Gresham College Lectures

DeFi, Crypto, and NFTs in Business

January 27, 2023 Gresham College
Gresham College Lectures
DeFi, Crypto, and NFTs in Business
Show Notes Transcript

How is the decentralised finance world organised?

This lecture discusses how cryptographic technology is applied in business. It discusses blockchains and their uses. It explains how smart contracts, open code that automatically executes contracts once certain conditions are fulfilled, are used. It will also look at non-fungible tokens, a type of cryptographic asset on a blockchain with a unique identification code and metadata that distinguishes it from any other.


A lecture by Raghavendra Rau recorded on 23 January 2023 at Barnard's Inn Hall, London.

The transcript and downloadable versions of the lecture are available from the Gresham College website: https://www.gresham.ac.uk/watch-now/crypto-nfts

Gresham College has offered free public lectures for over 400 years, thanks to the generosity of our supporters. There are currently over 2,500 lectures free to access. We believe that everyone should have the opportunity to learn from some of the greatest minds. To support Gresham's mission, please consider making a donation: https://gresham.ac.uk/support/

Website:  https://gresham.ac.uk
Twitter:  https://twitter.com/greshamcollege
Facebook: https://facebook.com/greshamcollege
Instagram: https://instagram.com/greshamcollege

Support the show

(whooshing music)- Okay, so to understand how the crypto world operates, let's start with the question of ownership. What does it mean when you say, I own something? I own something, let's say, in the physical world, like a piece of land. Well, in the old days, it was just a might versus right kind of situation, if you own something, someone stronger than you could come in, kill you, take all your land away, or whatever. But then it became a situation of social norms as well, for example, if you own a piece of land here in the UK, if there's a walk through that land which has been used for hundreds of years by people, they have a right of way over the land which you own, so that's also a social norm, that's saying, I own the land, but people have the habit of walking through that land, they can continue doing it. But there's a third form, which is the rule of law, that means if I own something, I basically have my details in a database somewhere, and so that is your title, if you have a house for example, you have lawyers who know where your house is, it's recorded somewhere on a database. So the question I want to address today is who maintains the databases? The first answer, obviously, is a central authority, maybe government. As a classic example of this, think of the "Domesday Book." Right after 1066, when William the Conqueror came in and invaded England, probably one of the first examples of a European national coming into the UK and taking away good British jobs,(audience laughs) he came in, and after he conquered England, he instituted the "Domesday Book," which is a central register of what people owned, so this is your property, these are your houses, this person, in this place, owned this. He did that for the sake of taxes, but it was a central authority. Today, there's also a whole bunch of financial intermediaries who own these databases. If you own shares, for example, it's recorded somewhere in some kind of depository institution, if you have a bank account, the bank has your details. And the problem is, here, the trust in these kinds of databases is a little fragile. Imagine the first one, the central authority. In many countries, it still is a question of might and right, so you own a property, the government says, well, I need that land, they take your land away, they throw you out onto the street. Perhaps not here in the UK, but definitely in some countries around the world. But even if you think about intermediaries, banks are not very good at maintaining databases either. You might think this is an easy job, I mean, you know, you have a bank account, you put 100 pounds into your bank account, how difficult is it for the bank to type in, even you can do this in an Excel spreadsheet, 100 pounds, bank account, and so on? It isn't, apparently, very easy, it's incredibly difficult. To take an example, Wells Fargo, it was recently fined several million pounds for messing up the way it kept its databases. What did it do? One of my favorite examples of Wells Fargo was, during COVID, the government said, you need to offer loan modification programs, for people who cannot pay back their loans. Wells Fargo looked at their records, and they found 190 people whom they said, oh, these are dead people, so we don't need to offer them loan modification programs. Now, this was a record keeping error because these 190 people were not in fact dead, so they continued sending in their checks to Wells Fargo, and Wells Fargo had one database that said these are dead people, they don't need to have a loan modification program, and another system that said these dead people are sending in checks. Maybe ghosts are incredibly financially responsible, who knows, but the point was banks have a whole bunch of legacy systems, written in ancient languages like COBOL, which don't talk to each other, so they're not extremely good at maintaining their own databases, they can mess up. So there's an alternative idea in the crypto world, which is trust a lot of complete strangers to maintain your database. You're kind of like, excuse me, is this more reliable than this? That's the thing that many people are worried about, but that is the essence of crypto. When we talk about this, what we are doing is, we have two transactions, a transaction between myself and one of you, and it could be any kind of transaction, it could be sending money, it could be writing a legal contract between two people, it could be creating a market, but the essence of all of this is there's a bunch of strangers who are economically interested in keeping all the records safe, that's the essence of crypto. Now, why are they economically interested in keeping these records safe? The answer lies in what you call a consensus protocol. What does that mean? Let's suppose I have a transaction, the transaction would involve me, say, paying 100 pounds to Victoria. Somebody has to write that, so basically deduct it from my account, augment her account by 100 pounds, and that's the transaction. So you have a bunch of complete strangers who are now saying, you know what? Let's augment one account, Victoria's account, let's reduce the amount in my account, and everybody has to agree this is the transaction that happened, 100 pounds, not 50, not 110. So how do you do it? Well, you've got a whole bunch of strangers who see that transaction happening, they have to say that's the transaction that actually took place. But how do you get all those 1,000 people together and say, this is what happened, 100 pounds? Maybe 10 people say it was 110, maybe another 20 people say, no, it was actually 90, how do we know it's 100 pounds? The answer is, possibly, majority rule. Majority rule says, well, if 501 out of 1,000 people say it was 100 pounds, we'll accept that 100 pounds was the amount that was transferred. The problem is majority rule by itself doesn't work because of something called a Sybil attack. Sybil, by the way, here, is not the Greek prophetess of old, it's basically a 1970s book written about a woman with multiple personalities. And what happens here is, you have, for example, one person who spins up 1,000 computers, all with their own IP addresses, and pretends to be 1,000 different people all voting for a fake transaction. As long as there's no cost to this, you can put up as many personalities as you want, all saying that this is the right transaction. So you need to make it costly for these people to vote, and how do you do that? Well, the answer is, originally from Bitcoin's perspective, what they call proof of work. That means if you spin up 1,000 computers, the system will automatically increase the level of difficulty of verifying a transaction, which means it's really not worth your while to go through with that whole process of trying to falsify a transaction, it's impossible to do it. And so the essence of crypto is, because nobody trusts anybody else, you can verify any transaction you want, that means you can trust it, because there's no trust involved, so you can trust it. So most people don't actually check the entire blockchain, they could, but they don't, because they could. All right, so as I pointed out last time, crypto's kind of like this generic term, it has a whole bunch of different things, and today, I want to spend some time talking about those different things. So for example, it could be cryptocurrencies, dealing with money, or it could be tokens, sort of dealing with money, it could be decentralized finance, which deals with automatic contracts, it could be blockchains, which deal with databases, all of them fall under this rubric of crypto, fine. It could also be Web3, which is sort of a decentralized evolution of Web 1.0, which was the earliest, chaotic kind of development of the Web starting with Tim Berners-Lee back in the 1990s, everybody had their own website, nobody could find anything, it was completely chaotic, the 1990s, and then Web 2.0, when centralized people like Google, Apple, Facebook, all of them started carving out their own walled gardens, and so they were collecting all your data, hoovering it up, Web3 is sort of, let's decentralize it and go back to it being owned by the people. All right, now, let's apply all this to finance. The first thing you have to realize about finance is that it's bizarre. Why is it bizarre? It's because, if you are a finance person, there are a whole bunch of arcane rules and arcane logic which you cannot understand unless you know financial history. For example, for a long time, you could only, in America, you could only buy shares in 1/8 of a dollar. That means, in other words, you could buy a share for 25 1/8 of a dollar, 25.125, but you could not buy a share for 25.2, you could not buy a share for 25.18, it was decimalized about 30 years ago, but before that, it was only in eighths of a dollar. Why? Because originally, when they started trading shares, you could break a silver dollar into half, that would still be sort of recognizable as a dollar, you could break it into a quarter, you could break into an eighth. Beyond an eighth, it becomes an unrecognizable hunk of metal, you're not really sure whether it is this or not. So unless you understand that, it's bizarre, why are shares traded in eighths of a dollar? It's just something that happens, nobody really knows why. The same thing with crypto, crypto is equally bizarre. In the last 15 years, crypto has built a whole new financial system from scratch. Well, it has basically reinvented and rediscovered things that finance has been doing for centuries. Some cases, it's found new and better ways to do things, some cases, it's done much worse ways, basically heading down paths which were abandoned by traditional finance a long time ago. Sometimes, it's doing the same thing that traditional finance is doing, but under a new name and a new motivation. So these are what you can do with crypto, it's a completely self-contained, elegant financial system, which is kind of interconnected with the traditional financial system. Let's take examples of these, let's start with cryptocurrency. What is a cryptocurrency? We talk about Bitcoin, we talk about Ether, we talk about all these cryptocurrencies, but what is it? Well, one thing people might say is it's a store of value, I buy a bitcoin, I'm storing it, it's like money, but is it? Well, if you think about how this started, when you measure value in finance, we need two things, we need cash flows and we need an estimate of the risk these things take, so cash flow and risk. To value something in finance, you need both. But there are no cash flows in Bitcoin, how do you figure out what the cash flows are? It's a currency in itself, but it's not even a currency. When Satoshi Nakamoto created this, what he essentially did was to create a way of transferring ownership of a number from one person to another. To put this in another way, if, for example, I were to send somebody the number 100, how do I do it? Type 100 into an email and send it over. But does that mean I lose my 100? No, I don't, I can type it in again, it's an Excel spreadsheet, I can type it in as many times as I want. What Satoshi Nakamoto said was, in your ledger, I can take the 100 out of your ledger, transfer it to another person, and you cannot spend that number again. Remember, last time, we talked about the falsification of entries, and we talked about the double-spending problem. Basically, he solved that double spending problem: you transfer ownership of a number, one person gets a number, the other person cannot use that number. And some people said, hmm, that's kind of interesting. Well, you artificially created scarcity for a digital asset, which makes no sense because a digital asset is just, if I send you a song or a book over email, I don't lose control of that; here, I do. So some people said that's like money, so we can associate those numbers with money. But how much money? That's arbitrary, it's an arbitrary number, this is arbitrary, so how much value you associate with it is also arbitrary, there are no cash flows, there's no risk. And so some other people said, wait, this guy has come up with an artificial number, or he's come up with a way of transferring one number to somebody else, and people are paying for it? Let me come up with my own method. So they came up with their own cryptocurrencies, and their own cryptocurrencies, and now we have thousands of cryptocurrencies, which are all methods of saying, I have a foolproof way of transferring a number to somebody else, and maybe people will give me some money for it. As a classic example, Dogecoin. The guys who came up with Dogecoin invented a joke currency, the Dogecoin was basically a Shiba Inu dog meme, and they said, let's come up with a currency which uses that meme, and people said, oh, this is cool, so they gave money for that, so Dogecoin is the sixth or seventh biggest cryptocurrency right now. Okay, but there's another interesting thing about cryptocurrencies, this part here. What is risk in finance? In finance, risk is all about correlation. If the economy is doing badly, all your portfolio, your entire portfolio is going down the tube, everything is crashing, you want at least one asset which is going in the opposite direction, it's going up while everything else is crashing, that acts like insurance, you're reducing your risk by adding that to your portfolio. When the world is going to hell in a hand basket, this thing is doing well, gold, for example, people think of gold as an example. But arbitrary number, arbitrary value, so a lot of people say, you know what? Bitcoin is a great store of value because it's uncorrelated to all the other assets we have. The world is going to hell, this one is just some arbitrary numbers, so it reduces your risk to have bitcoin in your portfolio. Unfortunately, that argument worked well'til about three or four months ago. Since then, all of these are more like meme stocks, they're fashionable, they're like, oh, let's give a lot of money to Bitcoin, the value goes up, now it's 1/3 of what the value was a year ago. Another way to think of a cryptocurrency is more fascinating, and that is, it's actually a computer system. Think about what a Bitcoin system actually is: you are instructing a computer to say, take 100 pounds out of my ledger, 100 bitcoins out of my ledger, and transfer it over to somebody else, it's an instruction, I'm giving an instruction, take 100 pounds out of this account, put 100 pounds in that account. And somebody called Vitalik Buterin came up with the idea, he was 20 years old when he invented that, now he's very old, he's about 28 years old, but he was 20 years old when he came up with the idea, he said, hold on a minute, that is actually a computer program, so how do I make it an explicit computer program? What he did was to add an if-then statement to that question of transfer money from A to B. So what is that if-then statement? Well, two possibilities, one is, let's imagine, for example, you're buying flight insurance, your plane takes off from London to Paris, you buy flight insurance, if your flight is delayed by more than four hours, you get compensation, but you have to file for the insurance, they have to check all that stuff. If there's an if-then statement in the insurance policy, you say, if the flight is delayed by more than four hours, money will automatically be transferred from the airline's account to your account. But that means you need an oracle, you need something to say that was the time the flight actually took off, something to connect the crypto world to the real world. Or you could have a different one, I can send an instruction to an address, and I can say, if I transfer money to that address, send me back a picture of a Bored Ape, for example, so that would be like a vending machine, I just send something to a particular address on the Ethereum virtual machine, and it'll send me back a picture, a CryptoKitty, or something like that. So that's the idea, so that once you have that idea, you can do a lot more with the blockchain. So what can you do with the blockchain? Well, the neat bit about this computer is there's no keyboard, there's no mouse, there's no monitor. Instead, what it is, a bunch of distributed apps, what we call dapps. The way a distributed app works is, it's a program that runs on the Web, runs on the Ethereum virtual machine, but the data is kept on the blockchain. For example, if you're doing a sword and sorcery-type fantasy video game, your character in the video game might have a sword, so the characteristics of the sword, however, are stored on the blockchain. Maybe it's three feet long, maybe it's very sharp, maybe it can cut through metal, whatever, the characteristics are kept there, but it's rendered virtually using the graphics card on your computer. Now, how does a distributed app work?

So let's have this instruction:

if the temperature is below one degree centigrade on this particular day, then transfer 100 coins from this address to that address, it's a program. So what happens? Well, it's sent to the entire network, and every node in that network executes that transaction. They'll check the weather, is it below one degree? Okay, it is? Fine, let's transfer. Some computers might say, no, no, no, in my area, it's not less than one degree, so they don't transfer, but then all these computers have to get together, and then they have to say, okay, you know what? The consensus is the temperature did in fact drop by one degree, and therefore the money should be transferred. So it's a consensus of all these programs, which are being run on every single computer connected to Ethereum network, so what that means, of course, is very slow. You do not want to have one of these running your Tesla, for example, when you're driving your Tesla and have autopilot on, because a pedestrian comes in, it has to check with 1,000 other Teslas, is there a pedestrian in front of me? And then, you know, there's a pedestrian in front of you, too late for the poor pedestrian. But how does it reach a consensus? Before September, 2022, Ethereum used a proof of work consensus protocol. That means all these guys had to burn electricity in order to solve a meaningless mathematical puzzle just to get their right to validate. But in September, 2022, last year, they changed the protocol to proof of stake protocol. How does that work? Well, let's say you want to be what they call a validator on the network, what the validator does is simple, he takes the money, he buys the network currency, and deposits it into a special smart contract called a stake. So I put my money, I bought a stake on this. That stake has limitations on its withdrawal, I can't pull the money out whenever I want to, it stays in the system. Now, if you're a validator, you compile all the transactions, all these programs into blocks. Now, every now and then, a validator is randomly chosen to pick one block, and then another bunch of random validators are chosen to review and vote on that block. So in other words, you don't have millions of people all trying to verify it, a smaller group of computers just getting together and saying, okay, hmm, I think this is the consensus transaction that happened, and therefore this is the outcome of that program. The good news is, because this is done in a very simple way, you don't have to mine, you don't have to solve a complicated problem, so in the proof of stake network, the electricity consumption drops by 99% relative to a proof of work consensus protocol, so this is all good stuff. But interestingly, what happens if you lie? So if you're a validator, and you say, this transaction happened, when actually, that one happened, you do this consistently, they cut off your stake, so you lose your money, so you have no incentive to act dishonestly. If you collude, so if 100 people are chosen to validate a transaction, and those 100 people collude among themselves to give a wrong transaction, people don't trust the system, and the value of ether falls. So all of this is to prevent the system from going down. So how do you earn money? Proof of work, Bitcoin, buy lots of computers, solve these meaningless mathematical riddles, earn bitcoin for that; proof of stake, deposit ether to buy a stake, validate a transaction, and earn fees, which are called gas fees, based on your stake. Now, gas fees are super important for Ethereum, because the idea of a gas fee is, let's remember that a transaction in ether is actually a program, so that means, in other words, I can have maybe a 100,000 line program, if this happens, do that, if that happens, do something else, if something else happens, I can have a 100,000 line program, and this has to be executed on all the thousands of computers, which means it is going to take up a lot of time. So Ether says this is the maximum gas limit which is allowed on any individual program you write. So if you write a 100,000 line program, you run out of gas, the program will not be executed. And there's also a limit on the amount of gas you can spend, so there's a price of gas which you're going to do, so if you run into an infinite loop, if A happens then do B, if B happens then do A, it's an infinite loop, it'll crash the system, but it'll stop, the gas fees will stop you from doing that. But the gas fees allow us to reinvent the concept of interest.

Let's take an example:

when you deposit money in a bank, why do you earn interest? Well, the banker takes your money and lends it out to somebody else, and that person does something productive, a productive activity helping society, they transfer the money back to the bank, and the bank gives you interest. Here, why do you earn interest? Well, if you want to be a validator, but you don't have the money, what do you do? You ask your friends, hey, I need to be a validator, I need so much ether, and if your friends say, fine, I'm going to lend you the ether, you put it into a pool, and use that pool to become a validator, and then you earn gas fees, and you return a portion of the gas fees as interest. Nothing productive being done here, all you're doing is verifying the security of your transaction. So interest is being earned for a purpose very different from what the banks do. Banks increase the value to society, here, you're just verifying the security of your particular transaction. All right, let's go now to fungible tokens. Why do you need a token at all? What is a token? Well, if you think about it, let's say you've created a computer program which does something interesting, one thing you might do is, say, flight insurance, we talked about that a few minutes ago. So if you apply through my blockchain system for flight insurance, how do I charge you for it? I mean, I'm giving you a service, I'm giving you insurance, so I want to charge you for it. One possibility, I can charge you in pounds, or dollars. Problem is, dollars live in the real world, this is the crypto world, it's going to take effort and time to transfer money back and forth between the crypto world, the real world. Second, I can say, well, you know, I'm running on the Ethereum system, so I'll take payment in ether, fine, or I can say, I'm going to take payment in my own made-up currency, like an FTT Token for example, which we'll talk about in a bit. Now, what's interesting about this? Well, think about something like Helium. Helium is a system which runs a network like this to provide Wi-Fi access to people around the world. The way it works is, if you open up your Wi-Fi system for strangers to connect to your Wi-Fi, you earn Helium Tokens. The advantage of this one is that you create something called a network effect. People say, initially, okay, why would I want to be part of this network? Think about Facebook, why are we on Facebook? I'm not on Facebook, by the way, but people are only on Facebook because their friends are on Facebook. If your friends are not on Facebook, there's no incentive for anyone to be on Facebook. So that's a network effect, you only join a platform because other people are joining the platform. The same thing here, you only join a platform like Helium because everybody else is on that platform, providing Wi-Fi. But how do you get that platform started?

The chicken and egg problem:

people won't join it until there's money in it, but if people don't join, then there's no money in it. The token solves that problem. Initially, when you join, you get cheap tokens, but as the platform increases in value, the value of tokens goes up. And creating a new token in Ethereum is dead easy, you just use, for example,"The Ethereum Whitepaper," there's a four-line code snippet for implementing a token system. So there are literally thousands of tokens everywhere, but this allows us to reinvent the concept of capital raising. How do you typically raise money as a new company? Well, you come up with a business plan, you're an entrepreneur, you go to a bank, you go to a venture capitalist, and you say, here's my business plan, give me money, and then you beg, and borrow, and whatever, and then, eventually, you do a roadshow, standard process, collect money. Now, think about Ethereum, in contrast, how did Ethereum raise money? Well, Ethereum could have gone the traditional route, the traditional route would have been to go to a venture capitalist, say, we're establishing a new system, give us money, and we'll collect 1%, or maybe 0.1% of every Ethereum transaction like royalties, because we're opening up the system to everybody, and that's our cash cashflow coming out. It would have worked, except these guys said, we want to build a decentralized system, so we are going to issue tokens, ether tokens, give us bitcoin, we'll give you tokens. And the amount of tokens they raised was about $18 million to start the Ethereum system, that eventually became $122 million, so a really good investment over there. But that exploded even further, so everyone was saying, well, Ethereum can do this, and so they started a world of coin offerings. The idea was, give me money and I will give you coins in my new offering. So the sheer number of initial coin offerings took off dramatically in 2017, and so on. This was like an initial public offering, except, give me money, and I'm giving you my made-up coin, my token, and hopefully, the value will go up if this thing takes off. Good stuff, except the SEC came in and said, hold on, this is not a currency, this is a security, so if you're issuing a security to a US investor, you have to comply with all these SEC rules, and the SEC rules are pretty stringent when it comes to offering this kind of stuff, so ICO has pretty much dried out, very few ICOs being launched in the last few years. In fact, this is my favorite ICO, it was called the Useless Ethereum token. So these guys say,"The first 100% honest Ethereum ICO." They say, you're going to give some random person on the Internet money, and they're going to take it and by some stuff with it, possibly electronics, maybe even a big-screen TV, and this guy says, the UET ICO transparently offers investors no value, so no expectation of gains, no gains means fewer investors, few investors means few transactions, few transactions means no Ethereum network lag, you're done, it's even called UET. And they sold $350,000 worth of them, which appreciated more than 240%, that's how obsessive people were about buying this kind of stuff. And then the next development after the ICOs dropped off was the concept of a decentralized autonomous organization, DAO. The idea here is you have a group of people who are like a venture capital fund, so all of us deposit money into this fund and somebody applies to us to say, let's start this project, I want money to fund this project, and people vote based on their token ownership, I want to invest in that project. The problem, of course, is, this sounds great, right, decentralized, the problem is, are you a corporation or a partnership? And in May, last year, a DAO was sued because it lost money, and they said there are no titles of incorporation here, you just gave money to be part of a DAO, which means you're a partnership, and if you're a partnership, in America, you have unlimited liability. If you're a corporation, the maximum you can lose is the amount of money you put in, but if you're a partnership, like Lloyd's of London nearly went under because all the names were called upon to give money beyond the amount they had already invested. So you're not quite sure, is this a partnership? You didn't think it was a partnership, but it can be treated as a partnership. Now, let's go to the next stage, so there's a fungible token, and a non-fungible token. What's a non-fungible token, an NFT? The idea here is simple, it's a unique token. What does that mean? Well, the protocol is Ethereum 721, which just basically says, add a token ID to your contract number so that that's unique. That means, in other words, when you write a program, you're just saying, send a message to this particular address with this token ID, it will return you something, a picture of a monkey, a picture of kitty, a picture of something else, and people said, whoa, that's worth a lot of money. How much money? Well, Beeple sold an NFT, this one here, for $69 million. And basically, until then, the maximum he'd ever sold something for was $100, this was $69 million. And what is it? Basically, a computer program which, every time you type it in, would go to one picture over there. You can download the picture for free, but in this particular case, of course, this guy who paid that money is the only guy who, the token would come back and say, this pointer was purchased by this person. Jack Dorsey sold his first tweet for over $1 million. To put that into perspective, if you decided to buy a physical painting of water lilies by Monet, that was only 54 million. Okay, well, you know, some people did that. Here's a CryptoKitty cat for $172,000, and of course, this is the Bored Apes, there were insider trading allegations. This guy said, a guy called Nate Chastain, who was head of OpenSea, basically would figure out what was being sold on Monday morning, buy them on Sunday evening, and then put them on the market on Monday. He was accused of insider trading, but insider trading, it sounds weird, right? I mean, this is a pointer. I mean, how do you know it's going to go up in value? What do you know about the true value of this stuff? If you don't know what the true value is, how can you be an insider? There's no true value, what do you do? So if you guys find this fishy, some heads nodding here, no problem. This is something called L'Eau de Distance, which is a virtual perfume.(audience laughs) This is a non-fungible token for a virtual perfume. So essentially, when you click on that, and you can go to that particular site, it will give you this. I downloaded it, I mean, it's free, but I didn't buy it. Somebody bought it, and so that's the smell of NFTs. But if you think about it, that's actually quite interesting. Initially, of course, you have all these stories about CryptoKitties, and Beeple, and all that, but how do you reinvent borrowing? Think about, you want to buy a house. In the traditional finance world, what do you do? Go to your bank, the bank will send an appraiser over to your house, the appraiser will say, this house is worth 1 million pounds, fine, I'll give you half of that, half, 500,000, security against your house. But now, you want to do this in the crypto world, how do you do it, how do you sell half your house? I mean, raise money, half your house? And first of all, number one, you don't want to leave your house, number two, how do you do it for half your house? That means without selling it.

Answer:

maybe you do something, an NFT on your house. So you issue an NFT on your house, which is pointing directly to your title deed on the house, and then somebody buys that, so five years from now, when the loan comes due, you toss a coin. It comes up heads, he gets the entire house, tails, you keep the house, so you've a 50% of keeping the house and a 50% chance of losing the house. Suppose you don't want to do that, you're like, now hold on a minute, I don't want a 50% chance losing my house, no problem, you go into the market, buy an NFT of a different house which is in the same class of houses, so a house in the London market, in this area, and then replace your NFT with that NFT, so you never actually go through with that. But the point is you can do this in the crypto world, like raising money, without ever actually having to go to a traditional financial authority. Now, to make this useful for business, we need to solve a problem, and the problem is called the blockchain trilemma. Vitalik Buterin put that up, and his thing was, you can have two of these three things, you can have security, you can have stability, or you can have decentralization, you cannot have all three. If you think of a traditional chain, like Bitcoin, like Ethereum, they're decentralized and they're very secure, but they're not scalable. If you think of something that's scalable and secure, it's high throughput chains, but they're not decentralized. Or you think of a multi-chain ecosystem, they're decentralized and scalable, but they're not secure. Solving this problem is one of the biggest things that crypto has, how do you make all three things happen? Very, very tough to do. And besides, you have a problem with regulators. Your regulator comes to you and says, okay, so how do you ensure the security of your transactions? You're like, hmm, I've got some miners in Russia who are looking at the security, your regulator's not going to be really very happy. So some banks said, okay, we can forget about all this and go with this route here, which is basically a permission blockchain, a private blockchain, nobody else is allowed to participate except for 10 banks who know each other, and they're the only people allowed to read or write on that blockchain, that's one possibility. But the last part of my talk today is, how do you move from the crypto world to the real world? And this is the biggest problem that crypto has. If you move from one blockchain to another, that's difficult enough, you need to have a crypto bridge, like, let's say I have ether but I want to convert it to, say, Dogecoin, how do I do this? They're in different blockchains, how do I do that? One thing I can do is send ether to an address on the Ethereum blockchain, that will go to an offline computer, the offline computer will then log on to the other computer, the Dogecoin blockchain, and they'll transfer money across from one to the other. It'll open an account for you on the Dogecoin blockchain with the amount of money you transferred over at that exchange, that's a crypto bridge, from one to the other. It's doable, it's tough, but it's doable. But moving from blockchain to the real world, that's really difficult. Matt Levine has a very nice article in "Bloomberg Businessweek" in October, the entire "Businessweek" that week is called "The Crypto Story," and it talks about all these issues, but he basically defines crypto as a bunch of things, like a set of tokens which are worth arbitrary amounts of money, a set of ways to create new tokens and distribute them, and try to make them worth money, a way of recording your holding of assets, a way to write contracts and programs which are interchangeable. But none of these get to that problem of how do you go from the crypto world to the real world. Let's look at that real financial system. One thing we can say is, the crypto world, as I said in the beginning, an elegant, self-contained system with permission-less innovation. What does that mean? If I want to write a computer program, I don't have to ask Vitalik Buterin's permission, I don't have to ask anybody's permission, I just launch it. Somebody wants it, they can pay for it, and then they can take it, I don't need anyone's permission. So there are two opinions, one opinion is a streamlined, modernized, innovative evolution of the traditional system, the other one says it's a chaotic devolution of the traditional financial system that never learned important historic lessons for fraud, leverage, risk, and regulation, two totally different opinions about crypto. So let's take a look at some of these things. Centralized intermediation, think about losing your password to your bank account. What happens if you lose your password to your bank account? Well, you call your bank, and the bank goes through a bunch of identity checks, and says, fine, I've reset your password, it's test123, please don't lose it this time, something like that, or on the website, there'll be a little link, saying, lost your password, you type in your thing, they'll do it. What happens if you lose your password to your bitcoin wallet? Well, here's an example. This poor sod, Stefan Thomas, was living in San Francisco, stored his private key on what we call an iron key, which is a hardware device which allows you 10 tries to get your password correct. He's done eight, he's got his private key on the iron key. After 10, it bricks up, there's nothing, no one can extract it from the iron key. He's got eight tries, and he's got two left, and the password is worth about $220 million, in theory, he's worth a lot of money. There's also, every now and then, stories about this poor guy in Wales who had his password stored on his hard disk, which he threw out into the garbage heap, and so every now and then, he begs the city to let him dig through the garbage heap to try to find his hard disk, but they have said no. So how do you invest in crypto? If you're a mainstream investor, you can't actually buy crypto, because what happens if you try to buy crypto is that the regulator will say, hold on a minute, what is this? You have to have assets which have some value, which have something, they're very uncomfortable with this. So you can buy Bitcoin futures, everybody understands futures, the traditional financial system, so some companies, some hedge fund will buy the bitcoin, and sell you futures on it, so that's one way. Or you can buy Bitcoin future ETFs, which is like a stock, so it's a stock wrapper around a future, which is in turn around Bitcoin. But suppose you want to directly buy crypto. Well, if you wanted to do it directly, I could go to one of you and say, here's some money, give me some crypto, here's my wallet, we can do that. But if you want to do this through an exchange, will the exchange send you the crypto? What do you think? No way. Why?

Think about it:

suppose you go to Coinbase and say, I want to buy one bitcoin. They say, sure, give us your credit card number, you send them your credit card number, they transfer $20,000, they send you one bitcoin. And then you call your credit card company, say, my card was stolen and someone's been using it to buy bitcoin, I want my money back. And so they'll call the exchange and they'll try to get it back from the exchange. The exchange doesn't want this. So once they transfer the money over, they can't get it back, so they're not going to transfer the bitcoin over, they will keep the bitcoin for you. But that raises a question, who guards the guardians? And there are lots of examples of exchanges which have been hacked. So if you've got the exchange with them, it could be that you lose your money forever, but again, that's as long as the people stealing your money stay in the crypto world, moving to the real world is tough, and this is the point I'm trying to make again and again. So in this particular case, it turns out the Bitfinex hack, over here, in 2016, this couple, Ilya Lichtenstein and his wife Heather Morgan, basically stole that money and moved it through a whole bunch of different transactions, but it's all on the blockchain, which meant that every single one of those transactions could be traced. These guys had stolen $3.6 billion, were they living high on the hog? No, they couldn't get it out. And eventually, they bought one Walmart gift card, which was connected to a real world address, and so they got caught. That's one of the issues here, going back and forth from the crypto world to the real world. And of course, if you think about other things which are in the crypto world, like stock exchanges, in a security exchange, you have all these decisions which are made, how much will they lend you, what's the size of your margin account, what role should the clearing house play? Unfortunately, the crypto world is supposed to be decentralized, it's not decentralized,

they have the same problem:

how do you set leverage limits, what is the level of collateral which you need to hold? And coming back in here, how to set a leverage limit?

FTX-Alameda is the classic example:

leverage limits for everybody except for Alameda, who could borrow a ton of money, and eventually almost bring down the entire system. Stablecoins. One way to think about this is, okay, a stablecoin is a crypto token that's always supposed to be worth $1. How do you do it? One thing is collateralized stablecoin, example, Tether. Tether says, you give me $1, and make sure your stablecoin, Tether, is always worth $1. So in a way, you're giving them money, they put it into an envelope, and they put the envelope in the safe. You come back and say, I want my money, they pull your envelope out of the safe, give it back to you. But this doesn't earn anything, so what they really do is they lend out your money to earn money, but they can't tell you this because if people lose confidence, they'll all ask for their money back, which is a bank run. So they say, we are auditing our stuff, but they won't tell you the result of the audit, it's all very, very secret, but they claim, trust us and everything will be fine. Or you can have an algorithmic stablecoin. What does that mean? That means you put in 100 bucks and get back something that's worth 100 bucks, but you guarantee it by a large amount of a volatile cryptocurrency. So if you want that $100, you're putting $500 of bitcoin, or something else to guarantee that value. But the problem here is, if people lose confidence in this, they will start immediately asking for their money back, which reduces the value of the algorithmic stablecoin, which reduces the value of the cryptocurrency, and that is called a death spiral. Okay, so last thing, crypto exchanges. In a regular stock market, you have market makers and limit order books, so in other words, you can say, I want to place an order for 100 shares of Microsoft at this price. If the price goes above that, I don't want to pay that, it's a limit order. So if its price stays below that, I can buy, otherwise I never execute the trade. This doesn't work in crypto because every time you put in a transaction, even if you've withdrawn an incomplete transaction, you have to pay for it with a gas fee. You're giving an instruction, the instruction means a gas fee has to be done, even if it's incomplete, it's too expensive. So how do they work, how do these guys set up a market making system when this problem is there? Answer is, and this is unusual, because it does not exist outside the crypto world, so this is an example of what I was saying, that these guys have invented new ways of coming up with finance that the traditional finance world never uses. So what is the example here?

It's simple:

an automated market maker is a market maker that develops equal amounts of ether and a dollar equivalent stablecoin, so in this case, say 1 ether is at $1,200 right now, so you deposit 1,000 ether and 1.2 million equivalent stablecoin. The idea is, one algorithm is called a constant product market maker, keep the product of these two constant at 1.2 billion. Now, let's take, somebody says, I want to buy 100 ether. The fund contains 1,000 ether, so now it has 900 ether in the fund.

All the algorithm does is simple:

I want the product to stay the same, the product, remember, is 1.2 billion, that has to stay the same. So how do I do it algorithmically? I say 1.2 billion divided by 900 is 1.33 million, I already have 1.2 million USDC in that so I need you pay an additional 133,000 USDC, which means that's the exchange rate. I don't need a human being, it's automatically done by the system. So this thing here is not there in the real world, market makers don't work like this, but you have to do this in the crypto world, which is an automated market maker. So if you are lending your crypto against real assets, you have margin lending, but we talked about unsecured lending for real assets, we talked about DAOs a few minutes ago, but now it becomes more complicated, because let's say you have staked to ether to become validator, so now you're a validator, you staked it, what does the system do? It gives you a receipt for your staked ether. But the receipt is valuable because it's saying, I have staked this amount. So what do I do? I take that receipt and use it as collateral to get another token in a different currency, and I can keep doing this to get something called yield farming, I move from token to token trying to get as much interest as possible. The perfect example was the Olympus DAO, they actually had this on the website, it's called the (3, 3) prisoner's dilemma. It basically said, as long as nobody liquidates their money, we all will have increasing values. It's kind of like, okay, excuse me, right. So you could do anything with blockchain, even buy a blockchain face cream. Real world arbitrage is super difficult to do. Why? Because if you have an idea, you go to a hedge fund, the hedge fund say, that's a great idea, thank you very much, bye, and they'll implement it themselves. In the crypto world, you can have lots of ideas because they're running public codes, so you can find lots and lots of ideas. How do you do that? Something called a flash loan. An example of flash loan is, suppose ether is trading at two different prices on two exchanges, what do you do? Borrow $100 million, which you don't have, borrow $100 million from this, two, use $100 million to buy a token on decentralized exchange A, sell the token on exchange B for 110 million, return 100 million to the lending protocol, plus a small fee, and keep the leftover $10 million. And if you're looking at this and saying, wait, where's all those 100 million coming from? The answer is it's a computer program, the entire thing is one program. So you send the entire program across, it will only execute if all the five steps work out. If the price of ether changes in the middle, the program will not execute, you don't lose any money at all, it is totally free. So that's a flash loan, you're writing an entire computer program,

and that means the problem is this:

somebody writes a contract, bugs occasionally let the user put in one token and get back two. What do you do? A flash attack, so you write an entire program, like this, send it through, and basically, someone writes a program, puts in 1 billion tokens, gets back 2 billion tokens, and blows up the smart contract completely. All right, last problem, front-running. Front-running is, somebody figures out what you're doing and trades against you. In the real world, this is tough, so lots of people pay money to be fastest in the line. There's a book called "Flash Boys," by Michael Lewis, which talks about this. In the crypto world, this happens all the time. To have a way to make money, you have to put into the blockchain, and everybody can see this, and so somebody says, hey, this is cool, I'll pay a higher gas fee and get ahead of you. I can see your program to make money, so let me make money ahead of you, just increase the gas fees. Very tough to actually implement this. And finally, because of all these things, crypto has done what we did in finance in 2008, we have got a crypto financial crisis. What have we done? If you think about Satoshi Nakamoto, he minted bitcoin to get away from all these problems, because of 2008, he said, I don't want to deal with this, but this is everything, it's a new financial system, but the problem is the traditional finance people who got in there said, hey man, this is fun, it's wide open, permissionless innovation, and everybody's getting rich, so why don't we get in as well? And so we have TerraUSD and Luna, we have Celsius, we have contagion, because all these tokens are tied to each other, one token goes down, everything goes down, and we have FTX-Alameda most recently. So, what's the future of crypto? Well with real assets, it's tough, I don't know how to do this, nobody knows how to solve this, how do you go from the crypto world to the real world? Digital assets, I can explain this, maybe that is the future of the world, all of us with virtual reality helmets, we stay in a digital world, real world is less and less important. In that case, maybe crypto is actually the way to go. But at the moment, I can probably say this part here is the part where crypto has a problem. Okay, and that's it.(audience applauds)- [Victoria] Do we have any questions from the floor?- [Questioner] This is a really simple question: putting aside Ponzi scheme concerns, which I'm not going to talk about, a very simple question, which is, when things fail, you need a backup plan. In traditional finance, you could usually find a written record, it might be written on a computer program, but it's written, it's in ink. I had the impression that that is not really possible to do in crypto, so if the whole, what my kids jokingly call the Interwebs, crashes, if we just lose power and don't have any access to the digital world, what happens?- You mean to say within the digital world alone, not connecting to the real, traditional finance world.- [Questioner] Right, because you've already, I think, adequately explained why that connection is difficult.- So, within the digital finance world, as long as people are willing to buy and sell different types of digital assets for different prices, that world is self-contained and it's very elegant. Our problem is, essentially, there are two things which happen, one is regulation, traditional financial intermediaries, regulators will not allow them to enter the digital finance world unless they have some kind of wrapper which makes them look like a traditional financial asset, example, Bitcoin futures, for example. I don't see that changing anytime soon, especially because the current SEC Chair, Gary Gensler, is highly against the whole idea of treating the crypto world as anything different from the normal world. I mean, he's using things like the Howey test to decide whether something is a security offering or not. So I think, in a traditional financial system, I don't think we need to worry so much at the moment about these guys invading this. You got a bunch of tech people who don't understand finance, and you've got a bunch of finance people who don't understand tech, so how do you get the two guys to talk to each other, that's the problem right now. If you look at Sam Bankman-Fried, with FTX, he had literally no idea how the actual world worked, if you read what he's talking about, it seemed to be somewhat of a Ponzi scheme, believe in me and the value of everything will go up. It didn't.- [Questioner] How about, if you go by real world as, let's say, buy land, or gold, backed by digital assets, will it make any difference, the best of both?- What do you mean? Can you elaborate on that?- [Questioner] For instance, I have a house, and I want to raise some funds on it, so I have a physical-backed asset on which, if I could able to tokenize and raise funds, is that doable?- Yes, that's what I talked about with Martingale fractionalization, that's the whole idea. But it sounds weird, because the whole point about this is, I'm borrowing the expected value of my house, which is the value of my house times the probability that I might lose it or not, is the crypto way of thinking of it. For a risk-neutral person, I don't care whether I get the actual value or the expected value.- [Questioner] But let's say, in a business, like shares?

- Well, I'll put it this way:

this is all the beginning of this. That whole thing about, going back to your question about the digital world and the real world, and how you bring them together, I think we are only now beginning to grapple with the complexity of that problem. The last five years, the crypto world has been going great, except for the fact that nobody has figured out that part. Now, there is almost like a crypto winter. Okay, there may be a way, there my be real-life solutions which they're trying to solve now, but it's still the beginning.- I might hijack the questions, if that's all right, with my privilege of chair. As you know, I'm very interested in regulation of cyberspace and digital technology, you touched on regulation. Very brief thoughts about the prospects of regulation of cryptocurrency?- Very good question. The thing, as I've talked about with Gary Gensler, is that Gary Gensler is talking about the crypto world as just like another asset in the real world, so a lot of regulation is, I'm sorry, you may call it not a security, you may call it an initial coin offering, but that's a security, which means these are the rules you need to jump through. So for example, the Howey test, which I talked about, was created in 1934, it's still being used today to do something which was never envisaged even 10 years ago, forget 1934. So one bunch of regulators are saying this is no different from a traditional financial instrument, so we can regulate it the same way. You've got another bunch of regulators who are saying, okay, let's find ways in which to treat this as new assets in themselves. The second problem is, again, going back to the real world, unless you can show that it has an effect on real cash flows or real assets, it's tough for a regulator to do something about it, you're basically saying one arbitrary number has crashed. How do you do this? Think about Nate Chastain and OpenSea, was that insider trading, was it not insider trading? Some people would argue that is not insider trading, you've got a joke card, how do you know the price is going to go up, where's the insider knowledge there? It's a very tough question, and I think the basic answer, quick answer, is, as far as the regulator can see a parallel between the digital world and the real world, they're going to use that to try to assert their authority over this.- And polarized views, as with a lot in the digital world around how to govern it. And with that, folks, I'm afraid we have run out of time, so it only now remains for me to thank those of you at home for joining in and tuning in today, and those of you in the audience for coming to see us in person, and of course to Professor Raghavendra Rau for another superb lecture. Thank you very much.(audience applauds)