Libra: A Governance Perspective
If you told a Bitcoiner in 2009 that Facebook would be developing a cryptocurrency, you would have been laughed out of the room. At the Bitcoin 2019 conference in San Francisco, Libra was an inevitable topic on most panels.
Libra is a stable cryptocurrency that’s backed by a reserve of assets and governed by an independent association that’s comprised of multinational corporations, non-profit institutions, crypto-focused companies, and investors.
While Fidelity, Square, and JP Morgan have made meaningful contributions over the years, I think the breadth and size of the Libra Association marks a milestone in the legitimacy of the cryptocurrency industry. In this post, I will focus on an underexamined aspect which I believe will make or break the project — governance.
It remains to be seen whether a group of organizations will be able to successfully coordinate around building and maintaining a public good. Moreover, it is unclear whether the benefits of this public good is a strong enough incentive for organizations with conflicting priorities and complex competitive dynamics to work together, especially with the inevitable regulatory pressure to follow.
In this post, I will cover:
- Libra’s currency and governance model design
- An analysis of Libra’s governance model and incentives for Association members
- Best- and worst-case scenarios for Libra adoption
Libra’s analogs
Before diving into the governance model, it’s important to define what exactly is being governed. To understand this, it’s helpful to first think about Libra in the context of similar currency designs:
Currency Boards
Libra’s design is actually similar to that of a currency board, which is a monetary authority that maintains a fixed exchange rate with a foreign currency. A currency board issues into circulation local notes and coins that are anchored to a foreign currency, referred to as the reserve or anchor currency. The anchor currency is a strong, internationally-traded currency (e.g. USD) and the value and stability of the local currency are directly linked to that of the foreign anchor currency. The most prominent example today is Hong Kong, which operates a currency board known as the Hong Kong Monetary Authority. Historically, the U.S. also maintained a special case of a currency board where the value of the U.S. Dollar was linked to the value of gold instead of a foreign currency.
Special Drawing Rights
Libra is not a 1:1 peg to any single currency, however. Instead, it has a floating exchange rate based on an underlying basket of assets. In this sense, Libra’s design also resembles that of Special Drawing Rights (SDRs), which is an international monetary reserve currency created by the International Monetary Fund (IMF) in 1969 that operates as a supplement to the existing money reserves of member countries. It is essentially a synthetic reserve asset that’s built from a basket of fiat currencies. The SDR basket is reviewed by the IMF every five years and currently consists of the U.S. Dollar (41.73%), Euro (30.93%), Renminbi (10.92%), Japanese Yen (8.33%), and British Pound (8.09%).
Similarly, one of the most important governance decisions for Libra will be around managing the reserve because it underpins the stability of the currency and thus the utility of the network. More broadly, governance will be focused on developing the network and allocating funds to social-impact causes.
Libra’s governance model
The whitepaper introduces several components that govern the project. I will provide an overview of each component before diving into the analysis. The source for the descriptions and diagrams could be found here.
The Libra Association
The Libra Association is an independent, not-for-profit membership organization headquartered in Switzerland. It is currently comprised of 28 Founding Members and hopes to have approximately 100 members by mid-2020.
The Libra Association Council
The Association is governed by the Libra Association Council, which is comprised of one representative per validator node. Together, they make decisions on the governance of the network and reserve, with major policy or technical decisions requiring the consent of two-thirds of the votes (i.e. a supermajority).
The voting powers in the Council are proportional to stake and voting rights are capped for any single Founding Member to avoid concentration of power. Each $10 million investment entitles one vote in the Council but a single Founding Member can only be represented by the greater of one vote or 1% of the total votes in the Council. The cap does not apply to validator nodes that join the network only through holding Libra in custody.
In addition, a party entitled to be represented in the Council can delegate its voting power to another party. That said, Founding Members who hold voting power exceeding the cap will make those excess votes available to the Board for delegation to Social Impact Partners (SIPs) or research institutions. Excess votes made available to the Board that is not delegated may be distributed by the Board among other Founding Members equally to maintain the relative aggregate voting power of Founding Members compared to that of the other validator nodes.
In addition, the Council could delegate many of its executive powers to the Association’s management and Council committees but retains authority to override delegated decisions and keep key decisions to itself, with the most important ones requiring a supermajority.
The Board
The Libra Association Board is an oversight body on behalf of the Libra Association Council, providing operational guidance to the association’s executive team. The Board consists of five to 19 members and decisions require a regular majority of the board’s votes.
Governance model analysis
Libra’s approach is a combination of a traditional corporate structure and proof-of-stake-like decision model with several important differences.
In a traditional corporate structure, the shareholders own the corporation. That ownership could be 100% in the hands of one individual or spread among tens of thousands of people. Although shareholders may not participate in day-to-day management or have a direct say in decision-making, major shareholders carry great weight in influencing corporate decisions, such as voting on the election and removal of board directors. Similarly, in Libra, the shareholders are effectively the validators, who directly vote on key decisions.
Separately, the primary responsibility of the Board of Directors in a traditional corporate structure is to protect the shareholders’ investment. The Board is responsible for drafting and amending the company by-laws, appointing committees, and appointing the officers (e.g. CEO), who report to the board. This is not the case for Libra because most of these responsibilities stay with the Council.
Checks and balances
It’s clear that the Council is the most important governing body because it has the right to change the governance structure itself. While there are limited checks and balances on the Council, since the Board’s decisions could be superseded by the Council, the issue is mitigated by requiring a supermajority of votes for major changes. At 28 members, one vote gives each member ~3.6% of the voting power and will require 19 members to pass a supermajority vote. This seems reasonable and will continue to improve as the Association grows, but there would be an issue if the Council significantly shrinks — for example, a five-member council would give each member 20% of the voting power and would require four members to pass a supermajority vote.
Balanced representation becomes more complicated when members could delegate votes to other Founding Members because power could become concentrated in an even smaller group of companies. To mitigate this scenario, the Council has limited the aggregate number of votes for a given company type. Specifically, neither crypto-focused investors and companies nor social impact partners could have more than 33% aggregate voting power in the Council. While there is a cap on aggregate Council members, there is no cap on the percentage of those members within the Board. For example, with 28 Council members and five Board members, a 33% cap on crypto companies would result in a cap of nine crypto companies, all of which could comprise the Board if voted in. Even if there were just three crypto companies voted into a five-member Board, that’s sufficient for a regular majority in favor of decisions that benefit those companies.
In addition, the Social Impact Advisory board (SIAB), which is led by nonprofit organizations and academic institutions, seems to hold limited influence. While it does make funding recommendations that could affect the network, the Board has to approve those recommendations. The Board also has to approve SIPs to become nodes, which directly affects the voting power of those SIPs in the Council.
Politics and conflicts of interest
With only 28–100 Council members, politics should be expected; after all, there will be individuals from organizations who know how to play that game. I define politics here as the use of power, which ranges from individual and informal to organizational and formal, to affect decision making in a way that is self-serving for a particular individual or group.
In the Council, this could look like individuals supporting decisions which benefit their organization or coalitions forming based on common interests. Speculative scenarios include:
- Voting to remove a Board member who is from a competing organization
- Appointing a Managing Director with personal ties to an organization
- Changing the Founding Member eligibility criteria to a higher annual revenue threshold
- Changing the governance policy to lower the aggregate voting cap for a particular type of organization
I expect the Board to have even more of this issue; at 19 members, each member has ~5.3% of voting power and a maximum of 10 members will be required for votes to pass.
Related to this is are potential conflicts of interest, which could be in the form of investor/portfolio company, competing company/product, and personal relationships. For example:
- What happens if Vodafone wants to launch a Calibra competitor similar to M-Pesa?
- What happens if Calibra wants to launch in the same market as Celo, a portfolio company of Andreessen Horowitz and Coinbase?
While competing applications are expected to be built on top of the Libra network, I expect a different reaction to applications coming from Council members rather than startups like ZenGo. These potential conflicts of interest could make relationships very complicated very quickly.
Is Libra incentivized to succeed?
In the early stages of the network, when there’s greater regulatory, financial, and reputational risk for Founding Members, I think one of the most important questions to ask is:
What is the incentive for companies to participate, and will it be strong enough to stand up against regulators and governments when push comes to shove?
One could argue that the $10 million buy-ins for Founding Members is an excellent design choice because those members have real skin in the game. At the least, each member will likely have one full-time employee dedicated to maintaining their validator node, which costs an estimated $280,000 per year (side note: this effectively rules out the possibility of individuals running their own node). That money, however, is a drop in the bucket for most corporations and not in the Libra Association’s bank account just yet, since the New York Times reported that members signed nonbinding agreements for the project and seven members said that “they weren’t obliged to use or promote the digital token and could easily back out.”
The incentive issue becomes more apparent on an individual company level. Let’s take Visa for example. Visa processed 124.3 billion transactions and $8.2 trillion in payments volume in 2018. For comparison, according to Blockchain.com data, Bitcoin processed 81.4 million transactions and $397 billion in payments volume in 2018 — orders of magnitude below Visa. Why would Visa want to support a project that circumvents their network and cannibalizes their ~2% interchange fee? Let’s assume this is a hedge against the scenario where a significant portion of the world’s payments transitions onto crypto payment rails (since Visa could still earn fees for value-added services on top of the network, similar to how Coinbase charges fees for Bitcoin purchases) as well as an opportunity to capture a portion of the 85% of transactions that are still made in cash.
Let’s look at Libra’s target market — the underbanked. If Libra successfully serves this population then Visa gets a piece of a large and growing market that could last for generations. There’s an issue, however. According to Findex, over half of the 1.7 billion underbanked come from just seven countries: Bangladesh, China, India, Indonesia, Mexico, Nigeria, and Pakistan; in more than half of these places, cryptocurrencies are banned, Facebook can’t freely operate, and/or heavy regulatory restrictions exist due to money-laundering and anti-crime concerns. What will happen if the U.S. threatens to revoke Visa’s money transmitter license because their node is validating transactions in one of these countries? Will Visa remain a validator and active member? My bet is no since it won’t be worth it for Visa to put their current business at risk.
And regulators are not just sitting on the sidelines until this happens. The U.S. House Financial Services Committee recently drafted a bill, titled the “Keep Big Tech Out Of Finance Act”, which proposes a fine of $1 million per day for technology companies with over $25 billion in annual revenue functioning as financial institutions or issuing digital currencies. Facebook, one of these companies, is not planning on picking fights with regulators. In his testimony to the U.S. Senate Committee, David Marcus, Head of Calibra, wrote:
“Facebook will not offer the Libra digital currency until we have fully addressed regulatory concerns and received appropriate approvals.”
It is also not planning on picking fights with central banks. Marcus goes on to say:
“[ The Libra Association] will work with the Federal Reserve and other central banks to make sure Libra does not compete with sovereign currencies or interfere with monetary policy. Monetary policy is properly the province of central banks.”
I believe this strategy is counterintuitive to Libra’s target market. While Libra intends to create a network that operates across any country, it’s actually creating a network that will have to comply with every country’s regulatory regime. This means that Libra probably won’t be able to reach countries where currency and banking are broken if those governments believe it undermines their power. Even though the Association is comprised of different companies, the country-specific legal entities of those companies can be ordered to take actions by courts.
More generally, I believe that running a permissioned network with companies that have a significant financial stake in regulated jurisdictions is not fully incentive-compatible if the goal of the network is to freely operate within those jurisdictions.
Addressing this issue will require either Libra to limit the scope of its operations, local regulators to change their position on cryptocurrencies, and/or enabling technologies that provide sufficient KYC/AML guarantees for those regulators while minimizing transaction censorship (false positives) and maximizing privacy guarantees for individuals.
The thing about consortiums
While consortiums provide a model for collaboration and collective ownership, I believe that no blockchain consortium to date has seen meaningful success beyond announcing initial launch partners.
When I was working for a global investment bank, I had a front-row seat for the R3 Consortium. The initiative was slow to start and was largely disregarded both internally and externally. Many organizations called themselves members just to listen in on meetings for competitive intelligence. While R3 eventually gained momentum and fostered usage on its Corda platform, most activity remains in the proof-of-concept stage.
While I still believe that there exists a need for corporations with different incentives to create shared infrastructure, the consortium model as it stands today suffers from politics and slow decision making.
With Libra you have a similar scenario — many corporations, some partners and others competitors, coming together to try to create a permissioned ledger that is collectively developed, governed and utilized by its members. As such, I believe that Libra will have a slow start and will need to address these issues to gain adoption over the years.
In addition, Facebook has never had to launch a product before using a consortium. They have typically launched by moving fast and breaking things in a centralized fashion so the team might not be able to execute in a decentralized environment.
The best-case scenario for Libra
What is the happy case for Libra? If these 28 organizations really do work together to successfully launch the network, what will be the implications? I believe that they’ll be at the scale of the Dutch East India Company (or “VOC” in Dutch).
With starting capital of roughly $644 million in USD today (unconfirmed) and the Dutch government granting it a 21-year monopoly on the Dutch spice trade, the VOC was founded in 1602 and flourished for nearly 200 years. By 1669, it was the richest company the world had ever seen, with over 150 merchant ships, 40 warships, 50,000 employees, a private army of 10,000 soldiers, and a dividend payment of 40% on the original investment. The Company even produced its own currency that was stamped with the VOC logo in gold, silver, and bronze, and was used in all trade and barter.
The Company was a transcontinental employer and an early pioneer of outward foreign direct investment, with its investment projects helping raise the commercial and industrial potential of many underdeveloped regions of the world. It was also a pioneer of corporate-led globalization; during those two centuries, the VOC sent almost a million people to Asia, more than the rest of Europe combined. Moreover, it was also the first multinational company and the first company to issue public stock. These innovations allowed a single company to mobilize financial resources from a large number of investors and create ventures at a scale that had previously only been possible for monarchs.
Similarly, Libra could receive the blessing of U.S. regulators, launch the network, bootstrap usage from the Founding Members’ existing customer base, and successfully expand into customers and geographies that are excluded from basic financial services. The reach of this network would be staggering — Facebook alone has 2.3 billion monthly active users, which is more than the number of followers of Christianity, Vodafone has 468 million customers, and Visa and Mastercard combined have 8.5 billion payment cards in circulation worldwide.
Other potential outcomes
There is a wide range of scenarios for Libra. Below are the ones I believe are possible, from happy to sad cases:
- Public support: Regulators try to stop Libra but face significant backlash from the public. The Association, piggybacking on grassroots support, “lawyers up” and regulators’ limited resources can’t compete with the lobbying. Regulators cave and amend regulations (e.g. removing capital gains tax on small purchases) to support the adoption for Libra and other cryptocurrencies.
- Launches but has no usage: We re-evaluate whether people need or value a permissioned cryptocurrency. Perhaps the reserve-backed stablecoin idea stays, but we go back to permissionless and censorship-resistant cryptoassets like Bitcoin. In the process, millions of individuals learn about the notion of non-seizable, self-sovereign money.
- Dead on arrival: Huge backlash from regulators scares away current and future Founding Members, leading to fewer than the five required to launch the network. Perhaps regulators deem Libra a security and, like with Basis, the $10 million is returned to the Founding Members and everyone walks home mostly unscathed.
Conclusion
Libra has taken a thoughtful approach to its governance model and its transparency has provided a new standard for consortium-based blockchain projects. It blends old ideas with new developments and attempts to provide checks and balances on the network in a permissioned state while innovations in fully permissionless cryptonetworks are being developed. While there are issues with the consortium structure, incentive model, and regulatory strategy, the Association’s resources support the possibility of a successful network launch and long-term adoption of Libra.
That said, there are still many open questions about how Libra’s governance will actually work in practice. We’ve never had companies operate together in this fashion at this scale before and, coupled with the regulatory risks, it’s going to be an extremely complicated launch. In his Senate hearing, David Marcus stated that the Charter and By-laws for the Libra Association is currently being developed and will be made public. This will provide additional detail on the governance model and I look forward to reading it.
Many thanks to Ash Egan, Devin Walsh, Spencer Noon, and Craig Burel for their feedback on this piece.
You could follow me on Twitter here.
