Kyle Samani & Tushar Jain

Multicoin – Emerging Frameworks for Cryptoasset Investing

That decentralized networks represent a massive investment opportunity is no longer a controversial view. In the last year alone, over 200 funds dedicated to investing in cryptoassets have been created.

But the principles and frameworks to understand this new world are still in its infancy. One fund at the forefront of advancing this understanding has been Multicoin Capital. Their Founders Kyle Samani and Tushar Jain joined us to discuss some of the concepts they use to invest in decentralized networks.

Topics we discussed in this episode
  • Google Glass and the origin story of Multicoin Capital
  • Why the most used smart contract platform will produce the winning store of value
  • Why they are bearish on Bitcoin
  • Why money is best thought of as an adjective not a noun
  • Differentiating between work, payment and burn-and-mint tokens
  • Why work tokens capture network value better than payment tokens
  • The future of Multicoin

Meher Roy: Today, we’ll converse with Kyle Samani and Tushar Jain who are the co-founders of Multicoin Capital. A thesis-driven crypto hedge fund. Multicoin Capital has become well known in this space for a lot of blogs that they write, and these blogs have insightful ideas guys. Kyle and Tushar, welcome to the show.

Kyle Samani: Thanks for having us, Meher. We’re big fans of Epicenter and really excited to be on here with you.

Tushar Jain: Hey, Meher, Brian. Super excited to be on.

Meher: Cool. So as usual, before we start discussing Multicoin, tell us a bit about your background and how you got to be involved in the cryptocurrency space. Starting with Kyle.

Kyle: Yeah. So I started programming and getting into technology when I was pretty young, probably around 10 or 11 years old. I went to NYU to study finance thinking I want to go work on Wall Street. Within my first year of college, I realized that was not my passion, that I wanted to be working in technology. After college, I ended up coming back home to Austin, which is where I’m originally from and worked at a health IT company for a couple of years along with Tushar. He was at the same company. After about a year, I got frustrated and left. And in May of 2013, I started my first company called Pristine. Pristine built software for Google Glass for use by surgeons. Yes, that Google Glass. That one. I was at the Glass hole running around, built that company up, grew it to a few million of revenue, raising of capital to kind of did the startup thing. Then Google pulled the plug on Google Glass. At which point, I had a lot of problems. I ultimately ended up pivoting the company and leaving the company and the company got sold. I found myself unemployed in January of 2016 and didn’t know what to do with myself. Discovered this thing called Ethereum in March of 2016. I was particularly drawn to Ethereum because no one could pull the rug out from under me after my experience with Google Glass and experiencing the pain of the platform kind of disappearing under your legs. I was particularly drawn to the whole decentralized thing. So that’s kinda how I got into crypto. Tushar and I have known each other since college. We met there. Were best friends and we kinda got into crypto together. I’ll let him tell his story.

Tushar: Cool. Thanks, Kyle. Yeah, so my background is more in the finance world. I also studied finance at NYU. I was an investment banker very briefly at Credit Suisse covering healthcare. Realized that the investment banking world was not for me. I didn’t really enjoy it very much. So I decided to leave that and actually moved to Austin about six years ago. And Kyle had convinced me to come to Austin and joined the healthcare IT company where he was working. That company was called VersaSuite. It was an electronic medical record provider. And after working there for about a year, I realized that I was much more interested in the data that was being created out of these electronic medical records systems than I was in the software itself. So I left in 2013 to found a company called ePatientFinder where we used data science techniques to match patients with clinical trials based on their electronic medical record data. That’s actually a really interesting data science problem because the data points that are collected in your day to day doctor’s visits are different than the data points that are necessary to determine your eligibility for a clinical trial. So that was a lot of fun. We did pretty well, raised some venture capital, had about 2 million patients interacting with the system, et cetera.

And in the meantime, I heard about this thing called Bitcoin back in 2013 and I bought a couple, literally just two. I wish I’d bought more. I think everyone does. And I bought these two Bitcoin thinking of it as a tuition check. Really, I wanted to have some skin in the game, so I would go and learn a little bit more about the technology. And as I learned more about the technology of Bitcoin, it was very interesting, but I really saw it as being inherently limited because it could only do one thing. The idea was a huge idea to have digital peer to peer money, but It wasn’t from a technological perspective as interesting to me at the time. And only in 2016 when I saw the Ethereum white paper and saw how this blockchain or distributed ledger technology could really enable the world to move towards a more decentralized paradigm where all of these businesses that have a rent seeking middleman are extracted that I got much more interested in crypto. Started investing really heavily in 2016. And in 2017, Kyle and I got together and realized that this was absolutely going to be a world changing transition and that we needed to be a part of it.

Brian: So once you decided you wanted to move into the crypto space, why did you decide to start a fund as opposed to a startup or something else? So maybe use some of your medical tech background that you both have.

Kyle: If you think about white collar jobs, there’s broadly speaking, two major categories of white collar job operator and allocator. Tushar and I have both done the operator thing and we’ve learned a lot from that experience, but I personally find my interest and passion has always been and just synthesizing tremendous amounts of information and then figuring out kind of what to do around it. That aligns more to an allocator skill set that an operator skill set. Tushar leans more to operator than I do. I like to, basically, if you could just give me books to read in all day and then write about it, I would love that job. And that aligns towards the kind of allocator skill set and I’m also happened to be also the extroverted, which is good to just meeting people and talking to people. So when we thought about, what are we going to do in crypto? We thought about building versus investing and we don’t—for me, certainly, I lean towards the allocator side. I think Tushar leads that way as well although he has more operator tendencies than I do.

Tushar: Yeah. To add to that, I mean, once I really started to wrap my mind around crypto, I realized it was combining my four absolute favorite things in the world, which are going to be cutting edge technology, new financial models, a healthy dose of economic models and also top that off with a bunch of game theory and understanding how the various incentives play out amongst the network pistons. So it was the most intellectually fascinating thing that I’ve ever done. And so being an allocator and running a fund allows us to get a lot more breadth in terms of being able to go look at every single project out there, you know, full time job is literally go and do research and understand how everything out there works and be on the cutting edge of how all of the projects operate and how everyone is thinking versus being a creator kind of is limiting to your own scope and going deeper. And at least for me personally, I was much more interested in going wide right now and going deep in the future.

Brian: And in what way could you see yourself going deep in the future?

Tushar: I won’t rule anything out. [Laughs] But it would be pretty cool to help move the crypto ecosystem forward in more ways than just investing.

Meher: So on your website, you position yourself as a thesis-driven cryptofund. What is your thesis?

Tushar: That’s a great question. And we don’t have a singular thesis per se. The theses are really more like a product market fit or a market that is uniquely enabled by the benefits of crypto. So this distributed ledger technology has three inherent strength from our analysis. One is that it is permissionless, another is that it is decentralized and therefore, censorship resistant and the third is that it’s trustless. And that everything that we invest in needs one or more of those attributes in order to exist. If there is a project out there that does not need one of those three attributes, then we think they should use a regular database and not go and mess around with distributed ledger technology. And so that’s really one of the theses is making sure that whatever we invest in is uniquely enabled by crypto. And we have a handful of examples of specific markets where that is the case.

Brian: So you guys started the fund in August 2017 about. Now what we have seen, I remember we did an episode with Polychain, I don’t remember when it was, but quite a while ago. Among the first, I think there was MetaStable and Polychain, that were started in 2016, I think. And since then, we have seen an explosion of funds. I don’t know how many crypto hedge funds they are, but it’s probably in the hundreds. How do you guys view this development, what was the market looking like when you guys started and how has it gone since and how does that kind of affect your strategy?

Kyle: Yeah. So we kind of made the decision in May of ‘17 to do this. There are really still the only two funds of substance. Polychain and MetaStable and Pantera. Those three. We did not foresee the number of funds growing this quickly. Over the summer, we started talking to lawyers and getting all that stuff together, kind of the initial fundraising. By the time I remember, I think it was the day before or the day after our fund went live, Forbes posted an article saying 15 new crypto hedge funds are launching. And I was like, wow, we are late to the party. I forget who was in that list. I think Polychain was in the list and MetaStable. I forget who else. I think now the official tally, if you go on some of the hedge fund databases, I think it’s something like 200. So it’s grown a lot since then.

I don’t have a super specific breakdown, but I can tell you just kind of what, how would I see in field from conversations with people in the ecosystem. I sense that a substantial majority of those funds are in New York and there were on mostly by finance guys or derivatives traders. So these are guys building quantitative models. They’re doing market making, their running arbs. They’re relatively short-term focused, definitely not technology oriented. I think that’s the substantial majority of the funds. Most of the funds of the West Coast are more run by tech people. They think more like venture capitalists than do like a traditional hedge fund manager. So we’ve kind of seen that just major bifurcation happened. We have really carved out our niche and we’re super happy with. And that’s kind of, as Tushar describes, having venture capital economics with public market liquidity. I’ll let him describe that.

Tushar: Yeah, for sure. So venture capital economics with public market liquidity is a really interesting concept because the things that we invest in have the attributes of a venture capital investment in that and they are really high-risk investments. They are investments in early stage technology projects or companies that potentially have a lot of upside. And obviously, the venture capital portfolio is based on the fact that you can only lose 1x your money and you can make 100 extra money. And that’s why VC is a profitable business at all. So we think that the fact that all of these assets are liquid really changes the game and actually makes it even more profitable because you can go and change your mind at any time. Unlike a typical VC who only gets to choose entry prices and doesn’t get to choose when they want to exit or when they want to invest more. We get to choose all of those things. So we can invest in something and if we get new information that changes our mind and says, oh, actually a competitor is going to beat this out. Or, oh, actually this doesn’t make a lot of sense. We don’t think it’s gonna have the returns that we think, then we can go and sell that asset. Or if we get information that makes us more bullish, we don’t have to convince the founder or the CEO to go and take another round of capital. We can just go buy some more and actually doubled down on our position. So we think that the public market liquidity really enables both risk mitigation as well as helps us, you know, generate more upside as investors when added onto the already quite lucrative venture capital economics.

Kyle: And to add a little bit more color there. If you look at a traditional venture capital portfolio, you have between 20 and 40 companies that’s pretty standard. The reason that’s kind of the normal range is that, you know, you want to catch a winner, and if you only have five bets, it’s just a low probability that one of those five ends up being a really big winner. The problem with venture capital of course, is once you make the investment, the investment becomes liquid and you got to wait 5 to 10 years to get it out. For us, because the investments are liquid, we believe that allows for more concentration. So rather than having 20 or 25 or 30 different assets in our portfolio, today we actually have under ten. I think we have seven or eight at the moment. We consider core positions and those comprise the vast majority of our portfolio. We have a handful of pre-ICO positions as well that comprise I think six or seven percent of our portfolio today. But the vast majority of our portfolio is liquid.

The reasoning is if we have the conviction to invest in an asset, we were not expecting that there’s something that comes out tomorrow that causes the asset to tank. I mean, it’s possible. There’s like a bug in Ethereum or Monero that everything goes wrong. But in general, those are very much edge case events that would cause an asset to crash that dramatically relative to the rest of the market. And we’re obviously very plugged in. We’re full time investors in the space. You know, everything we invest in, we make sure we’re close to the teams we work with. We make sure we value accretive to them as we just have a sense for like the vision, what’s going on, competitive dynamics, et cetera. And so if we’ve changed our minds about something even if we’re wrong and we’re down 20 or 30 or 40 percent, we’re not down 100 percent and we expect that we will be able to get out before an asset actually goes to zero, in the event of you know, ultimately something goes towards failure. So because of that, the fact that we can recycle capital, it allows us to be more concentrated and actually like doubled and tripled down on our winners.

Brian: Could you also see yourselves going short at some point and taking short positions?

Kyle: We are formally a long-biased, stage agnostic hedge fund. So our portfolio is generally long. You know, right now, we’re exploring derivatives, using CBOEs, CME, LedgerX as Nasdaq and that stuff comes online in the next few months. We’ll be exploring those as well. So absolutely, we will use derivatives. We’re not actively day trading derivatives. That’s not our expertise or our forte. We are tech people first, but we can definitely use derivatives to hedge risk in certain ways, and we will do that.

Meher: One of the unconventional views that you have espoused in public forums concerns these store of value market. So the key narrative in this market is money is a medium of exchange, store of value and a unit of account. And there’s this idea that the Bitcoin system and the Bitcoin token can just succeed as a store of value coin. And for a store of value coin, what is most critical is that a lot of people should accept it as a store of value and that would create a network effect. You differ in this opinion, right? On how the store of market value will evolve. So tell us your thesis there.

Kyle: Yeah. So if we look at store of value, medium of exchange, and unit of account, those are three kind of independent functions. They are related, and they can impact each other in terms of how things play out in the real world. If you look at the history of kind of time, there’s different goods and services that have acted as different, got different combinations of those three variables. I believe the network effect of money, meaning money being like general-purpose currency. So it is a medium of exchange and a store of value and the unit of account. The network effect of that device or that assets, I believe is measured at n squared in terms of the number of users who use the system. In particular, the fact that it’s a medium of exchange and that people will use, you know, accept it, is what gives it the network effect, right? The whole point of Facebook and phone calls and fax machines is that it’s the people who are on the other side that make it valuable. A store of value in and of itself does not create the network effect. Okay. The medium of exchange does. So when people talk about the network effect of Bitcoin being so high and so large that nothing could ever pass it. Hey, we have very substantial and vehicle proof today that that’s wrong, although Ethereum hasn’t passed Bitcoin. The fact that Ethereum has even approached Bitcoin suggests that that narrative is fundamentally false.

But then secondly, if Bitcoin exists basically is just digital gold and it has a vault somewhere and then you take it out of the vault, every so often you’ve got ten Bitcoins and hey, I want to buy a house so I withdraw one Bitcoin and buy a house, and then I want to pay for my groceries and I withdraw, you know, five Satoshi’s and pay for my groceries or like withdraw that and then I have it in like Litecoin or Ether or whatever, you know, something else, and then pay. That narrative to me does not imply a network effect of n squared for Bitcoin. That implies to me a network effect of login. What matter is if you have Bitcoin and you want to withdraw into some other thing and then use that as the thing to pay for stuff, is that you need to have liquidity to get your Bitcoin out of Bitcoin. So you need a marketplace where it’s being traded. Fortunately, we already have exchanges. Coinbase, Bittrex, Poloniex, all these places. Daily volume on Bitcoin is like, I think Bitcoin to Fiat pairs is somewhere on the order of $10 to 15 billion per day. So I mean, there’s real liquidity there. And so once you’re out of that liquidity, then I don’t see what the additional opportunities for Bitcoin. And so my sense is the network effect for a fewer store of value is closer to login than n squared and that the network effect for something that becomes general purpose money, where you both store it in your wealth in it and you spend it and other people on the other side tickets, that network effect ends up approaching something closer to n squared.

Brian: Yeah. And I’m curious here, maybe just jump in here of a question. So there’s certainly this idea that a lot of people have and the future we will have millions, billions of tokens and that there will be all these seamless decentralized exchanges, or these basically protocols that allow you to exchange any token for any other tokens, you know, hopefully, you know, we’ve had friction in very cheap, right? So I have token A, you know, you only want token B, but you know, I basically just pay that amount, it gets exchanged to you. Wouldn’t that kind of upset this n squared dynamic of medium of exchange if basically anything can become a medium of exchange?

Tushar: Let me reframe that for you a little bit and ask you to look at it a little bit differently because that’s a very valid point. But let’s say right now, we could wave a magic wand and solve the scalability issues with Bitcoin, right? And if we were able to do that, do you think that this world that you’re describing where you have many different tokens serving as medium of exchange would exist? Probably not. Why would you go and turn your Bitcoin into Litecoin or Monero or Zcash or whatever to spend when Bitcoin itself could go and handle those transactions. This idea or this vision of the world where you have a separate store of value from a separate unit or medium of exchange is really just a symptom of a backwards looking technological view, where if you look backwards at the technology and look at like, oh, we haven’t been able to solve scalability on blockchains yet. Then yes, you do need a separate store of value that has more security than the medium of exchange. That makes sense.

But there is nothing about scaling blockchains that is inherently unsolvable. This is a technical problem that I can say with extreme confidence will be solved. I can say with a little bit less confidence, but still strong confidence, it will be solved within the next five to ten years. So we’re not talking about generations of time. We’re talking five to ten years where you will have a blockchain that scales to support all of the transactions that are needed. And so in that case, why would anyone have that extra layer of friction of, well, I pay you in coin X and I paid that person in coin Y, but I really store all my assets and coin A or coin B and that’s just a lower friction version of the world that it seems more likely to happen.

Kyle: To add onto that, right? So once you accept that, you know, we will have scalability and the store of value and the medium of exchange can technically be one thing. I know we’re not there today, but if you take it for granted that will become the case, which I think it will. Then the next question from there is a balance sheet risk. So right, the problem is, no one wants to have their wealth started a thing that’s variable relative to all of the other things they commonly purchase. People in crypto complain that US dollars are terrible store of value and on a long enough time horizon, and if you’re like an Austrian economist or you know, these guys keep printing money and inflating my dollars. I understand that view. It’s technically correct. But on a practical basis, it turns out people like US dollars, that they worked pretty well, especially if you live in Argentina or Venezuela or something. But even most people in United States seem to be pretty happy with US dollars. So you know, although you can have call it two percent annual inflation, that’s not that bad. So people, you know, like having everything denominated in one currency means you don’t have balance sheet risk of oh my god, my X tomorrow worth two times the price they were today, or my gas is worth 4x the price. There’s real, when I say the network effect is n squared, it’s because it’s not just that they literally accepted the money on the other side and that they could then, it’s also that they denominate their wealth in that thing and whenever they denominate their wealth in that thing, that’s what makes the network effects are powerful.

Brian: So this is interesting because Tushar, you basically made the point, if I got this correctly, right? So we were going to have blockchains at scale, you know, before out almost, for almost no limit or too extreme amounts and I certainly agree. And then if you have a medium of exchange there, you know, that will be kind of universally accepted and that kind of store values almost a byproduct. And then Kyle, he made the point of the importance of stability for a store of value. So would you say that if somebody creates a blockchain that scales and that somehow also has a stablecoin, or maybe is some sort of stablecoin that that would then be likely the winner?

Tushar: Possibly. So I believe that money is an emergent phenomenon. And what that means is things become money by being used as money. Another way to look at it is don’t look at something being money as a noun, look at money as an adjective. So some things are more money than other things, like the US dollar is more money than almost everything else. And you know, anything that you’re using for barter in some sort of barter economy is less money than something like the US dollar, which hopefully will be replaced by a crypto that is even more money than the US dollar because it’s global, et cetera, right? So when you start to look at it this way, you’ll see that once something starts being used for transactions, because it has good attributes, it has the attributes of good money, then it actually becomes more widely used. And like Kyle mentioned, the network effect of something that serves all three purposes of money is n squared. And so that will actually quickly start to grow in value and attract more people to it. That’s just how that ended up playing out. So I would say that if we saw a blockchain that scaled and had the principles of sound money, a stablecoin, might be that perhaps depending on an individual’s view on economics, they might not think that a stablecoin is actually sound money. They might want it to actually be deflationary, which a lot of people in the crypto world seem to like. But I would expect that a blockchain that has the right amount of security and the right amount of scalability and also has the decentralization to make it trustless really comes out and is also able to be used as money that we will see that actually end up winning.

Brian: So how is this reflected in your Multicoin’s trading kind of strategy and thesis and positions today? I mean, it sounds like you guys aren’t very bullish on Bitcoin, so you kind of looking out for what are the new protocols, new technologies coming up that will most likely take on that kind of position?

Kyle: Yeah, so today I spend a substantial majority of my time kind of thinking about that issue. To tie this back to your stablecoins’ point, there’s kind of three, I think, fundamental hypotheses for what becomes the mega winner or mega winners in crypto, the things that are measured, not even in the single digit trillions but in the double digits trillions. And those could be that one hypothesis is that the store of value, so just like Bitcoin, Bitcoin Cash, Moneros, Zcash. Those are the four major contenders on that front. The other major hypothesis is the smart contract platforms become the stores of value because they’re the most useful because you need the native token of the platform to power all the transactions. There’s going to be all these useful dApps on top of them and so that the utility creates value and basically it transforms from a mender good to just full-blown store value. That’s second hypothesis. And then the third hypothesis is basically stablecoin is the killer app. It turns out that the big problem for crypto all along has been price stability. And that if you could just get price stability, and decentralization. I mean, once you have a stablecoin, then you could easily layer a smart contract platform on top of it. That wouldn’t be too difficult to do at some point down the line. So you could see that path potentially happening as well.

Right now, if you told me to pick one of those paths, and bet about my fortune on it, I would say pick number two, which is just smart contract platforms in general. And so I spend a lot of my energy today evaluating those, working with those teams, understanding pros, cons, tradeoffs, and then understanding how their ecosystems are developing, what’s being built on them, what are the challenges of building on them, and understanding how value will accrue in those ecosystems. But to be clear, we do have bets on the traditional store value bets. We have some Bitcoin, although it’s a very small allocation and stablecoins. We’ve talked about them publicly a lot. It’s a space we’re super interested in

Tushar: The other way to look at it as we’re looking at the portfolio probabilistically and trying to see what do we think has the best chance of winning. We do currently hold the view that the token that is most likely to become money or you know, digital peer to peer cash effectively is going to be the native token of a smart contract platform that becomes dominant. We are in the very early stages right now where this is still up in the air. However, I do expect that once we have a dominant smart contract platform that there will be a lot of social inertia behind that token being money and social inertia will carry the day. So we will see how that plays out. And I don’t know that stability relative to a Fiat currency is going to be a killer app for that platform. I think that’s a helpful transition. Maybe. But stability relative to a Fiat currency, just use the Fiat currency. It’s the whole point of using cryptographically bound protocols to define money is to not allow politicians to go and control your money. So if you want to peg your stablecoin to the US dollar, then use the US dollar.

Meher: So you mentioned, we can think of this as a tree, right? So there’s tree branches, so one branch is things that are trying to be money or stores of value. Bitcoin, Bitcoin Cash. Second is, second path could be smart contract, native tokens like Ether, Neo. One of these could become store of value. And the third branch is perhaps people do succeed in building a block one resistance like the stablecoin. Zooming in on the second path, what do you think of Ether’s chances to become a store of value?

Kyle: If you told me to pick one crypto today and come back in ten years, I will take Ether. My sense is the probability that Ether is kind of the big winner as call it 10 to 20 percent, somewhere in that range. So I think a very substantial minority percentage, but they are clearly the leader. It is theirs to lose. There are lots of ways they can lose. It’s theirs to lose. So I assign that probability somewhere on the order of 10 to 20 percent and then everything else I assign basically zero or near zero. It’s just the unknown, you know. Saying today, oh, well, definitely it’s the hot new thing and it’s going to change the world. That’s lunacy to say that. I personally think Bitcoin has gone the wrong way. Bitcoin Cash has a chance. Maybe I would assign maybe a substantial probability to Bitcoin Cash. The other smart contract platforms really haven’t launched in substantial capacity. So it’s again, assigning probabilities that are substantially larger than zero is not really justifiable.

Tushar: Yeah. And one of the things that we do well at Multicoin is we disagree with each other sometimes. So I am more bullish on Bitcoin than Kyle is, I am more bullish on the Lightning Network, et cetera. And it’s important within a firm to have different points of view. So that way, you don’t just create your own echo chamber. I do want to add to that that there is a substantial chance in my mind that Bitcoin could win just because it is the shelling point, it was the first blockchain that actually worked and really has lasted all this time and it’s very stable. And if they solve their scaling problems, they have a fair shot at it. So I wouldn’t put that probability at zero, let’s just say. This is a matter of extensive debate within the firm.

Meher: So I’d like to combine many of your views together into a question. So one of you have uses moneyness is an emergent property. Second view is a store of value is going to be tied to whatever wins the medium of exchange. Now if we combine these two things with Ethereum’s transition to proof-of-stake, my question is going to be whether if you combine these three things, you end up with something that’s not so good, that’s sort of illogical at some level. So the nature of proof-of-stake is if you have a coin like Ether that’s in proof-of-stake and there are validators that are bonding this coin in order to validate, then it’s likely that a majority of the coin supply will end up bonded and stationary. And the greater the fraction of coins that are bonded and stationary, the more secure is the proof-of-stake system, right? So the designers of Ethereum would want say 60 percent of future Ether to be locked down. If the coin is locked down, it doesn’t behave well as a medium of exchange because token velocity then is very low. The rate at which the money flows through the economy becomes very low. Whereas for medium of exchange, you want the token velocity to be higher. And then if a proof-of-stake system isn’t going to succeed at medium of exchange, since moneyness is an emergent property that emerges out of medium of exchange, a proof-of-stake coin ought not to win the moneyness or store of value market. How do you reconcile all of these three things?

Kyle: So let me just ask you to talk a little bit deeper into that scenario that you constructed. Why would the fact that 60 percent of Ether are staked in this scenario that you created actually affect the velocity of the remaining 40 percent of Ether? If you just see the velocity of that remaining 40 percent go back and increase significantly, so that the velocity of the total system stays at, let’s say 5 or somewhere between 5 and 10, which is the range of velocity for the US dollar for the past 20 years or so. Then it seems like that would solve that problem.

I find most of the economic counter arguments against proof-of-stake to be very theoretical in nature and not as useful to think practically. If you think about a network here, right? We’ve got an open trustless network and anyone can be validated or kind of a thing. Someone has to be the validator and they need to be compensated for doing that. There are fundamentally two ways to compensate them for doing this. Either you have transaction fees or you just inflate the money supply and pay the people you know, pay the validators with inflation. Those are the only two choices. IOTA has their idea for like you do your proof-of-work thing that like [laughs] I’m super skeptical of. You know, really, the only two conceptually viable models we have today are inflate or transaction fees. Bitcoin today is inflating and there’s some transaction fee bonus on the side. Ether is doing the same thing. I see no reason why that harmony can’t just continue to coexist forever. And even if you get scalability high enough, there’s even a world where transaction fees to go to zero, and you just pay with inflation. The EOS, that is the system by design. It just takes that to its logical conclusion. And I don’t see anything fundamentally wrong with that.

Tushar: To add to that real quick, actually, I saw Brian, you tweeted something about transaction fees recently on Twitter and I thought that was really interesting and I added a little bit to it, but I strongly believe that an inflation-based model wins because the fees that are invisible to the user are much more palatable to the user. And because we believe that money is an emergent phenomenon, I believe that the protocol that enables a predictable amount of inflation, I’m not saying crazy inflation, I’m not saying go and put the central bank in charge of it or anything like that, but a predictable amount of inflation in order to pay the validators for the security of the network actually ends up winning because now people are less hesitant to actually use that asset as money versus right now every time, like if I wanted to use Bitcoin on BitPay to buy something, even with like $4 fees, that means I’m never going to go and buy a t-shirt or, you know, pay for lunch with it, et cetera. So we really need to see zero transaction fees. I think that is the logical conclusion. Once we see zero transaction fees, you will see that emergent phenomena effect of money kicking, and the only way to pay for that and still have a secure network is to have inflation.

Brian: Yeah. Maybe let me briefly expand, kind of refer to what you mentioned. And I totally agree. Especially with the store of value use case, the inflation makes so much sense, right? Because let’s say you have Bitcoin and then people say, oh, it’s become a store of value. You don’t have to use it, you hodle it and then she just sits there. But that really means that you have people who may have, let’s say somebody has a million dollars in Bitcoin, they never use the coins. It sits in their wallets for a year. They have paid absolutely nothing towards, I mean, in a world where you don’t have inflationary or block reward anymore, they’ve paid absolutely nothing towards securing your network. So you have really unstable situation where on the one hand, you know, transactions have become so expensive to pay for all the people not using it. And the people not using it free ride on the rest, so it seems like a very ugly system. And I agree in general, in the long run, what makes most sense to me is that you have networks where you basically pay for the whole security through inflation and transaction fees are really just a anti-spam prevention, right? So you’re just trying to create people from creating crazy tax on the network by sending pointless transactions. But otherwise, there should be, you know, near zero.

Tushar: Yes, the anti-spam mechanic is actually really valuable. And I am confident that there will be an anti-spam mechanic in any future blockchain that ends up winning this use case. There are a number of other ways to do that as well where perhaps you have significant capacity enough to absorb standard daily global transactions, but if you start to get beyond a certain threshold, you start to implement certain either fees or prioritization for transactions. So you can also prevent spam attacks by identifying them and deprioritizing them rather than necessarily charging a fee. Like for example, when you look at your email inbox, you probably have very little spam in there actually, at least those spam emails that like everyone was so familiar with 15 years ago. And the reason for that is not because they charge people to send emails now, it’s because Google, primarily I would say Google because that’s my email experience. I’m sure the other providers have gotten good at this too. But Google has gotten phenomenally good at identifying what is spam and what is not, and they deprioritize it. They go and put it in that spam inbox. So there’s no reason to suspect that if we were able to do that with email, that we will not be able to do that with transactions on the blockchain.

Kyle: Yeah, I mean, this is the whole thing of like AI and you might call it censorship and like maybe de-censorship, but miners today already have the right to censor blocks. We see some miners mine empty blocks. Fundamentally, if you have miners mining things, if miners can censor things and if miners come to the conclusion that they could run some AI algorithm that detects spam and then filter those spam transactions and just not do them, then they will. And maybe there are some miners that say, no, I’ll take the fees and do it and that’s their prerogative. They can do that. And I expect over a long enough period of time, every miner is ultimately software and so he or she can like choose to censor these transactions however they wish. And if it turns out that all the people who are producing blocks at some point come to consensus about some algorithm that censor span transaction than like, sorry, spammers. [Laughs]

Meher: Yeah. Although I feel if I experience spammers, I wouldn’t call it a permissionless system anymore. So the other piece that I like to explore in this idea that Ether could be a store of value in the future. In your other blogs, you mentioned this other concept which is called the smart contract network effect fallacy. The idea here is that evolution of blockchain technology is proceeding in a way that the notion of one particular blockchain or the other blockchain mattering to the user is going away. All the user’s going to see is token and that token can move across different chains. Chain can also anchored their security to other chains, which means they might not be any advantage left on the infrastructure level itself for a particular token. On the other side, a lot of different programming languages for smart contracts are emerging, lot of ways of making smart contracts are emerging. So perhaps smart contract platforms are going to be like programming languages. There’s gonna be many winners and many different use cases in these many winners and there’s might not be one dominant smart contract platform. If there might not be one dominant smart contract platform, in that world, Ether does not have any particular advantage. And if Ether does not have any particular advantage, why audit to have a special position in the store of value market?

Kyle: Yeah, So the general case for that, what I said earlier, if I were going to pick one crypto and come back in ten years and I pick Ether. My reason is that I think Ether has the highest probability of Ethereum being the most useful chain and that ultimately value will accrue. People will assign value because they want their value in that chain of that token. The kind of thesis of the smart contract network effect fallacy is all of the technical things that make a network effect decays in traditional technologies, at by and large, are not relevant. The fact that everything is open source, the fact that you and I can have synthetic tokens and move them across chains, and the fact that you can anchor chains into one another for security purposes. All of these kinds of things do commoditize the chains themselves, but you can still never commoditize the notion of balance sheet risk and like people wanting to accept it and store it as their fundamental store value. So that is a real network effect. And that’s intrinsic to moneyness and that’s not in any way tied to technology specifically.

When I think about all the smart contracts, so I think that that’s a very theoretical and abstract view of the space. And I think on a long enough time scale, the kind of hypothesis presented in the smart contract network effect fallacy will play out. I think we’re at least five years away from that happening in any substance. What’s very clearly happening today is we have a Cambrian explosion of ideas of how to scale smart contract platform, what’s the kinds of features to offer developers, all that stuff. There’s a massive Cambrian explosion. You’ve got Cosmos and those guys saying, hey, there’s going to be 50 bazillion chains and every chain is going to be sovereign and they can talk. You’ve got EOS saying, guys, this is all too hard to scale. We over decentralized, let’s recentralize a little bit. And that solving engineering problems. You’ve got Ethereum, you know, and Polkadot and DFINITY basically saying, hey, we can shard these things. Look at our new novel consensus models. We can get to fast consensus and shard things and communicate across chain and have pool security and kind of have the best of all the worlds.

Right now, there’s a lot of ideas being thrown around. These ideas are at some level, they’re all competitive and another way is, they’re adjacent. In some ways, they’re kind of fluid. You could totally see a world where EOS is massively successful for some certain class of application and EOS plugs into Cosmos, that Cosmos is plugged into some Polkadot chain and that Polkadot relay chain has like 50 pair of chains in it. That’s totally possibly where we end up in this really interesting mesh hybrid thing. In fact, that probably, to me, seems probable over the next five years, that we get these really interesting heterogeneous networks. But that on a long enough time scale, the gravity will have—I say, once all the questions are answered, how do you architect these things? How do they work? How does all this stuff kind of come together? And once there’s no, I say, substantial technical risk and like the system as a whole, then you’ll start to see convergence around, well, what’s the one I just want to keep my money in and what’s the one that everyone else just wants to keep their money in? And that would be the network effect of the moneyness.

Brian: Cool. Fantastic. And I think that was a really compelling view of how this ecosystem could evolve. And yeah, I can totally see that. Now I want to jump to a blog post you guys or you wrote, Kyle, that is really interesting. It’s just came out. And you’re differentiating there between different types of tokens. So you’re talking about basically kind of a payment token and you’re talking about a work token and at burn-and-mint token and there are kind of very different dynamics that those tokens have. So can you first of all just briefly define what those are and then how they affect the different models, kind of affect how the values or how they would be valued?

Kyle: Yeah. And so, just for some context on this, so obviously, I talked with a lot of entrepreneurs in the space and that everyone’s coming to me saying, oh look, I have a new token. And basically my token is used to pay for my service and all of these tokens have kind of this problem that I call it the velocity problem, whichever. I think I wrote about it in November or December of 2017. The last problem basically says if you just have a proprietary payment currency then the money supply will turn over so fast, at some point, it will turn over so fast, that value itself will not accrue to the token. So that’s kind of the hypothesis. I call it the velocity problem. So lots more people agree with me. A couple people disagree with me. I have pretty high conviction that that will be a problem that will stifle value capture for basically all payment tokens. So I got really frustrated because I just kept getting pitched interesting things, they sounded cool, but I was like, guys, I can’t make money. I’m an investor. I’m a fiduciary to my investors. And I need to see how this thing will actually capture value and I would for basically everything I looked at for the number of months that I was just like, they all have the same problem. And a handful of teams I spoke to had like presented interesting ideas to solve the problem, and I kind of thought more about it and eventually kind of came up with some more coherent thesis around how entrepreneurs in general can try and solve the velocity problem.

And that’s what’s reflected in the post I published a few days ago. That post is called New Models for Utility Tokens. So one model is, right now with all these utility tokens are, and specifically, I’m referring to utility tokens. So I’m not referring to Monero or Zcash or Bitcoin. Those are general purpose stores value. I’m also not referring to the native tokens of smart contract platforms. I’m referring to things like Civic, 0x, Basic Attention Token, Raiden, this goal on these kinds of things, that are like functional utility tokens. So if you look at these tokens, they’re all payment tokens and the velocity just kind of accelerates towards infinity if these things become successful. So one model maintains that those tokens should stay payment tokens but introduces this interesting burn-and-mint equilibrium mechanic so that even as usage grows exponentially, price like the system will capture price, the price of the token will increase in and the system will actually capture value. And that is called the burn-and-mint equilibrium model.

The basic idea of this model is every time I want to pay for a service, let’s say I’m paying a 0x relayer, for example, for a I got facilitating some trade. Instead of paying them the 0x fees that I was, that I would have otherwise paid them, I actually take the 0x fees and I’ve burned them. You literally burned the tokens and then you issue, you know, send the notice to the blockchain and say, hey, I burned this money, here’s proof that I’ve burned this money and I burned this money in the name of that relayer, whoever that relayer might be. So you have that dynamic running completely, so it’s just constantly burning money every which way, and the system is deflationary. Then completely independently of that, you have the system means tokens on some predefined schedule. It could just be a flat number, it could be a schedule that has some curve or some growth curve to it or something, but it just prints a fixed number of tokens per day, per month, per hour, per block, whatever, sometime frequency. And then the idea is that for each time frequency that goes by, each time period that goes by, you as a service provider of that network, if you have tokens burned in your name, let’s say all of the tokens that got burned in the last 24 hours were burned in my name, then I have the right to claim one percent of all the newly minted tokens that are minted at the end of this 24 hour period.

And so what this system does, it sounds a little bit convoluted and it is a little bit inorganic, but what the system does is it solves the velocity problem. Because if let’s say you’re minting 10,000 tokens per month and let’s say people are burning 15,000 per month as a result of using the system, will at some point that’s going to create, supply will be decreasing and that will create upward price pressure. And as the price of the system goes up, then the number of tokens—if the underlying cost of the service is denominated in US dollars, then that means now for one token, you can do more of the same service, you get this fundamental equilibrium will come into play. And so that’s the idea of the burn-and-mint equilibrium model. It’s a little bit hard to kind of  follow verbally, but if you check out the post that walks through it, it provides some examples and stuff. The really important thing about the burn-and-mint model is that, it basically guarantees that are on a long enough time scale and I don’t think it needs to be that long, we’re not talking two years, I mean, even on the order of months, if the system, if demand organic, demand for the service is growing and people are using the thing more, price will go up because you are creating a life, figuring out a way to capture value in the system itself. So that’s number one.

The other model I propose is what we call work tokens. Work tokens are a fundamentally different way to think about tokens. Work tokens are not used as payment in the system. Work tokens, I think of it as a right to perform work for the network, kind of like a taxi medallion kind of a thing. And so the kind of first pioneer of work tokens is probably Augur. A more recent example that I’m also privy too is Keep and they have used these models is Keep it a little bit easier to understand. So Keep as a platform for multiparty computation. Basically, just think of it as people renting out their computers and doing some very interesting technical stuff in their computers. The idea is that you can see one anywhere in the world can rent out a computer that can do some interesting privacy computation stuff. Standard of Keep is instead of paying for the service and keep tokens. If you want to be a node operator in the Keep Network and run those computations on behalf of whoever, you have to buy the Keep tokens. You have to stake the Keep tokens and then as new jobs come in, as new people all over the world say, hey, I need to run a job at Keep Network, then basically, the system looks at the total number of tokens that are staked by node operators. It looks at the number of tokens that you have, and you get assigned that probability of getting assigned the next new job in the ecosystem, in the system. And so new jobs come in, you get a proportional number of those jobs and then you run the job and then you get paid and some other token. It doesn’t really matter what token you’re paid in. In the case of Keep, Keep isn’t ERC20 token and I believe payment will be in Ether since the vast majority of new tokens today are ERC20 tokens. It seems reasonable that most work token models, you know, the person being the service provider will be paid probably in Ether. And then the cool thing about this is, hey, it actually makes Ether more money as we were talking about earlier, but coming back to valuation models, it actually then becomes very easy to value Keep tokens. Keep tokens just become now valued using a pretty traditional net present value model where you can value the Keep token as a function of cash flows being paid to the node operator. And so you can have to just very rational buyers buying Keep tokens and earning yield on their Keep tokens and on the hardware they have been sitting around.

The work model works very well for basically any commodity type service. So you’re selling hard drive space, you’re selling life period, you’re selling transcoding services. You’re selling bandwidth, like any of those types of things that work tokens work fantastically for. If you’re going to host any sort of distributed application like high performant, distributed application or something, that would work great for those. And then like there’s people now playing with work tokens as well for even human jobs. I know Gems is doing this, for example. I’m sure there are others. What the idea is it’s kind of like Upwork where you just say, hey, you know, I want to do a job, pay me five bucks, or I’ll do the job for 10 bucks. But to do that, I will take some of my own money. And then that’s my name on it at risk, right? And so if I don’t do the job as advertised, then I will be penalized. And it’s all of these work token models have that model of like you stake your tokens, you take on some risk. If you fail to perform the job, as you said you would, then you get penalized. So you have accountability. And then if you do the job correctly, you get rewarded and you can value that as a function of cash flows.

The really design goal for both of these models is if you adopt either these models, there should be linear or even practice super linear relationship between organic usage of the system and price appreciation of the token. And I think that’s a very profound thing. You know, we have all of these entrepreneurs running around asking people for millions of dollars, investors are expecting a return. Given that’s the case, you need to design the system, the token to actually capture value and appreciate in price rather than simply just having a philosophy that trends towards infinity even if the underlying services aren’t very successful.

Brian: Cool. Now, I think that’s a really great mental model to think through that. Now the payment tokens don’t play very favorably in this model. Do you think there are use cases where that makes sense? So do you think in general kind of trend towards, you know, either these work tokens or these burn-and-mint tokens?

Tushar: I think something that needs to be taken into account for these payment tokens is that, if all it does is it’s a proprietary payment currency, can it be forked out very easily, right? And in the model that Kyle was describing, especially like the taxi medallion model, the tokens cannot be forked out very easily because you actually need them as a part of the incentive structure of the protocol. You need something to punish bad behavior on the part of suppliers to that protocol. With the burn-and-mint equilibrium, that is more vulnerable to that type of actual forking out. So there is a chance that you see a lot more of these proprietary payment currencies get forked out because that is a more stable, from a game theory perspective, it’s more stable equilibrium where you will never see a proprietary payment currency forked into something, right? So I actually think that the idea of proprietary payment currencies is going to go away almost entirely. And that there will be other valuable tokens. I know you know, matchless sitting here saying that there’s only one token that will have any value. I think that many tokens we’ll have a lot of value, but the idea of a proprietary payment currency is just not sustainable from a game theory perspective on forking.

Meher: So the proprietary payment currencies, what are good examples of projects that have such a system?

Kyle: So the most obvious examples are probably Raiden and Golem that comes to mind, but there are certainly others. I mean, most utility tokens are just payments. They say they are governance tokens, some of them. It’s not incorrect and there may be some governance stuff that happens, but assigning substantial value, there is no good formula at all and the value in and of itself is actually I could, you can question it very substantially, that it’s even should be above zero. So yeah, it’s a very hard problem.

Tushar: It’s really hard to ascribe value to governance functions when working exists, right? Yes, on-chain governance could be valuable, but any conversation about the value of on-chain governance or the value of tokens as playing a role in on-chain governance, it’s remember that off-chain governments will never go away. We always have the opportunity to coordinate off-chain and fork the protocol and so that puts an inherent cap on the governance value of any of these proprietary payment tokens.

Brian: So let’s move to one last topic that you guys brought up, which is the topic of competition between network and what kind of creates this sustainable mode and advantage for a network. Of course, we have this situation now that we have all these blockchain networks being public networks and thus the IP and the source code, you know, being open source, right? You can’t really have this IP patent advantage that maybe traditionally it was often used. So what do you think are the key determinants that makes you say, okay, this network has a sustainable advantage, it will be able to retain its position, or what are some factors that make you think, okay, this will kind of suffer under competition?

Tushar: That’s a really good question and it’s another way to look at, you know, the forking the token away, it was just, well, how do you look at competition in the space and understand how the inherent things about the crypto ecosystem affect competition. I think that one of the key things is that everything is open source. So because everything is open source, you can have what I just talked about in fork tokens away, but also you can have other implications of open source. There is no proprietary IP. Everything is public. A good example of that is Zcash has a technology called zk-SNARKs. It’s a very powerful technology that enables you to do zero knowledge proofs for blockchain. But while that is extremely valuable, and we have a lot of respect for the Zcash team who have some of the best cryptographers in the world for continuing to push the limits on this technology, what we do know is that Vitalik has publicly committed to bringing zk-SNARKs over to Ethereum. So that does defeat the main fundamental value proposition for the Zcash, both as a token and independent blockchain.

You can’t invest in something because it has this proprietary technology because it’s going to get absorbed into everything else. So really, one possible way to actually stay competitive that’s not based on technology is choosing a different point within the tradeoff sphere or the or the plane of tradeoffs. So if you choose a different tradeoff, like EOS is choosing a different tradeoff from Ethereum on scalability, you will see that EOS is choosing to centralize a bit more. And they’re saying that decentralization is only valuable to a point. After that, decentralization for the sake of decentralization is no longer valuable. And this is something that Ethereum could go and copy, but they won’t. Ethereum is very much committed to that specific point on the tradeoff space that they have. They do believe that decentralization is more valuable than Dan Larimer and the EOS team believed. And so that’s something that is not going to be taken by a competitor, and that is one really valuable way to differentiate yourself.

Meher: That’s so super interesting, right? So in any market, if you can write down a tradeoff space. So in this particular market, it might be like decentralization, speed of transactions and scalability, and there might be different points of operation on this triangle and different platforms might occupy different points and have a sustainable network and competitive advantage because they are occupying that particular point. So perhaps in the future, in the decentralized exchange market or in any other market, you might be able to sketch these tradeoff points, identify projects that are headed towards a unique tradeoff point and invest in exactly those projects because their tradeoff point is unique.

Tushar: Yes, exactly. And Kyle mentioned earlier that we’re investing pretty heavily in the smart contract platform space because we think that it has the best chance of capturing the value of money or being the digital money. And the way that we’re diversifying our investments in that space is looking at different places on the tradeoff space that different smart contract platforms are occupying, identifying based on our knowledge and our analysis which of those tradeoff points are actually interesting or that we think will have a chance at capturing all the value. But then of course, thinking about the world probabilistically and spreading out to the investments across multiple different places on that tradeoff space where we think that have a chance to capture a significant value. We can’t sit here and tell you that we know exactly what the right answer is. I don’t think anyone can, but we can tell you that it is extremely difficult to go and steer Ethereum very quickly because it’s a high inertia system and so Ethereum is not going to now suddenly adopt this piece from EOS or you’re not going to see EOS adopt concepts from Cosmos or Tendermint about how the security can be actually, just a different security model from pool security, right? So I think that this directly educates our investing philosophy as well.

Brian: We’re kind of towards the end of the episode but we would, I’d love to hear from you guys what do you sort of, how do you see Multicoin evolving over the next two years or maybe longer time horizon?

Kyle: So our immediate priority is, well, let’s take a step back. So kind of look where we came from. So we spent a year and changed kind of the two of us, were close friends, just kind of reading our crypto, learning and diving into and investing in it. Did that for a while. Not very organized. Made decision in May of 17, launch fund. So we’re getting more organized. Fund went live on August 1st. We started to launch the fund with like $3 million. Today, we’re managing about 50. So we’ve grown really quickly. I’ve been very fortunate to bring on some awesome and investors around us. We have 7 employees now and growing to 12 here in the pretty near future. So it’s kind of how we got here. As we look ahead for the company and for the firm are obviously fundamental priority is maximize returns for our LPs and independent the context of our strategy. But within building the firm, we’re really just trying to build like a world class organization and grow up. So we were very scrappy in the early days. We didn’t have an office for a while. We now are at a coworking space. We’re finally getting an office. We’re upgrading law firms, from small law firms and large law firm. We’re doing just getting from kind we got off the ground to how do we build a large, you know, multi hundred-million-dollar hedge fund and add some sort of crisis in the market. That should happen in the pretty near future.

And so our immediate priority right now is growing the firm up from scrappy guys running around who know crypto stuff to full-fledged world class hedge fund. Learning from the guys in equities and credit markets and derivatives markets and traditional hedge fund world, there’s a lot of things they don’t get about crypto. That’s fine. But there’s a lot of things they know about how to run a hedge fund, how to run an asset management firm, how to think about risk and these kinds of things. And so, you know, we’ve taken our forte, which has been great, which has been a really deep understanding of protocols, technologies, working with engineers, these kinds of things, crypto economic models, all that stuff. And now we’re saying, okay, well, if we’re going to do this at our current scale, we need to bring in all this other expertise around us and so we’re really growing the firm right now to, you know, get from seed to an operational hedge fund. That’s our priority and making money

Tushar: Yeah. And if I could use this opportunity to encourage any of your listeners who want to be employed in the crypto space, please check out our careers’ page. We are hiring, we’re looking for world class talent and when love to have fellow Epicenter fans working at the firm because obviously, we’re big fans.

Brian: Cool. Awesome. Well, thanks so much guys. It was a real great pleasure to have you on. Also, great pleasure to read all your amazing blog posts. So hope as you grow the firm, you’ll still have time for that. And of course, we’ll put a lot of links to those in the show notes. If people want to read up on them, which I highly recommend, then they’ll know what to do. So thanks so much for coming on.

Tushar: Awesome. Thank you, Brian. Thank you, Meher. I appreciate it. And thanks for having us.

Kyle: Hey guys, this was awesome. Brian, Meher, this was great. Looking forward to being in touch.