40,99 €
Mehr erfahren.
- Herausgeber: John Wiley & Sons
- Kategorie: Fachliteratur
- Serie: Wiley Finance Series
- Sprache: Englisch
Discover Bitcoin, the cryptocurrency that has the finance world buzzing
Bitcoin is arguably one of the biggest developments in finance since the advent of fiat currency. With Understanding Bitcoin, expert author Pedro Franco provides finance professionals with a complete technical guide and resource to the cryptography, engineering and economic development of Bitcoin and other cryptocurrencies. This comprehensive, yet accessible work fully explores the supporting economic realities and technological advances of Bitcoin, and presents positive and negative arguments from various economic schools regarding its continued viability.
This authoritative text provides a step-by-step description of how Bitcoin works, starting with public key cryptography and moving on to explain transaction processing, the blockchain and mining technologies. This vital resource reviews Bitcoin from the broader perspective of digital currencies and explores historical attempts at cryptographic currencies. Bitcoin is, after all, not just a digital currency; it's a modern approach to the secure transfer of value using cryptography. This book is a detailed guide to what it is, how it works, and how it just may jumpstart a change in the way digital value changes hands.
- Understand how Bitcoin works, and the technology behind it
- Delve into the economics of Bitcoin, and its impact on the financial industry
- Discover alt-coins and other available cryptocurrencies
- Explore the ideas behind Bitcoin 2.0 technologies
- Learn transaction protocols, micropayment channels, atomic cross-chain trading, and more
Bitcoin challenges the basic assumption under which the current financial system rests: that currencies are issued by central governments, and their supply is managed by central banks. To fully understand this revolutionary technology, Understanding Bitcoin is a uniquely complete, reader-friendly guide.
Sie lesen das E-Book in den Legimi-Apps auf:
Seitenzahl: 567
Veröffentlichungsjahr: 2014
Ähnliche
Contents
Cover
Title page
Copyright
Dedication
About the Author
Acknowledgments
Foreword
Prologue
Preface
Part One: Introduction and Economics
Chapter 1: Foundations
1.1 Decentralized
1.2 Open Source
1.3 Public Asset Ledger
1.4 It’s not Only the Currency, It’s the Technology
Chapter 2: Technology (Introduction)
2.1 Centralized Database
2.2 Addresses, Transactions
2.3 Distributed Database, the Blockchain
2.4 Wallets
2.5 The Different Meanings of Bitcoin
Chapter 3: Economics
3.1 Medium of Exchange
3.2 Store of Value
3.3 Unit of Account
3.4 Deflation
3.5 Volatility
3.6 Effect on the Financial Industry and Monetary Policy
3.7 Regulation
Chapter 4: Business Applications
4.1 Money Transfer
4.2 Exchanges
4.3 Payment Processors
4.4 Web Wallets
4.5 Multisignature Escrow Services
4.6 Mining
4.7 ATMS
Part Two: Bitcoin Technology
Chapter 5: Public Key Cryptography
5.1 Public Key Encryption
5.2 Digital Signatures
5.3 RSA
5.4 Elliptic Curve Cryptography
5.5 Other Cryptographic Primitives
5.6 Bitcoin Addresses
Chapter 6: Transactions
6.1 Transaction Scripts
6.2 Pay-to-Address and Pay-to-Public-Key Transactions
6.3 Multisignature (m-of-n) transactions
6.4 Other Transaction Types
6.5 Transaction Signature
6.6 Pay-to-Script-Hash (P2SH)
6.7 Standard Transactions
Chapter 7: The Blockchain
7.1 Hash Functions
7.2 Time-Stamp
7.3 Proof-of-Work
7.4 The Blockchain
7.5 Double-Spend and Other Attacks
7.6 Merkle Trees
7.7 Scalability
Chapter 8: Wallets
8.1 Symmetric-Key Cryptography
8.2 Offline Wallets
8.3 Web Wallets
8.4 Brain Wallets
8.6 Multisignature Wallets
8.7 Vanity Addresses
8.8 Simplified Payment Verification (SPV)
8.9 The “Payment Protocol” (BIP 70)
Chapter 9: Mining
9.1 Mining Technology
9.2 Pooled Mining
9.3 Transaction Fees
9.4 Selfish Mining
Part Three: The Cryptocurrencies Landscape
Chapter 10: The Origins Of Bitcoin
10.1 David Chaum’s Ecash
10.2 Adam Back’s Hashcash
10.3 Nick Szabo’s Bit Gold and Wei Dai’s B-Money
10.4 Sander and Ta-Shma’s Auditable, Anonymous Electronic Cash
10.5 Hal Finney’s RPOW
10.6 Satoshi Nakamoto
Chapter 11: Alt(ernative) Coins
11.1 Litecoin
11.2 Peercoin
11.3 Namecoin
11.4 Auroracoin
11.5 Primecoin
11.6 Dogecoin
11.7 Freicoin
11.8 Other Alt-Coins
11.9 The Case For/Against Alt-Coins
Chapter 12: Contracts (the Internet of Money or Cryptocurrencies 2.0)
12.1 Digital Assets
12.2 Smart Property
12.3 Micropayments
12.4 Autonomous Agents
12.5 Other Applications
12.6 Inserting Data into the Blockchain
12.7 Meta-Coins
Chapter 13: The Privacy Battle
13.1 Network Analysis
13.2 Laundry Services
13.3 Greenlisting
13.4 Privacy-Enhancing Technologies
13.5 Fully Anonymous Decentralized Currencies
Chapter 14: Odds and Ends
14.1 Other Transaction Protocols
14.2 Alternatives to Proof-of-Work
14.3 Merged Mining
14.4 Side-Chains
14.5 Open Transactions
14.6 Quantum Computing
14.7 Recent Advances in Cryptography
Bibliography
Index
List of Tables
Table 3.1: Bitcoin wealth concentration. Data from Wile (2013)
Table 4.1: Daily transaction volume. Data from Grossman et al. (2014)
Table 6.1: Types of transaction. Data from webbtc.com on June 17, 2014
Table 9.1: Distribution of hash rate among mining pools. Data retrieved from blockchain.info/pools on June 17, 2014. Average of 4 days
Table 11.1: Market capitalization of some alt-coins. Data retrieved from coinmarketcap.com on April 19, 2014
List of Illustrations
Figure 1.1: What Bitcoin is (and isn’t)
Figure 2.1: Double-spending problem
Figure 2.2: Central counterparty holding a centralized database
Figure 2.3: Central counterparty single point of failure
Figure 2.4: Analogy between BitTorrent and Bitcoin
Figure 2.5: User sending funds. State of the database after the transaction has settled
Figure 2.6: Bitcoin as a distributed ledger
Figure 2.7: Bitcoin issuance theoretical schedule
Figure 3.1: Transaction per day (with 14-day moving average). Data from blockchain.info
Figure 3.2: Estimated daily transaction volume in bitcoins. Data from blockchain.info
Figure 3.3: Estimated daily transaction volume in USD. Data from blockchain.info
Figure 3.4: Miners revenue/transaction value, including 14-day moving average. Data from blockchain.info
Figure 3.5: BTC/USD. Data from blockchain.info
Figure 3.6: Annual Bitcoin velocity, using data from blockchain.info
Figure 3.7: Cumulative coin-days destroyed. Data from blockchain.info
Figure 3.9: Bitcoin annual percentage price volatility using a 6-month rolling window. Using bitcoin price data from blockchain.info
Figure 3.10: Money in circulation versus Bitcoin market cap, from Normand (2014), all figures converted to USD
Figure 3.11: Quantity Theory of Money and Bitcoin’s possible impact
Figure 4.1: Money transmitters market capitalization versus Bitcoin. Data retrieved from Google Finance, Yahoo Finance and blockchain.info on May 10th, 2014
Figure 4.2: BTC/USD traded volume. Data from blockchain.info
Figure 4.3: Exchanges’ market share for BTC/USD volume. Data from bitcoinaverage.com
Figure 4.4: Payment processor
Figure 4.5: Cashflows from mining ASIC
Figure 5.1: Public key encryption
Figure 5.2: Man-in-the-Middle attack
Figure 5.3: Digital signatures
Figure 5.4: RSA encryption scheme
Figure 5.5: RSA signature scheme
Figure 5.6: Example of a 2048-bit RSA key. All numbers in hexadecimal
Figure 5.7: Key size of several public key cryptography algorithms to obtain various security levels
Figure 5.9: Elliptic curve point doubling
Figure 5.10: Elliptic Curve identity element
Figure 5.11: Elliptic curve math
Figure 5.12: An example of a toy elliptic curve with 14 steps (left) and with the full cycle 270 steps (right)
Figure 5.13: Parameters in secp256k1 vs toy example. All numbers in hexadecimal
Figure 5.14: Elliptic curve digital signature algorithm
Figure 5.15: Address generation
Figure 6.1: Transaction
Figure 6.2: Unspent Transaction Outputs Cache (UTXO)
Figure 6.3: Scripting
Figure 6.4: Pay to Bitcoin address transaction
Figure 6.5: Stack for a pay-to-address script
Figure 6.6: Elements of a transaction
Figure 6.7: Pay-to-script-hash transaction
Figure 7.1: Hash functions
Figure 7.2: Merkle-Damgård construction
Figure 7.3: Time-stamping a group of transactions by publishing their hash in a newspaper
Figure 7.4: Linked time-stamping
Figure 7.5: Partial hash inversion proof-of-work
Figure 7.6: Hashcash
Figure 7.7: The blockchain
Figure 7.8: Dynamics of the blockchain
Figure 7.9: Security of a transaction inside the blockchain
Figure 7.10: Investment required to pull off a 51% attack with 1% and 50% probability of success
Figure 7.11: Merkle tree of the transactions in a block
Figure 7.12: Pruning of transactions in a block
Figure 7.13: Blockchain size (MB) on a logarithmic scale. Data from blockchain.info
Figure 8.1: Encryption of wallet private keys
Figure 8.2: Paper wallet generated using bitaddress.org
Figure 8.3: Trezor
™
hardware wallet. Picture by SatoshiLabs (www.bitcointrezor.com)
Figure 8.4: Type-1 deterministic wallet
Figure 8.5: Hierarchical Deterministic Wallet (BIP 32)
Figure 8.6: Tree created by an HD wallet
Figure 8.7: Vanitygen address generator
Figure 8.8: Vanitygen address generator with a shorter prefix
Figure 8.9: Vanitygen address generator with a string close to the size of an address
Figure 9.1: Bitcoins in circulation. Data from info">blockchain.info
Figure 9.2: Hash rate of the Bitcoin network. Hash rate data from blockchain.info
Figure 9.3: Mining revenue compared to electricity cost of different technologies. Price and hash rate data from blockchain.info
Figure 9.4: Probability of mining a block (network-wise)
Figure 9.5: Q-Q plot (exponential distribution) of the empirical time between blocks
Figure 9.6: Expected time to mine a block
Figure 9.7: Prioritization of transactions by miners
Figure 10.1: Satoshi Nakamoto
Figure 11.1: Network effect
Figure 12.1: Digital assets
Figure 12.2: Smart property
Figure 12.3: Market capitalization of some meta-coins. Data from coinmarketcap.com on April 19, 2014
Figure 12.4: A Colored Coin genesis transaction
Figure 12.5: Ripple protocol
Figure 13.1: Transaction graph
Figure 13.2: The user graph corresponding to the transaction graph of Figure 13.1
Figure 13.3: Laundry service
Figure 13.4: Greenlisting
Figure 13.6: CoinSwap
Figure 13.7: Elliptic Curve Diffie–Hellman key exchange
Figure 13.8: Stealth addresses
Figure 13.9: A 3-colorable graph
Figure 14.1: Merged mining
Guide
Cover
Table of Contents
About the Author
Pages
ii
iii
iv
xi
xii
xiii
xiv
xv
xvi
xvii
xviii
xix
xx
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
Understanding Bitcoin
Cryptography, engineering, and economics
PEDRO FRANCO
This edition first published 2015©2015 Pedro Franco
Registered officeJohn Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, United Kingdom
For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher.
Wiley publishes in a variety of print and electronic formats and by print-on-demand. Some material included with standard print versions of this book may not be included in e-books or in print-on-demand. If this book refers to media such as a CD or DVD that is not included in the version you purchased, you may download this material at http://booksupport.wiley.com. For more information about Wiley products, visit www.wiley.com.
Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. Neither the publisher nor the author are associated with any product or vendor mentioned in this book. The material contained in this book is not related to any work the author has performed for any present or past employer. Opinions expressed in the book are solely those of the author and do not express the views of the author’s current or past employers.
Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. It is sold on the understanding that neither the publisher nor the author are engaged in rendering professional services and neither the publisher nor the author shall be liable for damages arising herefrom. If professional advice or other expert assistance is required, the services of a competent professional should be sought.
A catalog record for this book is available from the Library of Congress.
A catalogue record for this book is available from the British Library.
ISBN 9781119019169 (hardback/paperback)ISBN 9781119019145 (ebk)ISBN 9781119019152 (ebk)ISBN 9781119019138 (ebk)
Dedicated to Alvaro, Rafael, Luis, and Nayra
About the Author
Pedro Franco was born in Astorga, Leon (Spain). He holds a MSc in Electrical Engineering from ICAI, a BSc in Economics, and an MBA from INSEAD. Pedro has been a consultant with McKinsey and Boston Consulting Group and a researcher with IIT prior to gaining more than 10 years of experience in financial markets holding Quant and Trading positions in Credit, Counterparty Risk, Inflation, and Interest Rates. He has created various mathematical libraries for financial derivatives, and managed teams of software developers.
The author can be contacted at [email protected].
Acknowledgments
Thanks to Juan Ramirez for helping me gather the courage to write this book.
Thanks to Jon Beracoechea, Manuel Castro, and Robert Smith for exhaustively reviewing an early version of the book and providing many excellent suggestions. Thanks also to Eli Ben-Sasson, Alejandro and Alvaro Franco, Jeff Lim, Jan Pelzl, Stefan Thomas, Evan Schwartz, Rodrigo Serrano, Alena Vranova, and Bob Way for reviewing parts of the book and providing insightful comments.
Finally, thanks to my family for their patience and support; without them this book could not have been written.
Foreword
I have been hoping for some time for a good book covering the technology and ideas behind Bitcoin to be written.
There is certainly a wealth of information about cryptocurrencies, but the field advances rapidly and it is sometimes difficult for the non-initiated to understand the fundamentals and catch up with new developments. This book takes readers to a thorough understanding of the current state-of-the-art cryptocurrencies’ technology, as well as its future economic and technological implications, without assuming any previous knowledge of the many fields than constitute Bitcoin. This is an enabling book that empowers the reader to participate in and contribute to this great adventure.
The book clearly exposes many concepts previously mainly known to insiders of the cryptocurrencies’ world. It covers a wide range of topics, from the economics or the basic technology (such as elliptic curve cryptography, Merkle trees or the blockchain) to advanced cryptographic concepts (such as non-interactive zero-knowledge proofs), and explores many applications based on these ideas (such as multi-signature wallets or fully anonymous payment systems). All this is accomplished in a book that is very approachable and comprehensible.
Readers new to Bitcoin will surely be surprised by the ingenuity of the technology and the broad range of applications it enables. Those familiar with Bitcoin will find many sections, such as the sections on economics or advanced applications of cryptocurrencies, informative and thought provoking.
I believe Pedro’s book will be well received in the business and financial community as well as by the general public, spreading the knowledge about Bitcoin and contributing to this technology crossing the chasm to the early majority.
Jeff GarzikBitcoin Core Developer at BitPay, Inc.
Prologue
Preface
Opinions about Bitcoin are highly polarized between enthusiasts and skeptics. The author believes that the point of view of the skeptics is easier to grasp for someone not familiar with Bitcoin’s technology. The objective of this book is to present the technology and arguments from both sides of the divide so that readers can form an informed opinion of their own.
What drives the passion of the enthusiasts is that Bitcoin is a technological breakthrough that creates many new and interesting applications. As is often the case with brand new technologies, many future applications of the technology might not be envisioned today. Who could have imagined the success of video streaming services or social networks in 1994? Enthusiasts feel the technology will yield many unforeseen applications for many years to come. The fact that most of these applications are intertwined with monetary economics makes it even more interesting.
The economic and technical aspects of Bitcoin are so intertwined that, in the opinion of this author, they should be tackled together. Arguing about one of them without understanding the other would be like trying to run a car with only one pedal: just pressing the gas or the brake pedal. Sure, the driver could descend a mountain with only the brake pedal, but then she could not go much further. Similarly a driver with only the gas pedal could probably ascend a mountain, but she would be better off not trying to descend it. This book covers the technology behind Bitcoin, ranging from cryptography to software engineering to monetary economics.
References to Bitcoin’s source code are scattered throughout the text, especially in the technical sections. These references are intended as clues for readers interested in the implementation of the Bitcoin protocol, but can be safely skipped by other readers.
This book is divided into three parts. The first part serves as an introduction to Bitcoin’s technology and philosophy (Chapters 1 and 2). This part will also cover the economic arguments both in favor of and against Bitcoin (Chapter 3) and some business applications (Chapter 4). This part is designed for the time-constrained readers who are mostly interested in the business and economic impact of Bitcoin’s technology.
The second part covers in detail how Bitcoin works, starting with public key cryptography (Chapter 5), transactions (Chapter 6) and the blockchain (Chapter 7). The last two chapters expand on related topics: wallets (Chapter 8) and mining (Chapter 9). In this line, two additional great resources for developers are the Developer Guide (Bitcoin Foundation, 2014a) and the Reference Guide (Bitcoin Foundation, 2014b) maintained by the Bitcoin Foundation, and the forthcoming book by Andreas Antonopoulos (Antonopoulos, 2014).
The third part completes the cryptocurrencies landscape. First, digital currency technologies preceding Bitcoin are discussed (Chapter 10). Then alternative cryptocurrencies based on Bitcoin (alt-coins) are covered (Chapter 11) and new applications of cryptocurrencies beyond payment systems are explored (Chapter 12). Most of the action in the cryptocurrencies community is focused on these new applications and Chapter 12 will introduce several of the brand new projects that are being built. Bitcoin is not anonymous, and Chapter 13 explores techniques that can be used to de-anonymize users, as well as technologies that are being built to enable users to counter these techniques and enhance their privacy. The chapter concludes with an introduction to the technology, based on zero-knowledge proofs, to create fully anonymous decentralized digital currencies. The book concludes (Chapter 14) with a discussion of some additional technical topics and the latest developments being discussed in the community.
An earlier version of this book has been registered in the blockchain. The hash of this earlier version is
1324585ce12bdf2c16995835e1ba1a04246592e7755c6c1933419fe80f97f10eand was registered in the blockchain in transaction
e144275426185d0a0b85e7bdcfdfbbaa6f7f750a522007aeaae6f0f8708838bb.
The blog for this book can be found at understandingbitcoin.blogspot.com.
Madrid, July 2014
PART OneIntroduction and Economics
CHAPTER 1Foundations
There has been ample media coverage of Bitcoin, and many public figures have been compelled to state their opinion. As Bitcoin is a complex topic, covering cryptography, software engineering and economics, it is difficult to grasp its essence and implications with only a superficial look at it. Thus some commentators might not have a clear picture of how it works and the implications. It is the goal of this book to equip the reader with the knowledge to evaluate the merits of this technology.
Figure 1.1 summarizes some misconceptions around Bitcoin.
FIGURE 1.1 What Bitcoin is (and isn’t)
Bitcoin is a decentralized digital currency. This means there is no person or institution behind it, either backing it or controlling it. Neither is it backed by physical goods, such as precious metals. This might seem counter-intuitive at first glance: how could it exist if no one controls it? Who created it then? How did the creator lose control over it?
The answer to this seeming paradox is that Bitcoin is just a computer program. How exactly this computer program works is the subject of the second part of this book. The program has a creator (or creators) but his identity is unknown as he released the Bitcoin software using what is believed to be a pseudonym: Satoshi Nakamoto. Bitcoin is not controlled in a tight sense by anyone. The creator did not lose control of it because he (she?, they?) never owned the code. The code is open source and thus it belongs to the public domain, as will be further explained in section 1.2.
One of the most innovative features of Bitcoin is that it is decentralized. There is no central server where Bitcoin is running. Bitcoin operates through a peer-to-peer network of connected computers. Bitcoin is the first digital currency built in a decentralized way, a technological breakthrough. The decentralized nature of Bitcoin will be further explored in section 1.1.
Bitcoin creates its own currency called bitcoin, with a small b. The creation of a currency is integral to how the system operates, as it serves two simultaneous purposes. First, it serves to represent value. Second, issuance of new bitcoins is used to reward operators in the network for securing the distributed ledger. These two functions cannot be unbundled without significantly changing the design.
The heart of the Bitcoin network is a database holding the transactions that have occurred in the past as well as the current holders of the funds. This database is sometimes called a ledger, because it holds the entries representing the owners of the funds. Bitcoin is not the first distributed database to be created. However, the requirements of a financial database are different from those of other applications, such as file sharing or messaging systems. In particular, financial databases must be resilient against users trying to double-spend their funds, which Bitcoin solves elegantly. This is explored in the following sections and in Chapter 2.
Some critics have argued that Bitcoin is a Ponzi scheme. It is not. In a Ponzi scheme there is a central operator who pays returns to current investors from new capital inflows. First of all, in Bitcoin there is no central operator who can profit from the relocation of funds. Second, there is no mechanism to deflect funds from new investments to pay returns. The only funds recognized in the Bitcoin protocol are bitcoins, the currency. Transfers of bitcoins are initiated by the users at their will: the protocol cannot deflect funds from one user to another. Third, a new investment in Bitcoin is always matched with a disinvestment. Investors who put money into bitcoins usually operate through an exchange where they buy the bitcoins from another investor who is selling her investment. There is simply no new investment flowing into bitcoins: the amount of sovereign currency that has flown into bitcoins exactly matches the amount that has flown out of bitcoins.
However, bitcoin, the currency, can be a bubble. Whether the value of bitcoin crashes, holds, or increases depends on whether bitcoins will be used in the future for different applications. There are several interesting applications for Bitcoin, of which the most straightforward (but not the only) are to serve as a medium of exchange and a store of value. It is too early to tell whether any of these applications will become important in the future. The merits of bitcoins as medium of exchange and store of value are explored in Chapter 3.
Finally, Bitcoin is not just a currency but a whole infrastructure that can be used to transfer value digitally: see section 1.4 and Chapter 12.
1.1 DECENTRALIZED
Most currencies in use today are fiat currencies, where the currency is issued by the government and its supply managed by a central bank.
FIAT MONEY
Most currencies today (Euro, US Dollar) are fiat money. Fiat money does not have intrinsic value, as it is not backed by anything. It is called fiat money because there is a government decree (“fiat”) declaring the currency to be legal tender. The acceptance of fiat money depends on expectations and social convention. If confidence in a currency is lost, usually because of irresponsible monetary policy, fiat money can stop being accepted.
Experience has shown that leaving monetary policy in the hands of governments is usually not a good idea, as governments could have an incentive to increase the monetary supply to solve pressing short-term financial problems. This behavior can lead to high inflation and a loss of confidence in the currency.
The conventional solution is to entrust monetary policy to a semi-independent central bank. The central bank is tasked with managing the monetary policy, usually with the goals of economic growth, price stability, and, in some cases, stability of the financial system.
Bitcoin is based on a peer-to-peer network of computers running the software. These computers are called nodes. Participants in the network might be running nodes for different reasons: for profit as in the case of miners (Chapter 9), to manage full-node wallets (Chapter 8), to collect and study information about the network (Chapter 13), or simply as a social good.
Bitcoin’s decentralized nature contrasts to the structure of fiat currencies. Central banks make monetary decisions after evaluating evidence gathered from the evolution of the economy. In a decentralized system such as Bitcoin, discretionary decisions are not possible. The original creators of the system have to take most of the decisions upfront at the design phase. These decisions have to be carefully balanced, and take into account the incentives of the different users, otherwise the decentralized system is doomed to fail. In Bitcoin the monetary policy follows a simple rule: the final monetary base is fixed at around 21 million bitcoins and new bitcoins are minted at a planned schedule and paid to users who help secure the network. This serves the double purpose of providing the bitcoins with value due to their scarcity and creating incentives for users to connect to the network and help secure it by providing their computational power.
Control in a centralized system is usually concentrated in an institution or a small group of key people. Thus changes in a centralized system are relatively straightforward to decide and implement. Control in a peer-to-peer network is more subtle: changes in a peer-to-peer network have to be agreed by a majority of the peers at least. But even then, if a strong minority does not agree to a change, implementing the change can be technically challenging as the network runs the risk of a split.
One advantage of the decentralization of power is that changes that are contrary to the interests of most users would be rejected. In contrast, in centralized systems sometimes the outcomes are adverse to most of the participants, as in a currency debasement by excessive printing which usually leads to high inflation.
Another feature of decentralized systems is their resilience. Decentralized systems are robust against attacks either by insiders or by external forces. This feature might have been critical for the existence of Bitcoin. Earlier centralized attempts to create digital currencies (section 2.1) were forced down by governments. However, to force down a decentralized system, all individual users must be forced down, which is a much harder task. Bitcoin’s peer-to-peer nature makes it censorship-resistant, claim its supporters.
The technology to securely (cryptographically) transfer value digitally had been available many years before the creation of Bitcoin (Chapter 10). However, it had always required the creation of a centralized trusted party. Bitcoin not only does not require a central trusted party to operate, but it is also designed to resist the attacks of malicious participants in the peer-to-peer network. As long as these malicious participants do not control a majority of the network these attacks will not succeed (section 7.5).
The main technological breakthrough accomplished by Bitcoin is solving the double-spending problem in a distributed financial database. A double-spend attempt occurs when a user tries to spend some funds twice. All financial systems must reject these attempts. This is relatively straightforward in a centralized system, as transactions are recorded in a central database and future spending attempts are checked against this database first. In a decentralized system, many copies of the database are shared among the peers, and keeping a consistent state of the database is a difficult computational problem1. In the context of Bitcoin the problem is how the network can agree on the state of the distributed database when messages between the nodes can be corrupted and there might be attackers trying to subvert the distributed database. Bitcoin gracefully solves this problem (section 2.3 and Chapter 7).
1.2 OPEN SOURCE
Bitcoin is open source software. Open source software makes the source code available for anyone to use, modify, and redistribute free of charge. Some well-known open source software products include the Linux and Android operating systems or the Firefox web browser. A large portion of the internet infrastructure runs on less known (but no less important) open source software. The goal of open source is to make software development similar to academic peer-reviewed research. By publishing the source code for anyone to see and check, open source aims to increase the quality of the software.
The difference between open source software and proprietary software lies in their licenses. A proprietary software license grants the right to use a copy of the program to the end user. However, ownership of the software remains with the software publisher. In contrast, an open source license grants the user the right to use, copy, modify, and redistribute the software. The copyright of the software remains with the creator, but the creator of an open source software transfers the rights to the user as long as the obligations of the license are met.
Another difference between proprietary and open source programs is that proprietary programs are usually distributed as compiled binaries. This means that the software is usually distributed in machine language. Users willing to gain knowledge on what the software is doing must interpret the machine code in a time-consuming process called reverse engineering (Eilam, 2005). Most proprietary licenses forbid the use of these reverse engineering techniques. Thus under a proprietary license the user is usually not allowed to understand or seek knowledge of what the software is actually doing. In contrast, open source software is always distributed with a copy of the source code. A user who wants to understand what the software is doing can just read the source code. Cryptographic open source software has the advantage that it allows users to check that the code does not contain any backdoor or security vulnerabilities2.
It is unlikely that Bitcoin could have been released under a proprietary license. Had Bitcoin been released as closed-source, its creator could have easily inserted code that deviated from the specification: say, creating new bitcoins and sending them to an address controlled by him. Most users presumably would not have accepted decentralized cryptographic financial software distributed as a compiled binary and with a proprietary license. It is telling that most competing cryptocurrencies (Chapter 11, section 12.7), have either been launched using an open source license or have switched to an open source license.
Open source licenses grant the user the right to use, copy, modify, and redistribute the software. Different licenses may impose different obligations on the users. Broadly speaking, open source licenses belong to one of two families:
“Copyleft.”
These licenses impose the obligation to distribute derived works under the same license. If a user of the software makes modifications to it, she is obliged to release the modified software under the same license. This is referred to as the share-alike requirement. Thus “copyleft” licenses preserve the open source nature of the software as it is modified. An example of a “copyleft” license is the GNU
Public License (GPL)
.
“Permissive.”
These licenses impose very few restrictions on the redistribution of the software, usually just that the derived software acknowledges the original software and retains the copyright notice. Proprietary software that incorporates software released under an open source permissive license retains its proprietary nature as the license usually only requires that the proprietary software includes the copyright notice. Several common open source licenses belong to this family, such as the BSD license, the MIT License or the Apache License. Bitcoin was released under the MIT license.
Proprietary software requires that the company issuing the software maintains and updates it. In contrast, open source software acquires a life of its own once released. It usually does not matter if an original creator decides to stop working on an open source project, as other developers could take it over. For this reason it does not matter who Satoshi Nakamoto is, or that he has moved on. Open source projects are resilient: even if some developers are forbidden or discouraged to work on a project, other developers from all around the world can take over.
Under an open source license it is legitimate to start a new independent software project from a copy of an original project. This process is called forking. The threat of a fork can often keep the developers of an open source project honest. If the developers of a project introduce changes that are detrimental to the users of the software, anybody can create a fork, undo those changes and continue the development. Users will most likely follow the fork without the undesired features. Thus forking can be seen as a kill switch that prevents developers from evolving a project against their users. Most large open source projects are rarely forked3. Bitcoin is somewhat special in this respect, as it has been forked many times by developers wishing to test new concepts. This has given rise to many alternative cryptocurrencies called alt-coins. Alt-coins will be covered in more detail in Chapter 11.
Open source advocates argue that companies releasing proprietary software often lose the incentive to innovate once a product has achieved a dominant market position. Many software markets behave like natural monopolies where a product with first mover advantage can capture a large market share. Thus innovation in many software categories is low, these advocates suggest. In contrast, if an open source software captures the majority of the market this does not bring about the end of innovation, as anybody can keep on adding improvements to the software. Thus the pace of innovation in open source software can be higher than in closed source software.
One problem facing many open source projects is the tragedy of the commons. Although many people benefit from an open source project, few developers might have an incentive to contribute to it. Many open source projects face difficulties in getting appropriate funding or development time. There have been some indications that Bitcoin could be facing this problem (Bradbury, 2014b).
An exposition of the merits of open source software can be found in Raymond (2001).
1.3 PUBLIC ASSET LEDGER
The heart of Bitcoin is a distributed database that holds a copy of the common asset ledger. As this database is distributed, each participant in the network (a node) keeps a copy of it. Copies of this database kept by the different nodes are consistent by design.
On the other hand, every user is in control of her own funds, through a cryptographic private key. When a user wishes to spend some funds, she must use this private key to sign a message that states who she wishes to send the funds to as well as the amount to send. The user broadcasts this signed message to the network, and every participant in the network receives a copy of it. Then each node can independently verify the validity of the message and update its internal database accordingly4.
In traditional financial systems, value is represented in ledgers (databases) managed by financial institutions. Users must place trust in these financial institutions that these databases will not be subverted either by insiders or by outside attackers. The protocols and procedures that safeguard traditional financial databases are not generally revealed to the public. In contrast, Bitcoin makes the database public and creates an open source software protocol to secure it. This protocol is designed to be resilient against attackers participating in the network. Bitcoin users do not need to place trust on any entity: the system is said to be trust-less.
All the financial information flowing through the Bitcoin network is public, except the identities behind the transactions. Bitcoin does not use personal information to identify the holders of funds, but Bitcoin addresses. Addresses are long strings of seemingly random letters and numbers, such as “13mckXcnnEd4SEkC27PnFH8dsY2gdGhRvM”. Bitcoin is like making everybody’s bank statements public online, but with the identity blacked out (Back, 2014b).
Although in principle there is no way to associate addresses to identities, there are many techniques to analyze the information flowing through the network and acquire different grades of knowledge about Bitcoin addresses and the users behind them (Chapter 13).
Bitcoin is not anonymous, and it can sometimes be less anonymous than the traditional payment systems. In the traditional payment system, for instance, an employer does not gain knowledge of where an employee spends her wage, although the employee’s bank has that information. If an employee were paid in bitcoins, her employer could see where she spends the money simply following the trail of transactions emerging from the address where the wage was sent to. The employee could follow some practices to hide this trail of transactions (Chapter 13).
In other cases, this transparency can be an advantage. One such example is the case of public entities where a transparent destination of funds could help increase the quality of the administration and help avoid corruption. In the case of commercial enterprises some level of transparency can be beneficial, for example financial statements that could be verified against the public ledger. There has been some technological progress towards achieving different levels of transparency in a public ledger system (section 8.5).
1.4 IT’S NOT ONLY THE CURRENCY, IT’S THE TECHNOLOGY
Transfer of value has traditionally been a slow and highly manual process. In essence, Bitcoin is a protocol to create distributed consensus. This protocol allows transferring value securely in a trust-less way: it is an open platform for money. But it is not only restricted to money: Bitcoin and similar protocols can transfer any digital asset (Chapter 12). The technology is cheaper and faster than most of the alternatives, creating opportunities for new applications.
The digital transfer of value enables the adoption of smart contracts. Smart contracts are contracts that do not require human interpretation or intervention to complete. Their settlement is done entirely by running a computer program. Smart contracts are math-based contracts, as opposed to law-based contracts. A trust-less digital transfer of value opens the door to new applications that make use of smart contracts.
One such application is autonomous agents. Autonomous agents should not be confused with artificial intelligence. Autonomous agents are just straightforward computer programs, created for a specific task. One example is a computer program running in the cloud that rents storage space and offers end users file-sharing services. Up until now computer programs could not hold value: a computer program presumably could not open a bank account in its name. With the introduction of Bitcoin, computer programs can control their own funds and sign smart contracts with cloud service providers to rent cloud storage and computing power. Similarly a storage agent could enter into smart contracts with its end users. The storage agent can settle these smart contracts, making bitcoin payments to the cloud provider and receiving bitcoin payments from its end users (Garzik, 2013a). A more extensive discussion of autonomous agents can be found in section 12.4.
Autonomous agents are just one example, and many more innovative ideas are being devised (Chapter 12). Some of these ideas may turn out not to be practical, but maybe a few could become mainstream. A decentralized system is an ideal test ground for these technologies, as innovators do not need the approval of anybody to try out their ideas: a decentralized system enables permissionless innovation.
Bitcoin is an API (Application Programming Interface) for money and bitcoin the currency is just the first application. Bitcoin could be used as an open platform for the exchange of value in much the same way that the internet is an open platform for the exchange of information. It can be used as a protocol on top of which applications can be built, much like email, web browsing, or voice-over-IP are built on top of the TCP/IP protocol. This is where most of the excitement around Bitcoin and related technologies comes from. Regardless of whether bitcoins have a future as currency, the technology has shown that many applications are now possible and innovators will continue to push forward with new ideas. Bitcoin could become a platform for financial innovation.
One of Ronald Coase’s most important economic insights in The Nature of the Firm (Coase, 1937) was that one factor that contributed to the creation of firms was high transaction costs. If there were no transaction costs, an entrepreneur could contract any good she needs in the open market, and this would be efficient, as an efficient market would always achieve the best price for that good. However, transaction costs, such as information gathering, bargaining, policing the contract, keeping secrets and so on, can be a significant portion of the total cost of contracting out to the market. For this reason, it might be cheaper for an entrepreneur to hire some employees to produce the goods internally, thus starting a corporation. Transaction costs are also at the root of public goods and government action.
Bitcoin’s technological breakthrough creates an opportunity to lower the costs of entering and upholding contracts, say through smart contracts. More efficient contracts thus have the potential to change corporations and government action.
1
This computational problem is called the Byzantine Generals’ problem, introduced in Lamport et al. (1982).
2
This should not be interpreted that open source code does not contain security flaws or backdoors. Indeed, many security flaws have been found in open source projects (Green, 2014b; Poulsen. 2014). Open source advocates argue that it is more difficult to include flaws and backdoors into open source programs because there is a higher level of scrutiny, and that these flaws are typically discovered and repaired sooner than similar flaws placed in proprietary software (Raymond, 2001).
3
Most projects are really forked many times by individual users wishing to tinker with them or test new features. However, forks of large open source projects that split the developer base, such as the LibreOffice fork from OpenOffice (Paul, 2011), are rather rare.
4
The process is actually more involved to prevent double-spending attacks where a user sends different messages to different parts of the network. How Bitcoin prevents double-spending attacks is the subject of Chapter 7.
CHAPTER 2Technology (Introduction)
Until the introduction of Bitcoin, transmitting money digitally had required the mediation of a third party. The main breakthrough of Bitcoin has been to allow digital payments with no trusted third party. This chapter serves as an overview of the technology behind Bitcoin.
2.1 CENTRALIZED DATABASE
The most straightforward way to try to create digital value is to assign value to a certain data pattern, basically a string of zeroes and ones. The problem with this approach is that digital information is easy to replicate at basically no cost. This leads to the double-spend problem, exemplified in Figure 2.1. Say Alice has a digital coin, represented by the binary number 01000101. She could transfer this value to Bob, by sending him a message with this number, so that Bob had a copy of the number and thus the value. The problem is obviously that nothing prevents Alice from sending this same number to another user or indeed to many other users.
FIGURE 2.1 Double-spending problem
So digital value cannot be represented simply as a number because digital data is very easy to replicate many times and thus knowledge of the number does not have any value. As common sense suggests, for something to have value it must be scarce. The challenge then is how to create scarcity using digital technologies that allow the perfect copying of information.
The next step towards building a digital payment system is to create a central database, holding a list of the users and the funds held by any of them. This system is shown in Figure 2.2.
FIGURE 2.2 Central counterparty holding a centralized database
Now if Alice wants to transfer 1 unit of the currency, say a token, represented by the number 01000101 to Bob, she contacts the server running the central database and directs it to transfer this token to Bob. The server updates the database, and the token now belongs to Bob. If Alice tries to double-spend the token 01000101, sending it to Barry this time, she would have to again connect to the central server and direct it to send the token to Barry. However, upon checking the database, the server sees that the token 01000101 does not belong to Alice any more, and thus she is not authorized to spend it.
A central database solves the double-spend problem. However, there are issues associated with a central database. For a start, all users must have previously registered with the central server in order to operate. Thus the central database knows the identities of all the users and collects their financial history1. A central database is also an easy target to attack, either by insiders or by outsiders. If an attacker gets control of the central database, she could change the ownership of any funds, thus stealing them from their legitimate owners. Or she could create new funds (tokens) and assign them to herself.
Perhaps the main drawback of a central server is that it constitutes a single point of failure, as portrayed in Figure 2.3: the payment system can be easily taken down by shutting the central server.
FIGURE 2.3 Central counterparty single point of failure
Some early digital payment systems were based on the idea of a central database holding the positions of all the users. Two famous examples are e-gold and Liberty Reserve. E-gold ceased operations in 2009 (Wikipedia, 2014h), and Liberty Reserve in 2013 (Wikipedia, 2014i).
Figure 2.4 shows the analogy between BitTorrent and Bitcoin. Both are systems where the coordination of information is done in a decentralized way. In BitTorrent (the protocol) any user can create a torrent descriptor and seed the file into the network. Other users in possession of the torrent descriptor can then connect to the network and retrieve the file (Wikipedia, 2014d). Bitcoin’s ledger database is distributed and maintained by many computers called nodes. Bitcoin users can send new transactions to this distributed database, where they are recorded. Both systems are resilient, even in scenarios where a large portion of the network is forced down.
FIGURE 2.4 Analogy between BitTorrent and Bitcoin
2.2 ADDRESSES, TRANSACTIONS
At the center of the Bitcoin network is a decentralized ledger that contains the balance of every Bitcoin user. Bitcoin identifies users by large strings of letters and numbers such as “13mckXcnnEd4SEkC27PnFH8dsY2gdGhRvM”. The address is the public part of a public–private cryptographic key2. The private part of the key is under the control of the user. Figure 2.5 shows how a user (Alice) sends some funds to another user (Bob): Alice uses her private key to sign a message saying “I want to send 1 bitcoin to 1gr6U6...” that she sends to the network. Note that Alice does not identify the user she wants to send funds to, just the address to receive the funds. Thus Alice must find out Bob’s address through other means.
FIGURE 2.5 User sending funds. State of the database after the transaction has settled
Upon receiving Alice’s message, nodes in the network follow these steps:
They verify that the signature is correct. If it is not they reject the message.
They check that the sending address has enough funds to honor the transaction. If there are not enough funds credited to the address, the transaction is considered invalid.
Finally, they update the database, subtracting the funds from one address and crediting them to the other.
An important detail is that nodes in the network do not know the identities of either Alice or Bob, as users are identified only by their addresses. Bitcoin users are identified by a pseudonym: Bitcoin provides pseudonymity.
Another important detail is that addresses are not granted by the network. They are created inside the users’ devices when it runs the Bitcoin software that generates the cryptographic public and private keys. As the public and private keys are intimately related (Chapter 5), they have to be generated jointly and locally on the user’s device. The address generation process is straightforward and can be performed almost instantaneously by any device such as a laptop or a smartphone. There is also no restriction on the number of addresses that a user can create. Indeed, it is recommended that users generate many addresses to enhance privacy (Chapter 13).
No prior registration is necessary to use Bitcoin. In fact, new users do not even have to communicate their addresses to the network to be able to receive funds. A user, say Bob, can generate an address and communicate this address to Alice through other means, such as an email or the pairing of two smartphones. Alice can now send funds to Bob’s address and the network would accept the transaction even though it has never encountered that address before.
In a centralized system the funds are held by a central entity, which also holds the means to control those funds, say by changing the registries in the ledger. In contrast, in a decentralized system, the private keys that give access to the funds are solely in the hands of the end users.
Addresses, public–private keys and transactions are discussed in more depth in Chapters 5 and 6.
2.3 DISTRIBUTED DATABASE, THE BLOCKCHAIN
Bitcoin’s distributed database is called the blockchain. Transactions are grouped in blocks of transactions roughly every 10 minutes. These blocks of transactions are then recorded one after the other in a chain of blocks, hence the name blockchain. This may seem a strange way to record information, compared to, say, a regular relational database. The blockchain was designed to be resilient in the presence of attackers in the network. Blocks are linked to create a record of the history of transactions that cannot be altered. The link between blocks is a cryptographic link that cannot be forged unless the attacker has vast computational resources at her disposal. The blockchain is discussed in greater detail in section 7.4.
Aside from the blockchain, nodes keep an additional database called the Unspent Transaction Outputs cache (UTXO) (Chapter 6). The UTXO is a ledger that records the funds available for every address, in essence working as a cache for the blockchain. As new transactions come, the UTXO is updated: funds from the sending addresses are subtracted and added to the receiving addresses. The UTXO is more similar to the central databases at the heart of most centralized systems. Figure 2.6 shows a, sometimes useful, abstraction for Bitcoin: a distributed ledger with entries for the funds available to every address, which roughly corresponds to the UTXO. Every node in the network holds a copy of the distributed ledger. Furthermore, copies of the ledger are consistent across nodes, and new transactions have the same effect in all these copies.
FIGURE 2.6 Bitcoin as a distributed ledger
Bitcoin achieves consensus in the distributed database using several cryptographic constructions. The details can be found in Chapter 7, but roughly speaking, consensus is secured applying large amounts of computational power. This computational power serves the purpose of providing protection against attacks and is rewarded with the issuance of new bitcoins. The protocol encodes a schedule of new bitcoin creation, and all the newly created bitcoins are distributed among those who secure the blockchain, called miners. Miners compete to create blocks of transactions that are appended to the blockchain. A miner who creates one of these blocks is granted the block reward, consisting of a certain number of newly minted bitcoins. A native currency is essential to the design of Bitcoin, as the issuance of new currency is used to pay for the cost of securing the distributed ledger.
Figure 2.7 shows the schedule of bitcoin creation. The pace of new issuance is halved roughly every four years, so that eventually the total number of bitcoins will reach a total of roughly 21 million. The number of bitcoins in circulation, as of the time of writing, is around 13 million. Bitcoins’ value stems from their scarcity, as the number of bitcoins that will eventually be issued is fixed.
FIGURE 2.7 Bitcoin issuance theoretical schedule
Miners also collect fees from the transactions that are published in the blockchain. Fees are still a small fraction of total miners’ compensation, currently below 1% of their total compensation. It is expected that as the issuance of new bitcoins shrinks, transaction fees will take over as the principal compensation to miners.
During the end of 2013 and beginning of 2014 there has been an investment boom in Bitcoin mining equipment. It is estimated that over USD 200 million were invested in Bitcoin mining equipment in 2013 (Luria and Turner, 2014). This investment rush has been fuelled by the increase in the price of bitcoins and by technological evolution in mining equipment (Chapter 9). This investment trend will likely ease in time, barring another large increase in the price of bitcoin, with the future decrease in issuance of new bitcoins and the mining technology catching up with state-of-the-art semiconductor process technology.
An attacker who wished to subvert the distributed database to perform a double-spending attack must enter a race with legitimate nodes3. The result of this race is determined by the amount of computational power. A straightforward attack would require a computational power as large as the power of the legitimate network. That is, the attacker would need to control more than 50% of the combined power of the network. This type of attack is called a 51% attack. Other types of attacks, requiring somewhat lower fractions of computational power, are possible. In any case, an attacker would have to devote a significant investment to be able to mount an attack against the blockchain.
