Facebook’s Libra: blockchain, but without the blocks or chain

This is just one of a series of Alphaville posts on Libra coin, which we are calling Breaking the Zuck Buck, in which we will seek to show how nonsensical, pointless, stupid, risky, badly thought-out and blockchainless the whole thing is.

We’ve been told that Facebook’s shiny new Libra coin is “a low-volatility cryptocurrency”, supported by the Libra Blockchain, a “decentralised, programmable database”.

And at first look, Libra coin is just what it says on the tin: a cryptocurrency — or more specifically, a stablecoin — like any other. It has the characteristic “white paper”, the characteristic promises to empower people across the world, and even the characteristic handy charts to show the likely direction of the project. From the white paper:

Libra White Paper

But look a little closer and you soon realise that Libra coin isn’t really a cryptocurrency at all, and that the “Libra Blockchain” — which is, notably, always written with a capital “B” — isn’t really a blockchain.

What's more, it isn't at all clear why you would possibly need a blockchain, or anything calling itself a “cryptocurrency”. Given that Facebook simply appears to be trying to build a global pseudo-banking and payments network, there doesn't appear to be any good reason why you would want to do this using blockchain tokens, which as the authors point out themselves in the documents, have so far proven volatile and difficult to scale. China's social media-turned-payments giant WeChat Pay — who Facebook is surely trying to compete with on a global level — doesn't use the blockchain. Nor does PayPal. Or Venmo. Internet money doesn't need to be blockchain money. Which is maybe why Libra coin isn't really blockchain money in any meaningful sense of the word.

Could it be the blockchain stuff is mainly PR . . .?

Confusingly, sometimes it sounds like the white paper is almost explicitly telling us that the Libra Blockchain is not a blockchain (or not one that would fulfil the usual criteria for a blockchain). Like here (with our emphasis):

In order to securely store transactions, data on the Libra Blockchain is protected by Merkle trees, a data structure used by other blockchains that enables the detection of any changes to existing data. Unlike previous blockchains, which view the blockchain as a collection of blocks of transactions, the Libra Blockchain is a single data structure that records the history of transactions and states over time. This implementation simplifies the work of applications accessing the blockchain, allowing them to read any data from any point in time and verify the integrity of that data using a unified framework.

To say something is something, but unlike the previous version of that something because its defining characteristics are different, is a bit like saying:

“Unlike previous loaves of bread, which view the loaf of bread as a collectively baked set of ingredients, the Libra Loaf of Bread is a single ingredient structure that bakes the components sequentially over time.”

Meanwhile, in the rather grand-sounding “State Machine Replication in the Libra Blockchain” document detailing the Libra Blockchain consensus, we’re given this explanation of the way the Libra Blockchain works, which seems almost contradictory to the last description (emphasis ours, again):

Blockchains and cryptocurrencies have shown that the latest advances in computer science, cryptography, and economics have the potential to create innovation in financial infrastructure, but existing systems have not yet reached mainstream adoption. As the next step toward this goal, we have designed the Libra Blockchain [1], [2] with the mission to enable a simple global currency and financial infrastructure that empowers billions of people. Every block in the blockchain consists of a set of transactions, which are agreed upon by the participants using a consensus protocol. At the heart of this new blockchain is a consensus protocol called LibraBFT — the focus of this report — by which the transactions are ordered and finalised.

It almost feels like several dozen people wrote these documents and (ironically) failed to reach consensus on what they were actually trying to communicate. Oh wait:

Talk of decentralised structures is everywhere in the frustratingly wordy but detail-light introductory material. But one thing which can be unpicked immediately is that the system only aspires to achieve decentralisation at some point in the future.

For now the unique benefits of blockchain’s “distributed governance” are only theoretical because in the first instance the Libra Blockchain will be coming to market in fully “permissioned” form, effectively centrally controlled by the founder Libra coin holders.

To save costs and to introduce efficiencies even further, the Libra administrators even float the idea of archiving off data (possibly for a fee) in order to save space — something that would be considered anathema to most self-respecting fans of immutable blockchain technology.

In short, most of the features that make a blockchain a blockchain don't seem to be there.

But the problem with arguing that the Libra Blockchain does not actually constitute a blockchain is that — as we have pointed out before — outside of the world of the public blockchains that underpin cryptocurrencies like bitcoin, the word doesn’t have much meaning anyway.

So maybe a “single data structure” can be a blockchain after all. And maybe Libra coin, a bit like Ripple (XRP) before it, can be considered a cryptocurrency even though, per the standard definition via Investopedia:

A defining feature of a cryptocurrency, and arguably its biggest allure, is its organic nature; it is not issued by any central authority, rendering it theoretically immune to government interference or manipulation.

Moving things along . . . with the power of Jedi mind-tricks.

Another important aspect of the Libra Blockchain is Move, its new programming language. This programming language will, says Facebook, allow users to define their own smart contracts in the future. Smart contracts are agreements written in code whose clauses are automatically enforced when a set of pre-determined criteria is met.

So far, at least in the decentralised cryptocurrency context, they've proved highly unreliable. This is partly because they are written in code and not human language, which means even when they’re written by the world’s top programmers, errors can easily be introduced.

But Facebook has pre-empted the human error problem by creating programming language that, somewhat incredibly, can . . . read your mind.

From the white paper:

Move takes insights from security incidents that have happened with smart contracts to date and creates a language that makes it inherently easier to write code that fulfils the author’s intent, thereby lessening the risk of unintended bugs or security incidents.

What Facebook possibly failed to pre-empt, however, is that by calling its system Libra Reserve it brings to mind the Liberty Reserve, a Costa Rican-based attempt to create something similar.

We’re sure this one will end much better though.

Copyright The Financial Times Limited 2019. All rights reserved. You may share using our article tools. Please don't cut articles from FT.com and redistribute by email or post to the web.

Read next:

Read next:

Lookout, there’s a dollar crunch!

FT Alpha Tweets