Digital Currencies & Networks
Blockchain Beyond Bitcoin | Lecture 2
Recap
Welcome! If you’re a returning reader, thank you for keeping up with our series, Blockchain Beyond Bitcoin! If you are a new reader, we’d like to provide you with a TL;DR from the previous lecture.
In State of the World, we explored the concept of “trust” and how it has evolved over time. While “trust” existed initially as a social construct dictated by gestures, speech, and sharing, society organized into governments and companies who allowed for us to interact with strangers via institutional trust. Since the invention of the internet, we have discovered a new paradigm known as programmable trust, which is guaranteed by the soundness and objectivity of science and math. Many companies have attempted to create programmable trust via cybersecurity, but their approaches are both costly and brittle given the dynamic nature of attacks.
We also introduced a brief history of the internet, the development of a decentralized, permission-less technology which would enable exponential increases in quality of life, while simultaneously providing a sphere for freedoms and opportunities which never before existed. Despite this history, the re-centralization of the internet into the hands of gatekeeper companies like Google and Facebook, which control access and distribution, has forced users to cede control of their personal data and information in exchange for the ability to participate in far-reaching networks. This consolidation has undermined the internet’s integrity, allowing for it to be manipulated both functionally (hacking & ransomware) and politically (spam & fake news). In light of recent events, including online Russian political interference and the Equifax breach, it’s clear that privacy, transparency, and regulation on the internet stack all need to be reconsidered.
Last Week’s Reading
As mentioned in our last lecture, we will be assigning weekly readings. For our first selection, we assigned The New Vulnerability: Data Security and Personal Information by Daniel Solove, embedded below.
Last week, we considered the institutional trust relationships which serve as a foundation for the global economy. In our modern financial system, the items used as money often have little intrinsic value. As you might recall, the paper used when printing the US dollar is not particularly valuable. This leads us to dig deeper on what money means to us…
What is “money”?
In order to establish a working definition of money’s role in modern society, we need to understand how our instruments of trade have evolved over time. As any introductory economics class will tell you, trade exists due to specialization and division of labor (Adam Smith, The Wealth of Nations, 1776).
Trade exists between regions because of comparative advantage in production of a commodity. Different regions’ size, geography, and location offer different benefits for production. Trade has existed since prehistoric times, in the form of barter.
Brief History of Barter Systems
Bartering involves the provisioning of one good or service by one party in return for another good or service from another party. There are countless examples in human history of this type of exchange in practice: (1) Mesopotamian tribes were likely the starting point of the bartering system (6000 BC). (2) In the Roman Empire, soldiers bartered their services in exchange for salt from the emperor (27 BC–286 AD). (3) In Colonial America, the colonists used bartering to get goods and services from the Native Americans (1607–1776).
These systems were practical for their needs, but barter as an instrument of trade is limited by time and trust. Time must be spent: (1) finding another party with a mutually exclusive need. Without a standard measure of value of goods and services, time must also be spent (2) negotiating on and agreeing to terms. Additionally, social trust is a limiting factor because the (3) representation of goods and services must be accurate, as there is little recourse following the transactions completion.
Early, Commodity Money
As the need for more time-efficient and trust-less instruments of trade became apparent, a new invention was on the horizon. Primitive forms of money begin to replace barter, introducing a standard measure of value that is widely understood and through which everyone can identify a need. These early examples are classified as commodity money, which are objects that have intrinsic, material value due to scarcity. (1) Cattle are the first and oldest form of money (9000–6000 BC). (2) Cowrie shells are the most widely and longest used money in human history (1200 BC). (3) Modern coinage begins to appear made from precious metals in early civilization (500 BC).
Despite the improvements made upon barter, commodity money as an instrument of trade is limited by portability and durability. Irrespective of the value assigned to an individual unit of commodity money, there is always an upper bound on (1) how many units can be carried in a portable way. This limits an individual’s purchasing power. These commodity units are also quite clearly subject to the (2) decaying forces of mother nature: cattle have finite lifetimes, shells can break, and metal coins will rust.
Innovative, Representative Money
The need for more durable and practical-to-carry instruments of trade motivated the introduction of representative money, which are classified as objects that can be exchanged for a fixed amount of a given commodity. Again, we can recall many historical instances of this type of system. (1) In China, the Tang Dynasty introduced paper money, reducing the need to carry heavy metallic coins (618 AD — 907 AD). (2) In England, Gold was made the standard of value; printed banknotes could be exchanged for a certain amount of gold (1816). (3) Following WWII, central banks around the world could pay the US federal government $35 for an ounce of gold (1945).
Representative money proved extremely effective but, as an instrument of trade, it is limited by the protection and legitimacy of the backing physical assets. When President Nixon was worried about a run on America’s gold supply, he cancelled the WWII agreement (1971). The commodity that a unit of representative money can be exchanged for must be (1) protected physically. Somewhat paradoxically, there must also be (2) some legitimacy mechanism for proving possession of the commodity, which undermines the security of the assets.
Delicate, Fiat Money
As governments around the world began to gain power, they imagined a new type of money which would prevent them from worrying about accumulating and guarding a commodity store. Fiat money is currency that a government has declared to be legal tender without being backed by a commodity. The value of fiat money is derived from the relationship between supply and demand rather than material value. This is largely how money is managed throughout the world today. In the US, the Federal Reserve has a mandate to induce economic conditions which keep unemployment and inflation low.
Despite being a widely implemented approach to money, as an instrument of trade, fiat money is limited by government regulation of inflation and potential devaluation of currency. To understand this potentiality, we can look to Zimbabwe’s recent history.
In the late 1990s, the Zimbabwe government introduced a series of land reforms. This redistributed land from existing white farmers to untrained black farmers, which led to a large fall in food production. The economy experienced a sharp fall in agricultural and manufacturing output causing a collapse in bank lending. As a response, the government began printing money at an increasing rate, shifting supply. With the economy in decline, government debt increased. To finance the higher debt, the government responded by printing more money, leading to a record-setting inflation rate in 2008.
Important Properties of Money
What can we learn from these historical examples? most economists and experts agree upon 3 fundamental properties that all forms of money necessitate.
- medium of exchange := used to intermediate the exchange of goods and service by providing a common ground for determining value
Example: A community uses beaver pelts as a medium to trade for other goods - unit of account := standard numerical unit of measurement to compare goods using a common system
Example: Housing prices in Japan can be compared using the yen as a unit of account - store of value := maintains its value over time, and can be spent or exchanged at a later date without penalty
Example: 1 oz of gold could buy a toga in Roman times, and can still buy a nice suit today
With this new understanding of money, we turn to a discussion of digital money and the precursors to Bitcoin.
Previous Attempts
DigiCash
In 1983 David Chaum, a PhD student, publishes a scientific paper describing a new form of digital currency. Chaum’s digital money differs from credit card payments through anonymity. This is enabled through cryptography to create a blind, digital signature. But what does this mean? This type of signature can be analogized as the physical act of voting.
Imagine a voter encloses a completed anonymous ballot in a special carbon paper-lined envelope that has the voter’s credentials preprinted on the outside. An official verifies the credentials and signs the envelope, thereby transferring his signature to the ballot inside via the carbon paper. Once signed, the package is given back to the voter, who transfers the now signed ballot to a new unmarked normal envelope. Thus, the signer does not view the message content, but a third party can later verify the signature and know that the signature is valid within the limitations of the underlying signature scheme.
Chaum’s blind signature scheme enables the secure and anonymous payment of goods and services on the Internet, in stark contrast to the widely used credit card system. However, Chaum is smeared in the press as a paranoid cryptographer: deals with Dutch ING Bank and Bill Gates, who wants to integrate DigiCash into Windows 95 for $100 million, fall through after extensive negotiations. This reveals a fundamental flaw: DigiCash depends on Chaum. His errors reflect on the entire digital currency. Chaum is a fan of patents and copyrights, which is good for him but bad for bringing his technology to the masses.
E-Gold
In 1996, Douglas Jackson and Barry Downey propose a new currency. Their idea is to put gold coins in a safety deposit box in Florida, and create a website where digital portions of these coins are sold. These coins are denominated in e-gold units. After 4 years, e-gold becomes the first digital payment system to be used by more than 1 million people. E-gold can be integrated into online shops and thus enables pure digital e-commerce. The units are divisible into thousandths of a gram of gold which serves as the basis for the first functioning micropayment system. This type of system was not previously possible with credit cards because of high transaction costs. At its peak, e-gold has a market capitalization of $2 billion.
Despite e-gold’s ability to solve DigiCash’s adoption problem, its central systems are insecure and are hacked. Many users of e-gold are vulnerable to exploits in Windows and Internet Explorer, and some lose their deposits. Following the Sept. 11th, 2001 attacks, the Patriot Act allows for the suspension of many civil rights, including privacy. More than five years after e-gold had been launched, the US Treasury Department proceeded to prosecute USA-based gold systems, e-gold (and later e-Bullion) for not having money transmitter licenses, even though these companies had previously been cooperating with regulatory authorities and told they did not fall under the definition of money transmitter. Shortly after, the combination of adverse publicity and disrupted exchange markets led to a precipitous decline in e-gold usage and demand.
So what lessons can be learned from e-gold’s short-lived success and rapid collapse?
- A monetary system based on gold may lead to state-run blackmail
- A company domiciled in a country may be at the mercy of changing legislation
- Founders who can be identified as real persons may be targeted
PayPal
In 2002, a company called PayPal was acquired by eBay for $1.5 billion and was made eBay’s default payment setting. To setup a PayPal account, users could link their credit or debit card information to allow for online transfers of value. In 2007, PayPal partnered with MasterCard and made way for customers to pay on websites that accepted MasterCard payments but did not support PayPal’s infrastructure. In 2010, PayPal suffered DoS attacks (denial-of-service) reportedly linked to the WikiLeaks donation payment process being stopped. In 2011, fourteen alleged members of Anonymous group were charged with attempting to disrupt PayPal’s operations. In 2013, PayPal acquired Venmo for $800 million. We include PayPal and Venmo in this discussion because they are often confused with unique digital currencies when in fact they are simply a new manifestation of legacy banking infrastructure.
Bitcoin
Next week, we will be diving deep on Bitcoin. Before we move on to a discussion of networks, however, we can recognize briefly how it improves upon the flaws of DigiCash and E-Gold. Bitcoin was developed as open-source software, which increased interest and knowledge about the technology, in contrast to Chaum’s strategy. Bitcoin also has a pseudonymous founder, Satoshi Nakamoto, which prevents association between the tech and an individual’s shortcomings.
On Networks, Mental Models, and Market Power
Fundamentally, digital currencies require some sort of relationship between participants in order for users to relay transactions; this requirement similarly exists for any system where an individual can benefit from the participation of others in the same system. Thus, we transition from a discussion on the history of currencies—which exist in both physical & digital forms—to a more abstract conversation about networks.
As USV mentions in its investment thesis, networks are interconnected groups or systems. Examples of popular online networks are centered around person-to-person sharing of activities (SoundCloud), transactions and personal finance exchanges (Venmo, LendingClub), creativity (Tumblr, Instagram), marketplaces (eBay, Etsy, Toptal), and data aggregation (Indeed, DuckDuckGo).
Generally speaking, building meaningful networks has become the primary area of focus for entrepreneurs seeking to build long & lasting companies in the technology industry. As mentioned by Tren Griffin—a veteran of the technology industry who writes occasionally for a16z—that the cost of developing prototypes has reduced due to the availability of robust open-source software, defensibility for an idea no longer stems from the underlying technology (code or intellectual property) it is built on.
Instead, defensibility is measured by the growth of services which create more utility and opportunity with more participants. Thus, investors (more specifically, venture capitalists) seek to invest in businesses by evaluating the networks they have built. This leads us to ask…
A Quick Detour on Mental Models
Decision makers have at their disposal possible actions; checklists of things to think about before acting; and mechanisms in mind to evoke these, and bring these to conscious attention when the situations for decision arise. — Herbert Simon, Economist & Political Scientist, 1996
Charlie Munger, associate of Warren Buffet and Vice Chairman of Berkshire Hathaway, spoke extensively on the importance of using mental models to approach decision-making. Simply put, mental models are “psychological representations of real, hypothetical, or imaginary situations” (source), which involve applying judgment acquired from mistakes made in the past to future decisions. These models can be applied in parallel, and different models can be complementary or self-reinforcing.
Below is a list of mental models compiled by Shane Parrish of Farnam Street. You might find that many of these are intuitive conclusions about human nature, while others are conclusions derived from disciplines including statistics, biology, and psychology.
Mental models are pervasive in business; we will focus on the concepts of network effects and critical mass, which together describe a strategy used by internet companies to establish competitive advantages and defensibility.
Network Effects
A network effect can be described as a phenomenon where increased numbers of participants improve the value of a service. The hallmark example of a network effect is the telephone; owning one becomes more useful as more of your friends & family members do, since you’re granted the ability to communicate with them.
We can extrapolate that hardware enables physical network effects, which are the strongest form of network effects. Similarly, ethernet (the physical standard on which the Internet is built) has enabled computing devices to capture a tremendous amount of value by allowing developers to build applications, marketplaces, and new networks on the web, which benefit from communication between the 4 billion (and counting) online.
To read more about the different forms of network effects, how they manifest, and excellent clarifying examples, we encourage you to read the following resource:
Network effects are threatening brand, regulation, intellectual property, and other factors which create moats, and have become the primary avenue for software companies to build a barrier to entry against competitors. Hence, internet-focused venture capital firms often explicitly list network effects as a desirable business attribute. For instance, USV invests in:
Large networks of engaged users, differentiated through user experience, and defensible through network effects.
It is important to note that network effects don’t always lead to direct financial benefit. For example, open standards like Ethernet generate and benefit from network effects, but no party yields direct financial benefits from it. Moreover, even if network effects are initially strong, they can disappear over time depending on the growth and strength of connections on the network; examples include Blackberry Messenger (BBM) and Palm.
Network effects can be measured on two axes, for which examples are noted above. For context, we discuss the network effect of phone ownership:
- Positive vs. negative: an increase in the number of users on the network can either positively or negatively affect the utility captured by others on the network
- Direct vs. indirect: the additional value created for users on the network due to increased throughput or activity can either directly or indirectly affect user experiences on the network.
An indirect positive impact might be the increased availability of third-party applications or accessories for your phone, since developers are incentivized to allocate resources towards app development if there is a large user base. A direct negative impact might be network congestion due to a large number of users attempting to share content online (for instance, in a sports arena).
An Accurate Characterization
It is important to understand the limits of network effects. A classic definition from Robert Metcalfe, the co-inventor of the Ethernet standard, states that the value of a network grows in proportion to n², where n is the number of users.
But Briscoe, Odlyzko and Tilly describe in a piece on IEEE that this wrongly assigns equal value to all connections or groups on the network. The importance of the strength of the connection, defined by the value of the exchange between participants on the network, results in a proportion lying somewhere between linear and exponential growth.
Critical Mass
Critical mass is another essential mental model to understand internet networks. The term first came into use around 1941, in the context of an amount of uranium required to cause a chain reaction. The term was later adopted by other disciplines; in business, it came to refer to the size a company must be in order to competitively participate in a marketplace, sustaining both growth & efficiency.
In technology, critical mass refers to the number of users required to create network effects that generate moats (defensibility) for the business. Where network effects can be strong enough, there might not exist a second place, depending on how quickly a company can reach critical mass.
A classic example is the war between the Betamax and VHS formats for video in the late 70’s. While Betamax was superior in quality, VHS reached critical mass, penetrating the market via retail establishments like Blockbuster and resulting in the vast majority of consumers purchasing VHS players.
“The tipping and de facto standardization of the VHS format in 1981–1988 is believed to have been caused by network externalities [aka network effects]. In the time period, watching prerecorded videotapes such as movie titles became the most important reason to use VCRs. Hence an increase in the users of VHS VCRs could raise a variety of available movie titles and thus the demands for VHS VCRs. That is, indirect network externalities became signification in the home VCR market.”
— Sangin Park [source]
The “Extraction Imperative”
While businesses with network effects provide more value to participants the larger they grow, they are difficult to get off the ground, where growth is usually resembled by an S-curve. Those that do reach critical mass tend towards becoming natural monopolies, using their competitive advantages to squash competition and extract fees and rent from users and third-party developers who have nowhere else to go.
Given that these businesses are usually private companies with a duty to shareholders, extracting rent from consumers and developers is not optional; rather, it is a fiduciary responsibility. Network effects and critical mass combine to form what KJ Erickson describes as the extraction imperative. When the value of participating and switching are both sufficiently high, people consolidate around a network and create an environment where the business can extract rent without fear of losing users.
Relationships with users and developers move from positive-sum to zero-sum.
The negative impacts of the extraction imperative counteract the positive impacts of network effects on its users, until the business reaches a point where it may no longer be positively impacting the individuals and organizations which built on top of the opportunities it offered. Thus, network effects can be largely positive, but over time begin to harm services.
These downsides are exacerbated when these powerful network effects are controlled by monopolies, who assert market power. This results in businesses overreaching, extracting an amount of rent that is non-optimal and results in the classic economic circumstance of deadweight loss.
Market power arises when users or customers have few comparable alternative options for sources of the good or service being provided. This gives the seller the ability to raise prices, or in the case of some internet giants to charge transaction fees, compile and sell user data, all as a condition for giving users access to the platform.
— Cathy Barrera, MIT Cryptoeconomics Lab
Thus, network effects housed within natural monopolies who have reached critical mass are fundamentally problematic.
What’s next for networks?
Next week, we will discuss Bitcoin, the first attempt at a truly decentralized digital currency which leverages network effects to distribute value to its developers and users. Bitcoin sets the stage for a wave of decentralized protocols around which developers hope to build networks resistant to the perils of market power.
In the meanwhile, we hope you continue reading about the history of digital currencies & internet networks at the resources below.
Resources
If you’re interested in peer-reviewing our content, please feel free to make comments in the discussion or inline via Medium. We highly value feedback, and want to ensure that our content is accurate & meaningful — all help is appreciated. You can also reach out to us at blockchain@rice.edu with any private feedback.
){kind=link}