Now most companies, if you're multinational, you'd prefer the New York one, but New York one might be hard to access, has higher fees, has other properties that might prevent you from listing there or maybe like the A share exchange, give you some discounts or whatever. Now you list there or you might want to list both. You can do a listings as well. And I think you have the same paradigm happening in blockchains as well, where protocols are fighting for tokens. Tokens can choose which protocols to list on or to go on. And then you have this competition between both the tokens, the exchanges and the protocols. And so they're all kind of like different aspects of the Venn diagram that are all kind of like cooperating but also competing with one another. So to me, actually I guess the metric is, I kind of look at long Termism like just total market cap in crypto, like total, you know, what is the total amount of market cap? Because that's more like to me is like how much value has been transferred from the non digital financial systems into the truly digital financial systems.

Speaker A: And you know, I think when you go into CoinMarketCap, there's whatever 200 blockchains today and you could make multiple arguments, you could say hey, you know again, bitcoin maximalist is only going to be bitcoin everything. Listen, we have to shell situation which I personally agree with, but everyone can have a different opinion and you could talk about upgradability of bitcoin network and whatever, but let's leave it there. Then second you could say hey, you really just need Ethereum and performance is going to go up, gas cost is going to go down. Maybe have some of the side chains have scalability. Obviously it's part of the strategy. Do you need the other 198 chains? You could say no or you could say maybe it's going to be 2000. You know, you could.

Speaker B: But then if you look at the new chains being created, they're modularized, right? So you could use the Solana vm use like Celestia for data availability and then pull in like, you know, I don't know, whatever Ethereum for sequencing. It's like these terms are not fixed. Like what we call blockchain today is very different from what a um, blockchain was two or three years ago. I think the maximalist, honestly I think the word maximalist just feels weird to me because I'm like, you're in this game because you believe in technology. Do you think technology is not going to change? So I just don't get it.

Speaker A: I agree with you. And you know, do you think what we're going to see is rather 200 chains and less or do you think when we project out five or 10 years we have a very diverse ecosystem of chains? How do you envision it?

Speaker B: Uh, I hesitate to predict in those terms because I'm not sure that the word chain is going to mean the same thing today as it will 10 years from now. I feel more confident making more like what is the effect of crypto in a 10 year period versus where it is now? Because I think a lot of people are like what is crypto gotten for us kind of thing. Right. And so I feel pretty confident predicting in 10 years from now the assets we use on a day to day basis whether it's like the money that we're sending around or the assets we're investing our ETFs into or those kinds of things. I think those assets will be much more digital than they are today. And when I say more digital what I mean is there might be tokenized versions of USD, you have your wallet, you might have some weird NFT that's actually like an ETF that's investable but it looks like a board ape for some reason. Your savings account might have some weird like you know, yield features. So I think that these digital financial instruments will get more and more embedded in our daily lives. And ah, in certain industries like gambling, we completely tokenize so to speak, which is like honestly they should be because like what we're calling tokens are just ways to encapsulate value. And the easiest types of value to encapsulate are ones that are purely digital. Right? So like a bet between two people on a game or a right to some musical instrument or even right now you're seeing lots of athletes do name image like this that's kind of extended to the AI world. Uh, that stuff is actually really easy to tokenize as well. So I think that to me tokens are just like easier ways to wrap digital value than traditional ways to wrap digital value like SPVs. And so it's natural that they'll take over because everyone likes to save time and money and uh, they will allow people to save time and money once the technology becomes easier to use.

Speaker A: So given that you're obviously a believer in this big transformation and digital financial space decentralization, interpreting what I see from your product, you currently likely the majority of your functionality is usage of centralized exchanges, right? You must have a very particular vision and idea about what is today working really well in Defi. And Defi is doing really really well. But also about what the shortcomings are and what has to happen in Defi for this to be massively adopt. Right? As we said, this is going to be where Rome adopts Christianity. And there's clearly still a uh, disconnect between what we believe it's going to be and the situation we have. So I'm really particularly interested in this. What do you think is going well, what's not going well and what do you think has to happen here?

Speaker B: The first thing I would say, I'll just take note that anything new, especially in finance just takes a lot of time to roll out because it's like if you think about it, it takes a long time for a new type of market to become widely used to the point where everyone's really trusts deploying lots of capital into it. So I would say what's interesting, I think, is that there's been quite a lot of adoption of new things invented by people in crypto in past like 10 years. So to me it's a rollout of these new technologies and how they're being adopted into the market at scale. But just like it took the perpetual swap seven years essentially to achieve dominance as a trading instrument, I think it'll take it Amazon, you know, just as long. It's more like achieving a certain critical mass until the next new thing comes along and makes it better and those kind of things. So what is not going well, I would say, and this is an unfortunate thing with crypto is because money actually slows people down from an innovation standpoint. Because what happened was when the success of like Uniswap AMMs first were created, people were like, oh my God, you can create these decentralized exchanges and actually works. And then all this money poured into a sector. You had all these clones, people just copying what Uniswap doing, adding very little value. And then you had all these food tokens, the Solana tokens that went up and down. And so that whole cycle actually didn't result in any real product innovation. It just resulted in a lot of money being invested and lost. And unfortunately, I think this is actually the case with pretty much all technologies, which is if you actually want innovation, it happens when all the money is kind of sucked out of system and then people are just like, you know what, let's just build something new, try it out, uh, and eventually these things work. So what I'm actually the most excited about right now, I would say, are all the new perpetual dexs that are being launched. So you have the AMM types like gmx, but then you also have dydx, uh, but then these other new ones that are on Cosmos Injective, then you have these new order book ones like Dexlot on Avalanche, and then you have all these other models as well. People are trying like zigzag on Zksync and I think it's a good time for this kind of stuff because post ftx, all the quants, all the traders now they realize the risk of centralized exchanges. So we're starting to see a lot more activity in these perpetual indexes. And even in Hummingbird, we've seen usage of our perpetual connectors probably um, double in the past three to six months. And I think on both the centralized side and the decentralized side, the further Use of these perpetual futures instruments is going to start to move trading from the overall Wall street version into crypto exchanges.

Speaker A: Last question here. So what are you most excited about in this web 3 and DeFi ecosystem of the things that are in development right now during the bear market and what's coming next at Hummingbot and how that fits into the context? What's keeping you when you, you know, setting down your computer and you go like, hey, yeah, uh, you know, I should work a little bit more on that or you know, if I had a few more minutes, that's what I would spend time on.

Speaker B: Yeah, of course my answer would be governance actually. Because what governance really means is for people to trust that the system works well. They have power. And it's like if I own this token, like whatever, HBOT or uni or eth, whatever, I can actually quote, govern with it. Now the reason this is hard is because usually individuals don't want us to govern. No one likes voting or staking. So you have to kind of build this compensation mechanism into the process. And so what I'm also trying to do is trying to make the rules of governance clear, try to make the impact that people have with governance bigger and also try to create mutually reinforcing incentives, token based but also social incentives, status based incentives that get people to contribute to the system in various ways so that the system is self reinforcing and grows based on that. Unfortunately, I think that you can also do it incorrectly, kind of like create incentives and processes and things like that that reduce people's trusted system. System. The simplest way is to sell a lot of your tokens to a VC because then what you're telling the community is well, you don't really have power because fund that bought a bunch of tokens can outvote you anytime you do anything. So why should you even get involved, right? So I think there's that balance where you need to raise capital for your project. You need to sustain yourselves over that initial phase until you get sustainable. But recognize that the more capital you take, the higher the value, the more you're preventing your community from enjoying that episode.

Speaker A: I think that's a good ending note here, specifically thinking about everyone else that is building things in web3 to be mindful about this balance because you obviously need money to start stuff right, and it's not possible without it. But getting this balance right, as you put it in the future, especially when it comes to governance and the power distribution, is obviously fundamental. Michael, this was a awesome talk. I learned a lot, took a bunch of notes. I hope, um, everyone that listened in has the same. Thanks so much for your time.

Speaker B: Thanks Tim. It was a pleasure.

Speaker A: Dao Talks is brought to you by Grindery. If you enjoyed this podcast, consider subscribing to Dao Talks on Apple Podcasts, Spotify, Google, or any other platform you fancy. To find out more about Grindery, visit Grindery IO. Uh, thanks for joining me, Tim. Out.