Distributed Ledgers: Design and Regulation of Financial Infrastructure and Payment Systems

Distributed Ledgers

Design and Regulation of Financial Infrastructure and Payment Systems

Robert M. Townsend

The MIT Press

Cambridge, Massachusetts

London, England

© 2020 Massachusetts Institute of Technology

This work is subject to a Creative Commons CC-BY-NC-ND license.

Subject to such license, all rights are reserved.

The open access edition of this book was made possible by generous funding from Arcadia—a charitable fund of Lisbet Rausing and Peter Baldwin.

Library of Congress Cataloging-in-Publication Data

Names: Townsend, Robert M., 1948- author.

Title: Distributed ledgers : design and regulation of financial infrastructure and payment systems / Robert M. Townsend.

Description: Cambridge : The MIT Press, 2020. | Includes bibliographical references and index.

Identifiers: LCCN 2020004676 | ISBN 9780262539876 (paperback)

Subjects: LCSH: Financial services industry--Technological innovations.

Classification: LCC HG173 .T69 2020 | DDC 332/.042402855758--dc23

LC record available at https://lccn.loc.gov/2020004676

d_r0

Contents

Preface

Acknowledgments

1 Introduction

2 Economies, Obstacles, Welfare, and Measurement

3 Ledgers as Financial Accounts

4 E-Payments, E-Messages, and Trusted Third Parties in Payment Systems

5 Encryption

6 Smart Contracts: Contract Theory and Mechanism Design

7 Design Issues: Partitioned Ledgers, the Decision to Decentralize Implementation in Multiparty Contracts, and Incentive-Compatible Token Payment Systems

8 Building Financial Infrastructure on Distributed Ledgers: Practical Application in Emerging Markets

9 Payment Systems on Distributed Ledgers: Practical Applications

10 Regulation and the Use of Distributed Ledger Technology

11 Cryptocurrency: The Role and Value of Tokens in Economies with Distributed Ledger Systems

12 Summary and Conclusion

References

Index

List of Figures

Figure 2.1

Schemata of financial accounts and agent interaction through smart contract. Agents A and B with financial accounts are sending messages from an a priori message space; these messages are input into a designed smart contract and execute a transfer that alters the financial accounts. Source: Nicolas Zhang (2019).

Figure 2.2

Illustrative time line of events within a given period t under a contract. The agent observes its current underlying state, et, then takes an action a1t, determining publicly observed state yt, which in turn activates transfer τt. Then, toward the end of the period, the agent takes another action, a2t, resulting in utility as a function of these actions and events U(et, a1t, yt, τt, a2t). At date t+ 1, the state et+1 is realized (determined in part by previous actions) and the time line of t repeats at t+ 1, and so on.

Figure 4.1

Illustrative movement of currency balances for a selected household. Erratic but increasing levels of currency over time, with sharp drop at the end.

Figure 4.2

Swedish currency (SEK) in levels and relative to GDP.

Figure 4.3

M-Pesa adoption rates for the entire Kenyan population as well as for the poor, the lowest income quartile, and those with no bank account.

Figure 4.4

A schemata of the operational flows of the Kenyan M-Pesa system. An exchange of Kenya shillings, marked as $, for cell phone credits, marked as e, with a Safaricom agent; use of this e-money by the customer for purchases or remittances; and a larger picture of the flows of e-money and the cash throughout the system.

Figure 5.1a

Mesopotamia tokens.

Figure 5.1b

Figure 5.2

Medieval English tally sticks.

Figure 5.3

A decision to burn the obsolete tally sticks in 1834 nearly destroyed the Palace of Westminster.

Figure 5.4

Mesh communication networks and star communication networks. The figure on the left shows how nodes N are connected to each other. The figure on the right displays a classic star network with a central node.

Figure 11.1

The Turnpike model of monetary exchange. The arrows denote the direction of travel. Each agent type starts at some trading post marked with a vertical line connecting the two lanes of the highway. The numbers denote the endowment of the consumption good of the agent as the agent travels.

List of Tables

Table 3.1

Comprehensive financial accounts: balance sheet of household A.

Table 3.2

Comprehensive financial accounts: income statement of household A.

Table 3.3

Comprehensive financial accounts: statement of cash flow of household A.

Table 4.1

Money in terms of monthly consumption.

Table 7.1

Partitioned private ledgers and the gains from concealment.

Table 7.2

Colored coins and partitioned private ledgers.

Table 10.1

Circulating private debt: who meets whom when.

List of Boxes

Box 6.1

Key elements from mechanism design.

Box 6.2

Dynamic principal and agent problem with moral hazard and promised utility.

Box 6.3

Limited commitment.

Preface

Distributed ledgers have the potential to transform economic organization and financial structure. Yet the subject is embroiled in controversy, hype, and terminological inconsistencies. Rather than get waylaid by alternative possible definitions of distributed ledgers (also known as decentralized ledgers), we focus more broadly on an economic analysis of what distributed ledgers can do. We proceed by analyzing key individual components. We also compare and contrast the economic framework with the frameworks of computer science and data management disciplines to clarify the technology and take steps to combine these disciplines.

The familiar but key components of distributed ledgers discussed in this book are ledgers as financial accounts, e-messages and e-value transfers, cryptography, and contracts including multiparty mechanisms. Each component is evaluated and illustrated through the context of historical and contemporary economies, with featured applications in both developed economies and emerging-market countries. These use cases are a hallmark of the book. A recurrent focus is the general equilibrium impact of innovations and welfare gains from innovations featuring key components. This does not require that all components be introduced at the same time.

Contract theory is used to derive optimal arrangements, constrained only by obstacles to trade, featuring how the various aspects of ledgers can deepen infrastructure. Mechanism design and monetary theory are used to study public versus partitioned ledgers and improvements in payment systems. Prudential regulation, rather than being a barrier to innovation, can be improved with the use of distributed ledger technologies.

The goal is to provide blueprints for the ex ante optimal design and regulation of financial systems, including not only choices at the end points of the spectrum—of centralized versus decentralized systems, as in the hype—but the choice of hybrid forms in between. Each key component is assessed from both computer science and economic perspectives, and syntheses are offered. Overall, the book provides a vision for where we are heading, being clear about obstacles along the way.

Acknowledgments

I gratefully acknowledge research support from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD), grant number R01 HD027638; the Centre for Economic Policy Research (CEPR) and the Department for International Development (DFID) under grant MRG002_1255; and funding assistance for the continuation of the field surveys by the Thailand Research Fund, the Bank of Thailand, and the University of Thai Chamber of Commerce. I also am grateful for the collaborations with the Federal Reserve Banks of New York, Boston, and Chicago; Lightnet for its decentralized settlement banking network; and EvryNet for the decentralized custodian banking network in Thailand. The views expressed are my own. Thank you to Ernst-Ludwig von Thadden, James McAndrews, discussants of this work, and other participants at the Sveriges Riksbank Annual Macroprudential Conference in June 2017; participants at the Kellogg Conference on Development Economics at Northwestern University in September 2018; participants at the Cryptocurrencies and Blockchains Conference at the Becker Friedman Institute in November 2018; participants at the First New York Fed Conference on FinTech, March 2019; participants at the Penn State Conference in Celebration of Neil Wallace’s Contribution to Economics, April 2019; Daniel Aronoff, Neha Narula, Neil Wallace, Joshua Gans, and Jesus Fernandez-Villaverde for their very helpful comments; and Deborah Jamiol, Jennifer Roche, and Emily Gallagher for wonderful editing. Thank you also for comments on preliminary drafts to Marios Angeletos, Pablo Aznar, Zach Chao, Co-Pierre Georg, Michael Lee, Rhys Lindmark, Jacky Mallett, Antoine Martin, Nish Patel, and Nicolas Zhang. The reviewers for MIT Press made substantive contributions.

1

Introduction

Distributed ledger technology (DLT) or, better put, its various features in isolation and in combination, has the potential to be transformative. Nevertheless, this subject has engendered controversy and sharp debate as well as a lack of clarity in the terminology researchers use to discuss it.

The first part of this introduction outlines the debate, and the second part outlines the point of view of this book.

Of note, Bitcoin is thought of as having created the blockchain as part of its validation system, so some people consider Bitcoin, blockchain, and distributed ledger technology to be synonymous. They are not. Some distributed ledger technologies exist without the blockchain technology and without coins. Indeed, much of the technology for distributed ledgers existed before Bitcoin and blockchain.

So this book proceeds in reverse. It starts with distributed ledgers, works backward to blockchain, and defers a more in-depth discussion of cryptocurrencies to the end. Concepts, definitions, applications, and impact are discussed at each turn. The term distributed ledger is sometimes used synonymously (if incorrectly) with the term decentralization. Computer science and data science are needed to clarify the distinction, and we compare and contrast with the meaning of decentralization in economics.

1.1 General Motivation: A View from All Sides

We begin with selected quotes from policymakers and academics (not practitioners or fintechs) in support of the premise that technology is fundamental.

DLT refers to the processes and related technologies that enable nodes in a network (or arrangement) to securely propose, validate, and record state changes (or updates) to a synchronized ledger that is distributed across the network’s nodes. In the context of payment, clearing, and settlement, DLT enables entities, through the use of established procedures and protocols, to carry out transactions without necessarily relying on a central authority to maintain a single golden copy of the ledger.

DLT may radically change how assets are maintained and stored, obligations are discharged, contracts are enforced, and risks are managed. Proponents of the technology highlight its ability to transform financial services and markets by: (i) reducing complexity; (ii) improving end-to-end processing speed and thus availability of assets and funds; (iii) decreasing the need for reconciliation across multiple record-keeping infrastructures; (iv) increasing transparency and immutability in transaction record keeping; (v) improving network resilience through distributed data management; and (vi) reducing operational and financial risks [Mills 2016]. DLT may also enhance market transparency if information contained on the ledger is shared broadly with participants, authorities and other stakeholders.

—Bank for International Settlements (BIS 2017a, 2, 1)

Contracts, transactions, and the records of them are among the defining structures in our economic, legal, and political systems. They protect assets and set organizational boundaries. They establish and verify identities and chronicle events. They govern interactions among nations, organizations, communities, and individuals. They guide managerial and social action. And yet these critical tools and the bureaucracies formed to manage them have not kept up with the economy’s digital transformation. They’re like a rush-hour gridlock trapping a Formula 1 race car. … With blockchain, we can imagine a world in which contracts are embedded in digital code and stored in transparent, shared databases, where they are protected from deletion, tampering, and revision. In this world every agreement, every process, every task, and every payment would have a digital record and signature that could be identified, validated, stored, and shared.

—Marco Iansiti and Karim R. Lakhani, Harvard Business Review, 2017 (119–120)

But, of course, there are concerns and qualifications. We note some of these immediately, from the same sources. The Bank for International Settlements (BIS 2017a) lists risks associated with using DLT for payments that include potential uncertainty about operational and security issues arising from the technology; the lack of interoperability with existing processes and infrastructures; ambiguity relating to settlement finality; questions regarding the soundness of the legal underpinning for DLT implementations; absence of an effective and robust governance framework; and issues related to data integrity, immutability, and privacy. The Committee on Payments and Market Infrastructures (CPMI) chair, Benoît Cœuré, writes,

Central banks have traditionally played an important catalyst role in payments and settlements. This report will help central banks, other authorities, and the public to identify the risks as well as the benefits associated with the emerging technology, which could be the basis for next-generation systems (BIS 2017b).¹

Iansiti and Lakhani (2017) focus on the difficulty of adoption of transformative technologies. They distinguish between novelty and complexity, laying out various historical examples of innovation and current, ongoing experiments in DLT, in the end dividing innovations into four categories along the lines of high/low novelty and high/low complexity. This allows them to make predictions about not only where innovations are likely to succeed first, but also to identify those that could take considerable time, possibly decades, if they happen at all.

An implicit point: There is a distinction between invention of something new and its actual innovation and implementation. Lags in adoption are a murky criterion to use in the evaluation of the merits of inventions. This conflates the search for something new in DLT. Relatedly, innovation depends on context. In some settings, innovation is on the margin with much of the technology already in place. This may make the value of innovation marginal, potentially not worth the cost. But even if the gains from innovation could be incrementally large, vested interests with legacy systems can block change. In contrast, innovation can happen in settings where there is little if anything already in place on the ground, in which case implementation of key components singly, or in combination, can make a huge difference, even for innovations that are mundane and already adopted in other contexts.

An example of a low novelty–low complexity innovation given by Iansiti and Lakhani (2017) is Bitcoin. Their argument is that Bitcoin is another object like money for the transfer of value—hence, nothing novel. This, however, belies both Bitcoin’s creative algorithm and the controversy around it. To some, Bitcoin is singularly innovative. This has a lot to do with differences between computer scientists’ and economists’ perspectives, which we seek to clarify in this book. For others, Bitcoin is extremely problematic (we will return to this debate shortly). In any event, the bulk of innovations and experiments in new technologies occur under what Iansiti and Lakhani refer to as localization. That is, they introduce highly innovative uses and products but with a limited number of users. The list of these types of localized technologies is growing in length, some moving beyond commitments to experimentation and actual implementation.

One DLT use case that immediately reveals what DLT can do involves land-title projects such as those in Georgia, Sweden, and the Ukraine (Reese 2017). To buy property in these locations, the lawful owner must have a secure title to sign over to the purchaser. DLT uses hashes to record every real-estate transaction and make them immutable, publicly available, and searchable so that titles can be transferred quickly, without costly title searches. Propy.com is an example of a proprietary company innovating in this space, with a distributed ledger in active use. The same idea underlies the emergence of digital assets to facilitate ownership and transfer.

In practice, in many markets, there are gaps and pauses in transaction time lines even for the most obvious transactions. A key example: In financial markets, trade, clearing, and settlement are separated in time. An agreement to trade between two parties can happen quickly, but it is then recorded into the private and proprietary legacy systems of each party, hence requiring reconciliation later. Trades in equity on a central stock exchange can take two days or more to settle, and, in part, this is not a matter of choice as there is no immutable synchronized record on which all parties can rely. Digital Asset is a company that has entered into an agreement with the Australian stock exchange to allow trade and confirmation in equities in real time, which was scheduled to be in operation by 2020.

TReDS, in India, is a platform for the discounting and sale of trade receivables. There are two other competing platforms currently operating in India. Since 2017, these three platforms have been operating a common distributed ledger for the recording of submitted buyer-seller receivable transactions. Each transaction has a unique ID number, so there can be no duplicates and thus no fraud. The Hong Kong Monetary Authority is also implementing a DLT system to avoid double invoicing.

Stellar is a not-for-profit entity that Iansiti and Lakhani would place in the category of an innovation with low novelty and high coordination needs. Stellar focuses on banking, micropayments, and remittances for people without access to the formal financial sector or those who have access but at a high cost. Stellar has been operating since late 2014 and has a current market valuation, at the time of writing, of $2 billion. Ripple is a for-profit entity with an even larger valuation, $13.5 billion, founded in 2012. Stellar emerged from Ripple.

There is also innovation in nonfinancial markets. In 2017, Maersk and IBM implemented a distributed ledger technology for freight shipping, both for tracking and for improved logistics, sharing information and documentation among connecting nodes: port and terminal operators, customs authorities, customs brokers, transportation companies, and cargo owners. They project a substantial reduction in shipping costs.² Walmart has partnered with IBM to develop a system to track the supply chain of leafy vegetables from farms to stores so that in case of contamination, Walmart can quickly pinpoint