by Y Combinator4/18/2018
If you’d like to listen to more podcasts about crypto, here are episodes with Juan Benet (IPFS) and Dalton Caldwell (YC) and Olaf Carlson-Wee (Polychain Capital) and Aaron Harris (YC).
Craig Cannon [00:00] – Hey, how’s it going? This is Craig Cannon and you’re listening to Y Combinator’s podcast. Today’s episode is with Andy Bromberg and Ramon Recuero. Andy is the co-founder and CEO of CoinList and CoinList provides financial infrastructure for token creators and investors. Ramon is an engineer here at YC. Alright, here we go. Let’s just start, Andy, with a quick intro of who you are and what you’re working on.
Andy Bromberg [00:24] – Sure, my name is Andy Bromberg, I run CoinList. CoinList is a platform where the best digital asset companies manage their token sales and we’re also where investors find high quality deals in the space. We’re working on helping with things like compliance. Helping with things like transaction processing for deals in the space and, then, helping investors who are also our clients find these high quality deals and invest in them. Broadly, we provide full services for sales like Filecoin and Blockstack and PROPS and manage their sale on the platform and, then, our investors see the deals and, then, separately, we just run compliance for a lot of sales as well. Helping them with know your customer, anti-money laundering and investor accreditation, which would could get into but the nitty-gritty technical stuff you got to do if you’re selling–
Craig Cannon [01:05] – Okay, cool.
Andy Bromberg [01:05] – Yeah.
Craig Cannon [01:06] – And Ramon, for people who don’t know you, what are working on?
Ramon Recuero [01:09] – Yeah, I work for YC and before, I have my own company and I work in video games for Zynga for a while.
Craig Cannon [01:15] – Alright, you want to start it off?
Ramon Recuero [01:16] – Yeah, so let’s assume that we’re starting a project, how else would we think of this new mode of fundraising? What kind of projects could be perfect for a token sale versus going through an angel or a fund incubator?
Andy Bromberg [01:30] – Right, so, really interesting. Most projects, not a good fit for this funding model. And we see a lot… CoinList has done three public on the platform, so far. We’ve gotten over 900 inbound. Most projects not really high quality and not doing ICO for the right reasons. When we think about the right reasons, for me, it really comes down to one of a couple use cases. One is, raising money for protocols that need to be distributed. Filecoin’s a great example of that from Protocol Labs which is a YC company. And there, the insight is that, historically, protocols have only been funded by governments, academic coalitions, sometimes, really huge companies because they’re just expensive to develop. When you’re developing something that foundational, it’s too costly for a startup to develop, typically. The ICO model, giving stakeholders a stake in the protocol on that network, is actually the first time that we’ve had a way for new protocols to emerge and be funded for small teams with great ideas. What that’s enabling is this world of competitive protocols where, instead of, one just being the king by default because it’s the only one that’s capitalized well enough, now, we have competitive protocols going back and forth. Protocols is one example. The second is, projects that have a disproportionate advantage by having early users be stakeholders on the network. I’ll break that down a little bit. The idea there is that, if you can give the earliest users of a platform some upside if the project’s successful,
Andy Bromberg [03:04] – that’s, in theory, good for every project, right? If any company gave their early users some equity, if that was easier, then, maybe that’d be good, but we really think ICOs are good for ones where there’s a disproportionate impact there. One example might be if you had some sort of tokenized or distributed tor. Anonymous routing system where if there aren’t enough users on the platform, it just doesn’t work. It actually, it’s just anonymous or pseudonymous, it fails it’s mission. The idea there is if you could give those users an early incentive then, perhaps, that would be incentivized to grow the network faster, get to that critical point where it tips over and becomes useful much faster and then, be able to grow from there. The last category I’ll mention is, we’re early on this one, but… Separate from those two, securities token, so asset backed tokens that are taking an existing asset and tokenizing them, that’s appealing but almost a totally different category from the first one. We could talk more about the difference between those. The last thing I’ll say there is, I imagine we’re going to find a ton more use cases for ICOs. We’re, basically, a year into the market right now, it’s really young, and those first two categories for the more technical ones than the asset backed ones, are the ones that we find appealing so far, but there’s going to be way more beyond that as this market evolves.
Craig Cannon [04:22] – You guys have supported three out of 900 projects that have come through. What are the criteria that you’re looking at, why did you pick those?
Andy Bromberg [04:30] – Yeah, it’s a number of things. First of all, I think we’re looking at the team and the team quality and the technical details. Does the team have a history of shipping product? Did they build something meaningful before? A lot of the same things you’d look for, initially, in a venture capital investment, right. Does the product make sense? Have they put enough work in already, does it seem like it has a real chance of success? All of those components but, then, with the ICO market, what’s different from all the things you’d evaluate in the venture capital market, so, traditionally, team, product and market, the next step is a layer on top of that, because this is different. Tomponents there that I think are important aren’t as crucial in the venture capital world are, one, the actual structure of the sale. And when I say structure, I mean two things: I mean, legal structure and, then, how the sale is actually structured in terms of economics of the sale. In venture, both of those components are important but kind of nailed down, at this point. You don’t really see a lot of weird stuff happening there. But because the ICO market’s so young, figuring out how to structure these legally and then figuring out what’s the pricing mechanism for this sale. How do people get in? How do you decide who gets in, all of that, that becomes important to evaluate in the token industry and again we imagine it’ll become a little bit more standardized over time, but it’s early there.
Andy Bromberg [05:49] – But then, the second piece, which is a little bit unique to tokens, is what we call the token economic model. When you have a token and people are investing in the token, you need the value of the token to have a chance of going up right, otherwise, why would you support this investment?
Craig Cannon [06:06] – Right.
Ramon Recuero [06:06] – Yep.
Andy Bromberg [06:07] – There’s this question of, under what conditions will value actually accrue to the token? How can you make so that the token’s actually worth something? And there’s a lot of theories out there, another theory we could dive really deep into in how to value these tokens but we look at some of these systems for our tokens and maybe they have a really great team, maybe they have a market that’s really interesting, maybe they’ve built some real technology but, then we get to the token and either it doesn’t feel like value is going to accrue or it feels like the model’s broken somehow and, either, good actors are incentivized enough to do what they have to do on the network or bad actors are overly incentivized to bad things to the network. Digging in on that model is such a new concept that I think that’ll be a really big piece of, more and more people are become experts on that.
Craig Cannon [06:53] – Yeah.
Andy Bromberg [06:53] – We’re so early. The best experts I’ve seen are, either, traditional economists who’ve looked at game theory, economic incentives, that’s really what this is, at it’s core or actually from the gaming industry. Because what are games but a system of incentives in getting people to do certain actions? Thinking about game economies, and how those are designed, we’re going to see people move over from that industry into this token economic evaluation.
Ramon Recuero [07:21] – Cool, let’s go back a little bit to something you said before. You mentioned that, and I agree, that this is great for projects like Wikipedia or Linux that are foundational and traditionally have been underfunded. Wikipedia is always asking for donations because they don’t have enough money to continue. If I’m starting a project like that, how do you take on the recent statement by the SEC chairman that most tokens are securities instead of… He hasnt seen any utility token, per se?
Andy Bromberg [07:50] – Yeah, so I’ll tweak that a little bit. I should caveat that, this is not legal advice and everyone should get good counsel and talk to your lawyers but… My reading of chairman Jay Clayton, the SEC Commissioner and Chairman, his remarks over the past couple of months has been that he hasn’t seen an ICO, an initial offering of a coin, that wasn’t a sale of securities. There’s a nuance there. He’s not saying that the tokens will always be securities. He’s saying that the initial offering is securities and without getting too in the weeds on the securities law here… There’s this thing called the Howey Test which is a test in the United States whether something’s an investment contract or security and I think what he’s saying is, at the time of the offering, these things are going to be securities because one of the components of the Howey Test is, is this project dependent on the efforts of a single company, a single entity? And if it is, it’s probably a security. There’s other factors you have to consider but it’s probably a security. At the time of the offering, when the offering’s happening, almost certainly, it’s going to be dependent on the efforts of a single company and so it’s probably a security at that point but our belief and what’s being sorted out right now publicly and trying to be sorted out with the SEC and with these projects is,
Andy Bromberg [09:08] – is it possible for something to convert from being a security at one point to a non security at some point and, if so, where’s that line and how do you determine that? What we see is this idea that, yeah, it looks like these offerings are going to be offerings of securities and that’s fine. We think a lot about how to sell securities to certain sets of investors, there are other ways to sell securities to different sets of investors and we may just need to deal with the pain of that but it’s in the interest to protect the investors and in the interest of protecting the innovation that’s happening in the space. At some point, we will figure out where that line is and how these things can be non securities at some point and be used way more freely and traded more freely among all users and investors.
Craig Cannon [09:47] – And where’s the current thought on that conversion if ever it’s going to happen?
Andy Bromberg [09:51] – Yeah, there’s not a really clear guidance. The Howey Test is a facts and circumstances based analysis, that’s the terms you’ll always hear from lawyers, from the SEC is facts and circumstances. There’s these facts about what the thing is and how it works and there’s circumstances of current state. There’s no, at least right now, there’s no bright line. But, when it fails the prongs of the Howey Test, is when it, if you pass the Howey Test, you’re a security. If you fail, you’re a non security, so…. Counterintuitively, we want people to fail this test. When it fails the Howey Test, at some point, when it fails these prongs and that’s when it’ll become a non security but, the way to get there isn’t, you know, someone sitting in a room and saying, “Eh, I’m feeling like this isn’t a security now.” You go to great counsel who has deep securities law experience and you say, “How about now?” They do a legal analysis on it and they come to you and they say, here’s our memo, we think this thing is, now, not a security, it may have been before, it’s now, not a security and then, you can try and operate with that assumption.
Craig Cannon [10:53] – Okay.
Ramon Recuero [10:54] – In this line of reasoning, the SAFT that CoinList has used for Blockstack and Filecoin, it could be a security at the beginning and when the token is released, it will transition into a utility?
Andy Bromberg [11:04] – The idea is that SAFT, the Simple Agreement for Future Tokens, which was developed by Protocol Labs and Cooley is a security in the same way the YC Safe is a security. It’s a security just like a promise of something in the future as issued by a single company. The idea is that, at some point, when the network’s live and, perhaps, the thing’s not considered a security anymore by a legal counsel, at that point, SAFT holders will redeem for tokens and those tokens will be non-securities. The SAFT will always have been a security but the tokens that get outputted at some point are non securities, that’s the theory behind what the SAFT is and how it functions.
Craig Cannon [11:44] – Okay, and to go back a little bit on the point about sales, what is the test you guys run, like, “Okay, this seems like a legit sale,” and maybe give an example, “Oh, I don’t know about this other scenario.”
Andy Bromberg [11:57] – Right. For us, it’s actually formal legal diligence process as as result of how we’re structured. To make sure we’re in compliance, we have, it’s called, the Registered Investment Advisory Entity and CoinList doesn’t invest in these projects, but we make an investment decision and recommend the investment to investors that are clients of ours.
Craig Cannon [12:18] – Right.
Andy Bromberg [12:19] – There’s actually formal process we have internally that involved a technical advisory committee, so, some of the top developers in the space. A market advisory committee, so some of the top investors and luminaries in the space giving us feedback on the project. And then, on top of that, deep legal diligence. Doing things like bad actor checks on the projects themselves. Doing things like understanding how they operate, how the company operates, how it’s built out and then really writing an investment memo and digging in on the project itself on some of the facets we mentioned before and coming out with a recommendation. It goes to our internal investment committee, they come out and they say, hey, we’re good to go, we’re not good to go and, at that point, we can promote the sale, show the sale to the investors that are clients of ours on the platform. Separate from that, we have a whole separate compliance service that we’ll provide to people that, even if they don’t pass that or even if they don’t want to go through that process, we can still help them with this compliance piece. Even though we’ve only run three of these publicly, shown them on the platform, there are 20-ish, live right now that are using our compliance solution.
Craig Cannon [13:21] – Oh, okay. And then, do they move into evaluation area? Are you providing guidance there as well?
Andy Bromberg [13:28] – We, informally, help these companies figure out their sales structure. If we’re going to work with them, we want to be helpful to our partners but ultimately it’s not our decision, and if they, that’s part of the investment process, the investment advisory diligence process is understand how their value, we can’t make an investment recommendation without know what the price is. That’s part of the process, we’ll go back and forth with them on that and help them sort that out based on best practices and how the market’s evolving.
Ramon Recuero [13:54] – Yeah, and as for the next step there, once the due diligence is done and you say, “Okay, this is a great project.” What are the options that a great project will have because I know there are different regulations like Reg D, Reg A, which ones?
Andy Bromberg [14:04] – Right. This is under the assumption that you’re considered to be selling a security which, again, we believe all of these are or all of them have been, thus far. When you sell a security to investors, you have two options, and the second option has a million sub-options. One is, register the security with the SEC. That’s like when someone files publicly, right. They say, I’m registering this with the SEC and we’re selling a security. The second is to use an exemption. The SEC hasn’t accepted any securities registrations so far. We don’t expect them to, that’s the path any of these sales have taken but that’s fine. A legitimate second option is exempting. And just to give a sense there, when we do venture capital investments, we often exempt them. This is well understood. You’re allowed to exempt these things instead of registering them. There are a bunch of exemptions for selling securities and it’s actually probably instructive to run through a few of the ones that are being used a lot right now and we see as interesting. The one that CoinList deals with the most is what’s called Reg D 506(c) and this is selling only to accredited investors so people with specific criteria isn’t worth getting into but significant net worth or income, who are consider sophisticated investors, and you’re allowed to sell to them with certain restrictions and limits but that’s what we do for the most part. The Filecoin and Blockstack and PROPS sales that were helped by CoinList were all Reg D 506(c) offerings
Andy Bromberg [15:34] – and part of that is, you need to check these people’s accreditation status to get evidence of net worth or income. That’s something we help with, so that’s one. Another is, what’s called, Reg D 506(b), the one before that and that’s for very small private sales where you have an existing relationship that ensures an existing relationship with the investors. At that point, you have have the investors attest that they’re accredited as opposed to providing all this evidence. It’s a little bit easier but it’s way more restrictive in terms of how many investors you can have in and, of course, you can’t do it publicly. Part of 506(c) is general solicitation saying publicly, this sale’s happening. 506(c), you can’t do that. It’s only private, people you have a relationship with. Outside of that, there’s a bunch more. So there’s Reg CF, crowdfunding regulation which came in after the jobs act in 2012. And that is actually allowing you to sell to unaccredited investors. But there’s, similarly, a bunch of restrictions on that. You can only raise 1.07 million dollars with Reg CF right now, that may change soon.
Ramon Recuero [16:34] – That’s similar to Republic?
Andy Bromberg [16:37] – Exactly, I was just going to say, so Republic is a sister company of ours and they do Reg CF offerings. They sell to unaccredited investors, they have, actually, a whole Republic crypto division now that is doing just the crypto deals. For example, with the PROPS sale, sometimes you can get the benefits of more than one of these. For PROPS, we sold a bunch of PROPS to investors using CoinList and then, separately, Republic sold a million dollars worth to unaccredited investors. You got both those user sets. On top of that, just couple others that are worth mentioning Reg S is another exemption you can use to sell to international investors. So no U.S. investors, there’s geographic restrictions on how it’s transferred and all of that. But you, then, have to comply with those countries’ securities laws as opposed to the U.S. You don’t need to check them to U.S. accreditation standards, you can check them against that country where the investor’s from.
Craig Cannon [17:28] – Well, let’s go a little bit deeper on that because I know that with the crypto stuff, the international things are like, “Are you in Switzerland or Gibraltar or wherever?” Where are things going in the future? Right now, it seems completely fragmented, everyone’s operating at a different pace with different regulations, where do you see things going?
Andy Bromberg [17:47] – We’re really going to have different projects do different things and one of the beautiful things about crypto is that it’s global, it’s distributed, right.
Craig Cannon [17:54] – Yeah.
Andy Bromberg [17:55] – And that’s, both, I think just practically true. A lot of, way more big projects for the stage we’re at are around the world as opposed to being right in the U.S. as compared to other industries, that’s really interesting but it’s also a philosophical thing that crypto is distributed by its nature and so people take that to its fullest extent and distribute themselves. There are advantages and disadvantages to each of these jurisdictions. I’ll probably say this 40 times on this podcast, we are so early in the industry that most of these countries have not really nailed down exactly how these things are going to be treated. Some of them need to clarify their securities laws as a result of this and we’re going to see a lot of things shift. I really believe that the U.S. is going to do the right thing. The regulators have been incredibly thoughtful so far. They’ve taken their time in understanding the space. They’re now starting to make statements but what we love to hear from them and what they’re saying is we’re going to apply existing regulations to this. They’re not trying to write a whole new book of laws on how to govern ICOs, not interesting to them, not interesting to us. What the answer is, securities law, commodities law, currency law, pretty well lit, well understood areas of the law.
Andy Bromberg [19:05] – There’s some edge cases around crypto that we need to figure out exactly how to apply them, so that’s they challenging part, this question of when does something transition to being a utility but we’ll get there. We’ll still see a ton of sales happening in the U.S., and at the end of the day, just practically, if you want to run big token sales, there’s a lot capital in the U.S., there’s a lot of money here.
Craig Cannon [19:25] – Yeah.
Andy Bromberg [19:25] – If you’re selling to U.S. investors, even if you’re not in the U.S., you’ve got to comply with the U.S. securities law. So, at the end of the day, I think a lot of projects will be headquartered in the U.S. or even if they’re not, they’ll sell to U.S. investors, they’ll need to abide by these laws.
Craig Cannon [19:38] – Right.
Andy Bromberg [19:39] – But there are certainly advantages to being in other jurisdictions as well, for all sorts of reasons, from regulatory to even just where the talent is in some cases.
Ramon Recuero [19:47] – Yeah, another point there is that, many networks assume that the only path is to do a token sale. Are there other distribution mechanisms like airdrops, for example? what do you think about that?
Andy Bromberg [19:58] – I love airdrops, I’m obsessed with airdrops.
Craig Cannon [20:00] – Can you define that?
Andy Bromberg [20:00] – Yeah, so an airdrop, well, I’m going to loosely define because people are sorting out how to do it right now.
Craig Cannon [20:05] – Okay.
Andy Bromberg [20:07] – An airdrop is a mechanism to give users tokens without them paying for those tokens, usually around the launch of a network or early on in a network’s lifecycle. Couple points to make here. One, airdrops are not a way to skip securities law. If your token is still a security, which many of these are, you can’t airdrop tokens to everyone, that’s actually considered, again, not legal advice, you should talk to counsel, but considered a securities offering and the case law for that dates back to the dot com boom, interestingly, when companies were giving away shares.
Ramon Recuero [20:39] – Oh, wow.
Andy Bromberg [20:41] – The SEC said, hey, you can’t give away shares, that’s securities offering because the user, the person’s exchanging some information, they’re giving, maybe their email address, their address or some information you’re giving them securities, that’s a securities offering, you’ve got to abide by securities law. You can’t just airdrop, that’s not a way to skip the issues with securities law. But to give some examples of airdrops, because I think it’s a really powerful concept, and taking a step back for a minute. ICOs have, kind of, couple together two ideas, fundraising for an underlying company and distribution of the token to early users. We’re trying to both with ICOs. But it turns out that maybe the investors in a sale aren’t the same people as the users. The idea that maybe we could decouple those a little bit, in some cases, and have a sale to investors and then, an airdrop to users, makes a lot of sense. I break airdrops down into three categories. One is what I would call broad or universal airdrops. The way this works is that, and OmiseGO is an example of this. The way this works is that you take an existing blockchain, so Ethereum, and you just give everyone who has a balance on that blockchain, who has some of that token, your tokens. You can do it in different distributions, you could do… And so everyone gets one token. You could do it so they get it proportional to how much they have. You could use all sorts of functions there. But the idea is just broad distribution.
Andy Bromberg [22:07] – Give everyone who has this, and this is interesting, we’ve never really been able to do this before, give someone a product because we haven’t been able to know how to reach people, right. If I wanted to just give everyone a phone, how do I give people a phone, how do I know where to find them? But because these addresses are public, you can just airdrop it–
Ramon Recuero [22:26] – And this goes back to what you said before about aligning the incentives and then everyone has a skin in the game because if I get this free token, something valuable–
Andy Bromberg [22:31] – Yep.
Ramon Recuero [22:32] – I’m incentivized to grow the network.
Andy Bromberg [22:34] – Absolutely, and then, the two there categories that I break airdrops down into is, one, based on off-chain data and one based on on-chain data. What I mean by that is off-chain would be if I have an existing service with users and I just want to give those users tokens. I know something about these users, not because of some blockchain status but just because I know something about them, they’re my users. Numerai is an example of this, distributed network for data scientists to solve these financial problems and make predictions about the market. They gave their users tokens, they just gave it to them. Or, recently, they actually just announced that they’re going to give all Kaggle users tokens. They’re using off-chain data to give tokens to people. The last category and one that is really interesting and hasn’t been done much, if at all, yet is using on-chain data. What I mean by that is, these blockchains are public, you can see the transactions and you can’t necessarily link them to real world identities, so, gibberish Ethereum address, you don’t know it’s me. But, you can get data out of that. We’ve got this whole ledger of transactions. There is by analyzing that, you can pick out which addresses to give tokens to. An example might be, if I was going to start Augur, a prediction market platform, already exists, but if I was starting Augur, maybe I would go to the Ethereum blockchain and I would look at, I would do transaction analysis and I would figure out
Andy Bromberg [24:05] – which address is which users were transacting frequently with Ethereum gambling sites.
Craig Cannon [24:09] – Hmmmm.
Andy Bromberg [24:13] – Right, so there’s a bunch of gambling sites you can spend Ether, gamble, yeah sure. I could look at that, those addresses are known. I know which addresses belong to the gambling platforms. I could users transacting a ton with those. Those are likely gamblers, I’m building a prediction market. Seems like there’s a nice Venn diagram overlapping those users and then I could just give those users tokens immediately. The idea that you could bootstrap network effects off of existing networks because all their transactions are public–
Craig Cannon [24:43] – Yeah.
Andy Bromberg [24:43] – Is a really cool idea. It hasn’t been explored a lot yet, but I think we’re going to see more on-chain analysis for decided who to give tokens to.
Craig Cannon [24:50] – That is super interesting and going back to getting things rolling. Really getting the ball rolling with your company, you mentioned Big Money before. Some of these companies, a hundred million dollars, two hundred million dollars, three hundred million dollars, someone asked, kind of, the inverted question which is, Fintech Hub asked, is there going to be a liquidity crunch for post ICO companies?
Andy Bromberg [25:13] – Yeah. It’s such a different model, right. The way these things are developing is changing constantly. Most of these companies are trying to raise enough that they never have to raise again.
Craig Cannon [25:27] – Right.
Andy Bromberg [25:28] – Part of the idea there is, and this just requires, I think, a mindset shift in how these businesses are run. These companies, in their ideal form, don’t make money from transactions on the network. A big piece of this is removal of rent-seeking middle men in the process.
Craig Cannon [25:43] – Mm-hm.
Andy Bromberg [25:45] – A lot of these companies’ business model is actually more like an investment company. They own a bunch of the tokens, they do work just as a contributor to the network trying to make the network worth more so that the value of their stake increases, but there’s nothing coming in, it’s just they have a bunch of tokens and the value should increase over time. The idea is that they raise enough to get the network out into the wild and built by them, built by other people, contributing to it, it’s totally open, and increasing the value of their stake. The investors are investing and increasingly are getting locked up for some time period, you want long-term support from your investors, so, maybe they don’t get their tokens for 12 months or 18 months or some vesting schedule, maybe it’s way longer, it’s years and years. And so, the idea of a liquidity crunch is interesting but I struggle to find out how exactly it applies–
Craig Cannon [26:35] – Right, yeah.
Andy Bromberg [26:37] – Because the companies, again, ideally should it work, raise the money, start building it, get it out into the wild and then, just keep building it because they want their value to increase. The investors are holding, for some period of time, way shorter than how long they hold for startups, right.
Craig Cannon [26:55] – Which, I mean, you Dropbox IPO’d today, 11 years later–
Andy Bromberg [26:59] – Exactly.
Ramon Recuero [27:00] – One way to think about it is that for a company, when their goal is to maximize the value for the shareholders, but if you’re a company and they’re starting a network–
Craig Cannon [27:09] – Yeah.
Ramon Recuero [27:10] – Then, your goal, ideally, is to take the company out–
Craig Cannon [27:09] – Yeah.
Ramon Recuero [27:14] – So then, you are just one more participant in the network with, hopefully, a lot of tokens because do you think that a lot of networks are going to suffer from not having enough tokens to support the underlying protocol?
Andy Bromberg [27:27] – As in the company building the network or the network itself?
Ramon Recuero [27:30] – The company building the network that, years on, three or four years, they didn’t design the token economics right.
Andy Bromberg [27:37] – Actually, the inverse is true. I think a lot of companies have too much of these networks. And it’s going to really hurt the incentives for other people to contribute. And what’s going to happen is, these networks are going to get forked.
Craig Cannon [27:50] – Yep.
Andy Bromberg [27:52] – If you’re sitting there and there’s a really cool, amazing technology built, this technology is open source, by nature, right. If I’m sitting there and, say, there’s an amazing bandwidth sharing, tokenized network, and the team’s great, technology’s awesome, it’s like really getting going but the team owns 50% of the network. I sit there and I say, well, that seems wrong. And I’m going to build a great team, we’re going to fork this network, we’re going to do a great public push to make people aware of that and we’re going to reduce the team stake down to 10%, 5%, 1%, who knows where these things end up.
Craig Cannon [28:32] – Will the market push it to 0%?
Andy Bromberg [28:34] – Well, so I don’t, I think the market, this is what I love about the token market as a whole.
Craig Cannon [28:39] – Yeah.
Andy Bromberg [28:40] – It is the purest form of markets, right. It’s all open, people can make decisions. Zero percent, I don’t see ’cause I do think there is value to strong core contributor teams. Having a set of core contributors that are building the protocol and they’re not the only ones, they don’t have special rights or anything like that but they’re core in building the protocol, I think the market will want to reward that and when I think about that and you see, maybe, an amazing team with a 10% stake, a solid team with a 5% stake, a weak team with .2% stake and no team with a 0% stake, I think you end up, starting with the great team but, perhaps, over time, as the network value increases, the incentives change such that maybe a great team is willing to do 5% or a great team is willing to do 2% and you may see a lot of forks on the way down. I certainly think the intuition’s right. I agree with that it drives down over time but I don’t think it drives to zero.
Craig Cannon [29:38] – Okay.
Ramon Recuero [29:39] – It’s really interesting because networks own a lot of the percentage of, these teams own a bigger percentage of the network. Worst of all, if the investor crpyo funds targeting an early releasing of a lot capital there and they’re getting a huge chunk of the network that is supposed to be owned by the all the participants?
Andy Bromberg [29:57] – Yeah, the role of the investors is to get the network started, right. Same as an early stage company and seed investors, is to be helpful, get the relationships they need, help them build a team, do all those components. It’s the exact same situation, if the market decides at some point that too much is owned by certain parties, it may get forked. We’re going to see a lot more governance struggles over these networks as they go live because the thing to remember here is, a lot of this, and again, we’re early in the market, but a lot of this is so theoretical, there’s not a lot of networks that are live and being used right now. There’s no real incentive for governance struggles, for forks.
Craig Cannon [30:34] – Yeah.
Andy Bromberg [30:35] – I mean, what are you forking, there’s not much to do.
Craig Cannon [30:38] – Right.
Andy Bromberg [30:39] – And we’ve seen a few, obviously, the Bitcoin forks, Ethereum forks, but there’s not a lot that are live. As networks go live, which I think is going to happen this year as the ICOs of 2017 start to go live, we’ll see a lot more battles over this and it might be that if investors own a huge stake and the network doesn’t think that’s right, they may try and fork it but the piece to remember there too is, you still need to have a team behind it. Focus on building the protocol. If the best team is building it and investors own some percentage, that may be worth it to stick with, the investors owning that percentage if it means that you still get to keep the team there working on making it the best possible protocol.
Craig Cannon [31:14] – Another question from Twitter related to this, Jordan Jackson asks, do you think there’ll basically be one dominant protocol for each of these categories? Hosting, payments, whatever it might be.
Andy Bromberg [31:26] – I believe that, I believe a couple things. One is that, I believe a lot of things.
Craig Cannon [31:31] – Related to this question, specifically.
Andy Bromberg [31:34] – A couple things.
Craig Cannon [31:34] – Yeah.
Andy Bromberg [31:36] – One is that, I do believe that, at any given time, there’s probably one but it goes back to this markets point. It is such a free market here that there will be constant competition, whether it’s from a fork, whether it’s from a new protocol. I don’t think we’re going to get to a place where there’s a dominant one and then, we’re just in stasis and nothing every changes. I do think that there will be constant competition, things will pop up and things will go down. At any given time, there’ll probably be a couple that are competing for that title, but it does feel like there’s going to be consolidation. Interoperability is a huge, huge issue. The other belief I have, and this is one I’m, is strong belief, weakly held, I think we’re early enough that we don’t know how it’s going to play out. I lean towards fewer but more generalized protocols with platforms, networks, capabilities, technologies to allow for interprotocol, interoperability. What I mean by that is, different protocols have different advantages. We could build a generalized one that’s, to oversimplify a little bit, really fast, or one that’s slow but highly programmable, right. Or they might have all sorts of different advantages and the idea that we could then use layers on top of these protocols to link them together, and there’s a number of projects out there, so, Polkadot, Cosmos, that allow for these interprotocol interoperability is interesting to me because there would be a real pain in managing a wallet that had a thousand crypto currencies,
Andy Bromberg [33:12] – one for every single use case I have. I go to CVS and buy something, I pay with one token. I go to this, pay with another, seems very unlikely. The idea that we’ll have a few generalized ones, but I find it hard to believe we get to one, and there’s some people that believe that there will only every be one, right.
Craig Cannon [33:27] – Yeah, yeah.
Andy Bromberg [33:29] – I find it hard to believe that we’ll get to a place where all those components can be satisfied.
Craig Cannon [33:32] – Right.
Andy Bromberg [33:33] – Because it does feel they’re just some real limitations on the trade-offs between those different components. Some are more distributed, some are less, faster, slower, secured, not secured, all these different components, so I think there’ll be a bunch but not a million and then, they’ll be interoperable between them.
Ramon Recuero [33:48] – In this train of thought, did agree with the theory of the fat protocol that they’ve thrown out, agree to that, most of the values going to accrue to the base layer that can be Ethereum and everything that is on top is going to get less and less of that value?
Andy Bromberg [34:01] – I do generally agree with that. With a twist, in that there’s a world where some, closer to the application layer, things do well, but as securities, this is a little bit of a wonky point, but… I’m not the only person whose had this idea, there’s a few people thinking about it, but an interesting concept around, what I’m calling, index tokens right now but it’s not a great name, if anyone has a better one, please tweet at me and let me know. But the idea is that there’s a… We don’t have to have a no middle-man world. If some protocols take some fees for being helpful, or some slightly closer, some higher up protocols take some fees, it may not be the end of the world if it’s the best option. There’s kind of like a knee-jerk reaction of like, no middle man, that’s crypto. Ideally, I think we get to that place but there’s some that could be worth it, so, if we were to be designing a solution that did something for protocols. So let’s say, we built a tokenized network that made protocols more secure or encrypted protocols transactions, something like that, so it’s helping protocols, it’s sitting on top. Perhaps that layer, that token, that network could take fees from those protocols per transaction. Maybe the network chooses to pay so much, it’s helping Ethereum, so if they pay so many ether to that layer on top per transaction, what would be interesting there is that token would then be an index across all the projects that it’s helping, right, Ethereum’s paying with some ether,
Andy Bromberg [35:51] – Bitcoin’s paying with some bitcoin, Filecoin’s paying with filecoin and it all goes into this thing and then, that one has its own token but the token isn’t used on the network, it just represents this pool of assets that are being accrued for the network. That thing, again, not a lawyer, but is probably a security, it’s an investment, you’re saying this is. But value is accruing to that token, possibly, in a very meaningful way, the more projects that get on it, the more it’s worth and so, this token, all of a sudden, has value. Maybe the protocols are accruing a ton of value because they’re being used and the value’s accruing to them but maybe there’s this layer on top that is also gaining value as the protocols do. As Ethereum gets more transactions, it’s paying more fees to this network, still worth it, but that, now, is accruing a ton of value as well. I am a believer in the fat protocols but, I think, we’ll see more layers on top of it as well.
Ramon Recuero [36:41] – Let’s talk a little bit about the end user layer. It’s nine years since the Bitcoin whitepaper was published, but many people say that and fairly we have really finite end consumer cases that are in use right now. What is your answer to that?
Andy Bromberg [37:02] – Nine years since Bitcoin, absolutely, but it hasn’t been that long for other tokens. And I think Bitcoin’s use case, from my perspective, is store of value, it’s a digital gold. It’s doing that, so, there’s an end user use case. But that’s the one for Bitcoin, I think. These new ones are so young.
Craig Cannon [37:22] – Yeah.
Andy Bromberg [37:24] – We had a period of experimentation a few years ago where a few come out, obviously, Ethereum, some others but it was so early, we were figuring out how these tokens economical models work, we still are, but the ICO market and this wave of new tokens is really just 2017. And, really, just mid to late 2017. Without trying to make excuses for the industry, if you’re doing a sale in 2017 and you’re building something meaningful based on the result of that sale, you know, takes nine, 12, 15, 18, 24 months, we’re just now getting to the point where that time period’s being hit. I do think the next, call it, 12 months will be a real trial period for these networks that are just now going live. Post ICO, after some time to build and actually make that meaningful technology. Without meaning to weasel out of the question and say there’s nothing, because I agree, there are not a lot of real end user use cases out there right now, I just think we’re just now getting to the time where that’s going to happen.
Ramon Recuero [38:27] – collectibles, how many do you…
Andy Bromberg [38:30] – I can’t answer any questions about what I own but I’m a big fan and that’s one that is worth mentioning, I think.
Craig Cannon [38:35] – Yeah.
Ramon Recuero [38:36] – Collectibles. Is one of the few.
Andy Bromberg [38:38] – Collectibles are interesting because they’re… collectibles protocol, the technology behind it, the ERC721 which is the component added to Ethereum to allow for collectibles or, at least, the first one, very technically challenging, CryptoKitties’ team built that and that’s really impressive but, then building a collectible on top of it, just a raw collectible with no kind of interaction mechanics is actually not that hard. That’s, technically, challenging but you don’t have to build a whole new protocol. We will see a lot of value in collectibles accrue because those can just be spun out really quickly.
Craig Cannon [39:17] – Yeah.
Andy Bromberg [39:18] – Then, you get into more interesting mechanics around collectibles interacting with other, can you build a game where the collectibles, that’s challenging to build but I do thing with CryptoKitties showed us is you can put something out there that, in the distributed Blockchain world, that doesn’t need to have pure technical value to the end users, you can just build a really great product and people want to use it.
Craig Cannon [39:39] – I say this all the time, most people don’t care about how websites work. They don’t care, it’s not an important part. It’s important in how in functions, it’s important in being secure and all that stuff but, at the end of the day, how many people know how Facebook functions?
Andy Bromberg [39:52] – To that point, we are just now entering a time of… What I call, the professionalization of the space but more people with expertise building businesses, more people with expertise shipping products coming to the space.
Craig Cannon [40:04] – Yep.
Andy Bromberg [40:05] – So far, it’s been so beautifully driven by pure technology. Let’s build these deep protocols, the infrastructure layer that needs to exist, that has to continue forever, and it will continue forever, but we’re just not getting to a place where the user bases are big enough, the incentives are strong enough that real product builders, real product people are coming in and thinking about how to build it, and I’d argue that CryptoKitties’s team, Axiom Zen team, that’s a great example. They build products and they said, the infrastructure exists now, we see a use case, let’s build a great product.
Craig Cannon [40:34] – Yeah.
Andy Bromberg [40:35] – And we’re going to see a lot more of that over the coming months.
Craig Cannon [40:37] – Another question from Twitter, Johannes asked, what are your recommendations for how someone ought to think about implementing and on-boarding users to a protocol, or something like CryptoKitties, whose primary users are non-technical? In that case, you’re talking about MetaMask, you’re doing all this stuff, what do you guys think?
Andy Bromberg [40:57] – It’s just a function of needing better products. One really good example of this, it’s an interesting questions because it’s almost not a blockchain question.
Craig Cannon [41:07] – Yeah.
Andy Bromberg [41:09] – In the same way, it feels like asking, how do you on-board someone to a social graph and connect with other nodes? And the answer is, you build Facebook–
Craig Cannon [41:20] – Right.
Andy Bromberg [41:21] – Let me click an “Add Friend” button and that’s me on-boarding to a social graph. And so, the UIs aren’t really there yet. One example that I thought was really impressive is, for an airdrop, a company called Bloom which is a decentralized credit scoring application protocol, they did an airdrop in partnership with Earn.com to airdrop a bunch of Bloom tokens to users. And they just did a great job building this product. Airdrops are hard, you have to use MetaMask and some other wallet, send an Ethereum address, this, that, the other thing. They just built a beautiful process where you basically just had to click a bunch of times and you said, I want this, and then it redirected and something else popped up and said, this is the button you want to click and here’s what it does, click that. Sure, it was still a long process but that’s just a function of the ecosystem being young and the tools not being developed fully yet but they made it really easy, explanations, click buttons. At the end of the day, the answer to that question which is a good one, how do you on-board users to the protocol if the users are non-technical is make it obvious what is happening and make it easy to click buttons and type things and have that happen. Building UIs and building the technology to support that.
Ramon Recuero [42:30] – It comes back to, what are you using the Blockchain for. For example, in the case of CryptoKitties, you’re just proving a scarcity. You have this unique kitty that nobody else has it. But the user doesn’t think about anything else, MetaMask. Even what Coinbase did that is that you own the private keys, so the user doesn’t need to deal with MetaMask. That can be a really interesting approach on this. Think more projects we’re going to see taking…
Andy Bromberg [42:52] – Totally agree with that.
Craig Cannon [42:53] – Yeah, and actually, on the Coinbase side, do you see index funds becoming the default way of investing?
Andy Bromberg [43:01] – Yeah, it depends on what you mean by default, because there’s a lot of different potential parties here.
Craig Cannon [43:04] – Sure.
Andy Bromberg [43:06] – One note, and I’ll get back to the core question, is a tiny percentage of the money in crypto is professionally managed right now.
Craig Cannon [43:14] – Yeah.
Andy Bromberg [43:15] – Depending on who you ask, about 10% of capital in crypto is professionally managed by professional investors. When you look at a developed market like the U.S. equities market or any other major market, that number is 95, 96, 97, 99% professionally managed. For those people, largely, index funds won’t be the answer, although, in some cases, they’ll make significant investments in index funds but the amount of professional investment as a percentage of the total capital is going to increase dramatically and we’ve got a long way to go. Part of that, as an aside, contributes to the volatility. Professional investors tend to, in most cases, help manage volatility for an asset class because they’re professional investors and I think a lot of the volatility comes from the fact that it’s minimally professionally managed right now. For retail investors… I think that we will get to a place where index funds are really significant investments for retail investors, but in these early days, before the market really matures, people like making their crypto investments. They like thinking–
Ramon Recuero [44:20] – It’s fun, it’s gambling.
Andy Bromberg [44:22] – Yeah, it’s gambling, it’s speculation on which ones are going to win but when people are making investments, I do think that they enjoy it at this early stage. We’ll see a significant inflow to the index funds, there’ll be some very successful index funds products that come out. But I’m not convinced that right now it’ll end up being a massive percentage in the same it is in the U.S. equities markets.
Craig Cannon [44:45] – Do you have a default asset allocation that you recommend to people?
Andy Bromberg [44:49] – I don’t make investment recommendations but–
Craig Cannon [44:51] – Oh right, sorry.
Andy Bromberg [44:52] – But I do think that, broadly, as with any asset class, it’s smart to, first of all, we’re very early, in the system for the millionth time, invest money you’re willing to burn. Like investing in seed rounds and startups, if not even more risky, invest money that you are willing to just have disappear. Don’t over index into this. And then, once you do, as with everything, a pretty traditional strategy is, of the money you’re willing to burn, most of it, you may want to put in the stronger, more established ones. It’s all early now, but breaking down kind of into the stronger, more established ones makes sense and then, if you want to, throw some bets into things you’re interested in or you have some insight into and dig in on those spaces. It makes sense for people to invest in projects they understand and so, aside from core investments, in the space or indexing into the space, one you understand is always a good bet.
Craig Cannon [45:45] – Well, ones you understand, that should be exploded out as well, right? If you’re in retail, non technical investor, how do you come to understand some of these whitepapers?
Andy Bromberg [45:57] – I hope we’re moving away from just whitepapers. So a whitepaper is kind of an overloaded term and it’s just become this defacto thing in the crypto industry. The best projects are moving, generally, they may have a whitepaper but they’re moving away.
Craig Cannon [46:12] – Okay.
Andy Bromberg [46:13] – What a whitepaper originally was is a reasonably technical explanation of the system works.
Craig Cannon [46:18] – Yep.
Andy Bromberg [46:21] – Some people want to really understand that. And if we were making an investment here in a big data company, maybe there’s some investors that really have the capability and the ability to understand the deep technical piece of how some set of algorithms works and maybe they do their diligence on that, but plenty of people invest in very technical companies without understanding exactly how the system works and I don’t think that’s wrong, I think that’s okay. As long as someone’s doing their diligence and then, you’re doing the diligence on things that you care about for investments because everyone will have different weightings. People will care more about the team, the market, the product or the technology and that’s totally okay. I do think these protocols should publish whitepapers or something like it, but that should be a deeply technical exploration of exactly what it is they’re building and you shouldn’t have to assume, as an investor, that you need to read and understand that. You should be able to invest in something without understanding what proof of space time is for Filecoin, that’s a–
Andy Bromberg [47:19] – Crazy technical concept and I didn’t make that up, that is actually what it’s called. I think they’ll start to release more, we’ll see investors get more sophisticated about doing their diligence and we’ll see companies get more sophisticated about what they disclose and what materials they put out and there’ll be a happy medium where investors can get the information they need without needing to understand right how the protocol works.
Ramon Recuero [47:42] – That point, I totally agree. I think the startup fundamentals is the same, the team, the market always apply but, then, I hope we see more working networks before they try to token sell. To do a token sale.
Craig Cannon [47:52] – Yeah.
Ramon Recuero [47:53] – So the network is already working and then, you don’t just have a whitepaper, you have a whitepaper and a working network, and especially for products that are more end-user applications. The user, as we talked before, he doesn’t need to know how it works but if it’s something that you care about…
Craig Cannon [48:09] – Yeah.
Ramon Recuero [48:09] – And it’s working already, you may believe in that project.
Craig Cannon [48:12] – Yeah, absolutely. What about the people who want to join these teams? What are your thoughts around vetting a project as an employee?
Andy Bromberg [48:23] – Yeah, I think it goes back to some of the things I said earlier about how CoinList evaluates these projects.
Craig Cannon [48:30] – Yeah.
Andy Bromberg [48:31] – There are a lot of components that are the same as evaluating any other company, right. It’s not like a whole world, it’s–
Craig Cannon [48:36] – The same–
Andy Bromberg [48:37] – Evaluate those things but, then, layer on top of that some of the token specific pieces and, again, there are not a lot of people that are experts in the token specific pieces so far. Structuring these different things and token economics, so, ask smart people. In the same way that, if you’re a non technical person going to join a deep learning startup, ask some technical friends, what are the merits of this, is it interesting, ask some investors in the project, have those conversations and be intentional about finding out things about the team and the product and history and all of that but a lot of that you can do on your own, if you have good intuitions there. Spend more time with external sources, evaluating token economics, maybe the technical details if it’s not something you understand, the history of the team and digging in on that, but at the end of the day, there are people who join companies whose technology and products they don’t fully understand all the time, as necessary.
Craig Cannon [49:34] – Yeah.
Andy Bromberg [49:36] – And I think it’s a lot of the same mechanics there.
Ramon Recuero [49:38] – And how did you get interested in the space?
Andy Bromberg [49:41] – I got interested in the space from a professor of mine university, Balaji Srinivasan, who now runs Earn.com and he was a professor and a number of took this class at Stanford called Startup Engineering. He developed a relationship with him and after the class ended, he told a bunch of us, Bitcoin, it’s a great thing. And, you know, we looked at him, we’re like, this–
Craig Cannon [50:10] – Leave school, buy Bitcoin.
Andy Bromberg [50:11] – Magic internet money. I don’t think this is a real thing. But he convinced a few of us to buy just a couple and then, we ended up, as a result of that, actually starting the Stanford Bitcoin Group, this was back in 2012, 2013. We did some really cool stuff. We did research on some of the economic impacts of Bitcoin, so this was around when Cypress economic collapse happened, so how could crypto have helped that. We built some really cool projects in the space and did a lot of evangelism. Ran up and down Sandhill Road pitching Bitcoin to investors. We didn’t want to raise money, we’re just saying you should be interested in Bitcoin–
Ramon Recuero [50:51] – How did that work, at the time?
Andy Bromberg [50:53] – It was a toss up, there were some people that got it, some that didn’t. We may have needed to hone our pitching skills a little more. It’s interesting not to see some of those people we talked to and how they’re thinking it’s evolved on the space. There were seven of us in that group, most of whom are now still in crypto. We did a lot of cool stuff there and my interest evolved from then.
Craig Cannon [51:15] – I guess my last question is, if you weren’t working at CoinList, what would you be working on?
Andy Bromberg [51:21] – That is a really hard question. Part of what I love about CoinList and what would inform that decision is, supporting building tools for or services for the companies that are building great protocols in this space, right. I don’t have a protocol in my head that needs to be started and tokenized, right now. Built out. There’s some great protocol teams. It is very early and, so what I love is this idea of supporting this ecosystem and bringing value to the people that do have these brilliant ideas and rather than going in and adding skills to a single project, because I like a lot of them. I would much rather say, I can I help all of these projects. It would be something else, kind of, services or tool oriented, picks and shovels for the token industry because I, fundamentally, believe in what’s happening in a meaningful way
Andy Bromberg [52:14] – and want to support it but what I would say is, for other people thinking about the space, especially for people that are not deep in crypto already, these projects need you. They have every role open, they’re figuring out how to run these new types of businesses, whether you’re technical, or a marketer, or sales, or operations, or anything else, these companies need you and going out and finding a team that you identify with, as with any startup, but also a use case that you understand, again, you don’t understand the technology but something you believe in, whether that’s Filecoin.
Craig Cannon [52:48] – Sure.
Andy Bromberg [52:48] – File storage. They are potentially solving a massive problem here and introducing a whole new economic model or any of these other projects, it is worth while and they need help from people that are coming from outside crypto. It’s an exciting time in the industry.
Ramon Recuero [53:03] – Yeah, there’s a lot of energy in the space, it reminds me a little bit of the early time of the startup before the dot com boom, you can see the same kind of–
Andy Bromberg [53:10] – Yeah, all that, energy and also, one of the things that’s so attractive about the space is just the number of A players coming into the space. The brain power concentrated here is… When I say here, I mean in the industry, but it’s worldwide, it’s so high and so much fun that you get to work with such brilliant people excited about making a meaningful change in the way that companies’ projects, technologies are built in the future, that I think it’s just hard to find in any other industry right now.
Ramon Recuero [53:36] – It’s a network affect, the best people are going to attract the brightest people.
Andy Bromberg [53:40] – That’s exactly right.
Craig Cannon [53:40] – Yeah, totally. Alright man, well, thanks for coming in.
Andy Bromberg [53:44] – Thanks for having me.
Ramon Recuero [53:44] – Thank you.
Craig Cannon [53:46] – All right, thanks for listening. As always, you can find the transcript and video at blog.ycombinator.com and if you have a second, it would be awesome to give us a rating and review wherever you find your podcast. See you next time.
ICYMI: Watch 40+ founders pitch at YC's Work at a Startup Expo
December 11, 2020 by Ryan Choi
Y Combinator created a new model for funding early stage startups. Twice a year we invest a small amount of money ($150k) in a large number of startups (recently 200). The startups move to Silicon