Recently I had the pleasure of stepping into the studio with the illustrious Mark Pesce to talk about Libra. Mark had been promising me that he wanted to get me onto the podcast for some time for various topics, and the timing was finally right for us.
Our time on the air was preciously short, and these are questions with deep answers, so here are my thoughts and notes that I had written ahead of recording (somewhat updated to reflect some of the conversation, so please excuse a degree of fragmentation of the information).
You can find Mark's excellent podcast here: https://cryptonomics.show/2019/08/14/episode-2-01-sign-of-the-times/.
CRYPTONOMICS Podcast Notes
It's quite interesting to consider what Libra could become, and how it could change the world.
It can be argued that all blockchain projects are maximalist games; you need a critical mass of users for a blockchain network to be worth something, but you need all users for it to be maximally effective and maximally valuable. This is reflected in Metcalfe’s Law - that the value of a network is the square of the number of users of the network - referred to earlier in the podcast.
So it’s important to think in terms of network effects when thinking about the future of blockchains, because the technology of blockchain gives us the tools to capture the value of a network effect in a decentralised way. Most people don’t realise this but network effects are among the most valuable assets in the world. They are sources of extreme market efficiency, and that efficiency is valuable.
Getting back to Libra, it's interesting to consider the relative sizes of the user base. Let's say a user base of ~350m for the US dollar (i.e. the popualtion of the US). Now consider Facebook's 2.7b plus the other unique customers that could be potentially onboarded via the other companies in the consortium. Who’s got the biggest network effect now? And what does that mean for fiat currencies?
Is Libra a Cryptocurrency?
This is a bit of a rabbit-hole-like question.
I actually argue for the negative, that Libra is not a true cryptocurrency. In short, this is because the unit of value is not intrinsic to the system. It's essentially, functionally, a depositary receipt (or a bearer promise to repay), that's deployed on a blockchain (the form of the instrument). My view is that the "crypto" in cryptocurrencies doesn't just refer to the fact it uses cryptography, but that it uses cryptography to develop a completely trustless system. I think this was the point of Satoshi's innovation.
I've been involved in blockchains long enough to see the technology evolve, and terminology along with it. Over time bubbles of jargon appear, based on some particular subgroup’s use case, and they typically don't entirely agree with the rest of the DLT landscape. I gave a brief talk last year on this called “Coins, Tokens and Stablecoins” at the Spark Festival here in Sydney in an attempt to provide some draft definitions - I'll get into those later. In all cases however it's super important to try and agree on terminology that works across the industry to give us a common language.
Before we get into that though, it’s even more important to understand money a little bit.
Money is actually a tricky thing to understand, because it seems to be a combination of phenomena and technologies that need to be individually appreciated, including human psychology, society-specific shared beliefs, network effects, macro/micro economics, fractional-reserve banking and debt, and national governance.
Understanding money first of all means understanding value. There are actually still a lot of people who dont realise that fiat currency, the stuff you have in your bank account and in your physical wallet, isn't “backed” by anything any more. We dropped the gold standard a while ago.
In essence, fiat is backed by guns, as well as its own network effect. You can think about the government, governing things for the prosperity and welfare of a population, as a well-armed insurance company. So some say fiat is "backed by force", and the mandate of a democratic government, as elected by the population, is to perform this function of macroeconomic management. And backing that mandate is the judiciary (armed police) and in some cases, the military. The biggest tool in the macroeconomic toolbox is the issuance of money and the management of its supply.
So 'money' is just the instrument that everyone in a given country has agreed to use to represent value in general, even though it’s (arguably) intrinsically valueless. And so enter the debate: because the currency is used in this way by a whole population, and has generated a maximal network effect (on the scale of a country), then perhaps the aggregate ability to efficiently undertake commerce is actually the thing that gives money its value, captured and represented by these fungible "tokens" we call dollars. Okay so maybe it does have value given its utility in commercial scalability. And certainly, people believe dollars have value. At least most people do.
People have been trying to invent an alternative to fiat currencies for a while, often citing displeasure with the corruption of government or the inferiority of the way fiat has been implemented to date. Let's recap some of that. The technology of money has three key functions: to act as a unit of account, a medium of exchange, and a store of value.
Now, the first decentralised attempt was Bitcoin in solving the double-spend problem. Bitcoin has so far been a poor unit of account due to its price volatility. In order for a token to effectively act as money, its purchasing power must remain stable against goods and services over time. That may change if its network effect and functionality improve.
But after Bitcoin came many others. And now we have a landscape of Cryptocurrencies, Tokens, and Stablecoins.
However the defining feature of Bitcoin was the fact it was permissionless. Anyone can use it. You don’t need anyone’s permission to receive it or send it, unlike money in electronic form in a bank account where you require a company to provide you a “money service”. The network wasn’t secured by a set of intermediaries you had no choice but to trust; instead it was secured by cryptography with a dash of game theory. (Reading about how proof of work actually works is pretty interesting).
Hence the term Cryptocurrency: a decentralised computer program that securely manages a set of units of value. We can call these units of value a 'currency', and the techniques to manage them heavily rely on cryptography. The subtlety here is that the units of value need to be intrinsic to the system - blockchains can only be completely trustless (secure) if what they’re managing is “on-chain”. As soon as you start to reference data outside of the information universe that the blockchain can see, you require a trusted entity to inject that information from the outside world. While this is very useful in many use cases, if the information being injected is about the units of value, I argue that you’ve broken the original definition of cryptocurrency.
So, Is Libra Really a Cryptocurrency?
Not according to me because it’s merely an instantiation of external valuable assets. It’s a form over function thing. Tether sits in this camp for me too. It’s not a true cryptocurrency, but rather more like the gold standard where the instrument borne by the holder was a promise to redeem some external value.
Libra is certainly using a decentralised system, but only to operate what is effectively a centralised (off-chain, trusted) product. Facebook could do this centrally, but I think there’s a degree here of using blockchain in order to appear more decentralised than what they are in reality.
So while they call Libra a cryptocurrency, it's really only about a 2 or 3 in terms of decentralisation, rather than a 7 or 8 on the scale for say Bitcoin and Ethereum.
That’s not to say Libra is necessarily a bad thing, or that Ethereum is the right way to get the technology adopted either. And that’s not to say it wont be more decentralised in the future - this thing could really take on a life of its own, rather like the invention of fiat last century.
Now in terms of the Libra consortium, a permissioned network is often just a convenient way of implementing/governing these “trust bridges” (i.e. oracles). They’re often a proxy for each other: if it’s permissioned, look for the external information to be brought into the blockchain; if it’s dealing with external information or assets, look for the permissioned actors. If you know who the actors are who are acting as this points of trust throughout the network; it’s easier to manage that using their reputations, audits, regulation, etc)
And now of course we’ve seen many others since bitcoin: many public, permissionless blockchains such as Ethereum, Z-Cash, and others, but also other configurations of blockchain, including private ones (the data is not open), and permissioned ones (the validator set is restricted).
Getting up to speed on Stablecoins
Some draft definitions:
A tradable unit of value to act as a store of wealth and medium of wealth exchange.
Examples: bitcoin, bitcoin cash, dogecoin, litecoin, ripple, z-cash
A tradable unit of access to, and ownership of, a (utility) network or market. Tokens are network-specific and may be non-fungible.
“Network Effect Access Token”
Examples: Ethereum, Synthetix, Golem, Quantstamp (some good ones, lots of bad ones)
A cryptocurrency (coin or token, usually coin) whose value is held stable w.r.t. an external asset. They come in trusted and trustless flavours.
Examples: Synthetix or MakerDAO, and Tether/ Libra/TrueUSD
Is it hard to keep stable?
Algorithmic - yes; but should theoretically work with sufficient users/demand
Trusted - no; need to maintain ‘trust’ in the existence and redeemability of the external asset.
Will Libra be redeemable? That is currently unclear. It may have a big enough network effect to not matter.
Move - A viable smart contract platform?
Okay so if you consider Libra to be a money technology, it’s bringing something that fiat has struggled with for a long time for all the wrong reasons: programmability.
But that’s not all they’re bringing, to my earlier point, Libra is many times more viable as a money technology because Facebook is not just bringing the technology, it’s also bringing billions of users. In a sense, Libra is on track to being the decentralised version of Apple and Android. If Libra continues development unimpeded, in no time at all we’re going to see the equivalent of an "app store" built on Move for decentralised applications (a dapp store), but an app store that is able to offer things like savings accounts and insurance instead of Fruit Ninja. It very much has the makings to accelerate wealth creation to the underbanked and materially improve the lives of billions of people.
However, it’s still permissioned, and there’s still huge scope to exclude a lot of the global population due to both domestic and international laws. Just consider the Indian ban on cryptocurrencies, domestic sanctions, and AML/CFT regulations. And as we know, questions remain about how the transaction data will be managed. I.e. what controls are in place to maintain privacy on a user level, and indeed at a population level given the big-data possibilities for authoritarian control? Who will have access? If we can use big data science today to shift democratic outcomes on Facebook, imagine how much more powerful it would with financial data on Libra.
Will Libra succeed?
Libra is equal measures exciting and terrifying.
There are lots of opinions on their ultimate strategy, but I'm not sure that complex strategy is that simple. Good strategy is thinking in terms of risk and probabilities.
Libra gives facebook optionality. It could:
- Sell data
- Sell services to on and off-ramp onto the system, e.g. KYC, or indeed provide these for free to benefit another monetisation strategy
- Benefit from returns on the investments of the basket
- Retain the relevance of Facebook, or rebuild using a utility token economy
- Ultimately own the blockchain space
To answer the question, I believe yes, it will succeed, and I also expect it to mutate and grow over time. For example if it doesn’t succeed in America, they can just block Americans and still be very successful elsewhere such as as huge swathes of Africa and Asia, particularly when you consider the wider financial product possibilities of programmable money.
I think it's important to note that I’m an engineer, and an optimist. Whilst I understand the power of incentives, I tend to look at this stuff through a rosy lens of technical and product possibility by default, rather than through probable human failure modes.
Admittedly, Facebook has succumbed like many companies before it to perverse incentives, but I think they have an opportunity with Libra to do things better. And if you consider Zuckerberg the person, you'd have to think he's considering his legacy. He's got enough of that fiat money.
There are all kinds of capital barriers to a better life. Even in wealthy countries such as Australia. Capitalism is a lumpy road, but Libra has the power to smooth out the journey.