A brief history of insurance, and what it means for cryptocurrency and blockchain.
With risks undeterred, opportunities are deferred. The growth of an industry necessitates sophisticated insurance products in order to mitigate the inherent risks of exploration and entrepreneurship. It is no different in the nascent crypto/blockchain space …
The idioms “save for a rainy day” and “don’t put all your eggs in one basket” originate in the 1500s. They still resonate half a millennium later, in invisible shape and infinite form. Indeed, insurance — in some form or another — dates back to prehistory. Various ancient cultures have independently discovered the benefits of preventing, hedging, and distributing risk—beginning with public community granaries protecting against food shortages.¹
In ancient times, insurance-like provisions were attached to marine loans, and burial societies offered something analogous to life insurance, but the earliest known true insurance policy dates to 1343.²
These pre-modern societies arrived at the practice of distributing risk and pooling future compensation early on. It wasn’t until the late 1600s that the field of statistics emerged—along with the formalization of the Law of Large Numbers— allowing for mathematically precise insurance models to emerge. This was the birth of modern-day insurance.
The Code of Hammurabi — one of the first legal texts in existence, written around 1750 BC — included an early insurance policy in the form of a clause that safeguarded a debtor, allowing them to forgive their loan if some personal catastrophe impeded payment, thereby offering the debtor some relief from risk.
If any one owe a debt for a loan, and a storm prostrates the grain, or the harvest fail, or the grain does not grow for lack of water; in that year he need not give his creditor any grain, he washes his debt-tablet in water and pays no rent for this year.
— The 48th code of law from the Code of Hammurabi⁴
The first recorded forms of insurance are attributed to the Babylonians and ancient Chinese traders⁵. Merchants would divide their items across ships having to cross treacherous waters, minimizing the risk of total loss. Even without the explicit knowledge of the Law of Large Numbers, ancient civilizations intuitively began to see the value in spreading their assets in order to minimize such devastating losses.
The birth of modern-day insurance emerged from the ashes of the Great Fire of London in 1666.
The Great Fire devastated London. There were few recorded deaths, but estimates put the destroyed property value at £10,000,000 (£1.5 billion in today’s money). From the ashes rose an unlikely development: the world’s first property insurance policies [covered by Lloyd’s, an insurance empire which stands to this day].⁶
Collections (read: donations) at churches across England were the initial solution to compensate for the losses. This quickly proved insufficient. Further, due to various squabbles between landowners and tenants regarding who was to pay for damages, an emergency ‘Fire Court’ was set up in 1667 to handle disputes. In the same year, the first insurance company was created, plainly called “The Insurance Office”⁷. Increasing demand led to insurers assembling shops throughout London — with names like the “Friendly Society” and the “Hand-in-Hand”— and by 1720 “they had underwritten 17,000 policies totalling £10 million — enough to cover the estimated cost of property that the Great Fire had destroyed.”⁸
Insurance — in its modern form — quickly shifted from a mere convenience to an urgently needed institution for preserving the well-being of society.
Perhaps it’s too obvious to draw a parallel to an event that occurred last year in the crypto/blockchain space. On Crypto’s Black Thursday, cryptocurrency markets collapsed, where bitcoin fell to half its previous day’s price. Indeed, there were insistent whispers of insurance talk shortly after the event:
- some DeFi platforms fared better than others due to having some form of a built-in insurance fund⁹
- some discussion of secondary insurance markets to protect against the risk of undercollateralized loans¹⁰
Ultimately, however, the concept of insurance hasn’t fully penetrated the collective mind state of the cryptospace. This can be easily seen by the limited number of insurance projects in the space — compared to, say: other financial projects, infrastructure projects, and media/entertainment projects (the latter most well-exemplified by the burgeoning NFT space).
As the mirror opposite of gambling, the entrance of insurance into the crypto landscape was inevitable. The two have in common the fact that participants buy into a probabilistic mechanism with the expectation of being compensated in future. (The difference is that gambling deals with gains, while insurance deals only with restitution and should never form a source of income for the insured — this would be considered fraud.)
Pre-modern insurance practices give us insight into what has and has not worked. These insights can be used in identifying opportunities for insurance products within the blockchain space.
This article will touch on several parallels between (proto-) insurance practices of the pre-modern era and certain patterns we see in the nascent pre-insurance cryptospace.
With an instinctive understanding of risk, human societies created informal systems to assist their communities in times of disaster. One common example is death. Ancient Greeks and Romans formed “benevolent societies” — community guilds that pooled resources in order to ensure the deceased were given a proper burial, and that their families were reasonably cared for.¹¹ This was the ancient precursor to life insurance.
Many ancient societies had insurance guilds. The justification for “benevolent societies” stemmed from religious or ethical, rather than financial considerations — i.e. no profits were taken, as there was no beneficiary of unused funds.
Even in the case of the Great Fire of London, there was an initial attempt to support the rebuilding effort via church collections.
Before insurance companies existed, communities would group together themselves. They would pool resources to protect individual members from risks they all faced. If an unfortunate event occurred the senior members of the community would decide whether to provide assistance or not. All funds raised were used to benefit the members of the community.
— from the Nexus Mutual whitepaper
Although not explicitly utilized for mitigating future uncertainties, we see a similar communal energy in the cryptospace. Crypto gifting has been a practice seemingly since bitcoin’s inception — “the yin to crypto-capitalism’s yang” as Camilla Russo puts it in her article dedicated to the subject:
It’s always been just as much about public goods as about markets, and just as the basis of markets is freely transacting, public goods are about freely sharing.
Another thematically similar topic is social consensus. This can be seen through how the DAO hack was remediated: the informal assembly of community stakeholders conceded to rewrite this historic wrongdoing by forking the network into what are now two separate chains: Ethereum and Ethereum Classic. That is, a solution to the devastation of an unforeseen event emerged from the informal coming together of the community.
Early entrepreneurs learned to manage risk on their own. Marine traders quickly discovered that it was safer to spread goods over a number of vessels rather than risk total loss of a single ship. These entrepreneurs often take the road of self-insurance, making their own independent risk mitigation decisions without using a third-party — especially if such parties have not yet been firmly established.
Just like these early explorers had to grapple with sea storms, fires, pirates, and enemy attacks; so too do the financial explorers of our day — crypto traders and blockchain protocol users — have to grapple with volatility, smart contract bugs, vampire attacks, and MEV bots of the often treacherous cryptospace.
And, just as early explorers self-insured by spreading cargo over multiple vessels, blockchain users manage their own risk — sometimes by diversifying and hedging, though often doing nothing at all. However, there are downsides to DIY risk management including: cost, complexity, and, (ironically) risk.
Left to their own devices, a user of the DeFi ecosystem is left to managing the risks associated with these activities for themselves; the management of which amounts to cost, risk, and difficulty.
— from the UNN Finance whitepaper
Critically, when individuals are managing risk on their own, the Law of Large Numbers cannot be leveraged, resulting in costly risk mitigation:
Insurance economics are driven by diversification. The more individual risks that are pooled together the less capital is required to be confident all claims can be met.
— from the Nexus Mutual whitepaper
Returning to the example of maritime commerce, ancient Athenian lenders advanced loans to individual voyages with extremely high interest rates. Repayments were cancelled if the ship was in any way damaged on the voyage. The maritime loan thus had “an effect like insurance, in that the one who sustains the disaster gains a financial advantage in partial compensation, so that the risk of the venture is spread”. The “unusually high” interest rates differed depending on the time of the year — and thus the perceived riskiness of the voyage — suggesting implicit risk pricing.
In ancient Christian and Jewish societies (and many modern Islamic ones), usury meant the charging of interest of any kind and was considered wrong, or was made illegal. Insurance permitted the circumvention of usury laws.
The usurers of yore were now redeemed as benevolent speculators. Though leasing illusions of power to be fearless was still akin to being usurious in the face of God, “don’t push your luck” protection was the opiate of the masses in 14th century Genoa, when “investment contracts became separate entities and could be contracted with different parties, greatly improving the spreading of risk”¹⁴, and shylocking became invisible trustlines of leverage separating the investor from the insurer as credit from debit.
We see an undeniable analogy to a common DeFi design pattern, where the usury is disintermediated from the usurer: namely, the overcollateralized loan, as seen in some of the biggest projects in the space: Maker/DAI [Total Value Locked (TVL): 5.8B], Compound [TVL: 6.2B], and Synthetix [TVL: 620M], to name a few. Users taking out loans on such protocols provide collateral (in the form of a given cryptocurrency) exceeding the value of the loan itself (issued in the form of a “stablecoin” — a cryptocurrency pegged to some “stable” asset, often USD).
Additionally, these overcollateralized loans often have high deposit/withdrawal fees and/or high interest rates on account of crypto’s volatility. Users may lose all of their collateral in the event that its value breaches a minimum threshold. Users individually mitigate this risk by overpaying their loan in various ways. Meanwhile, overcollateralization requirements limit the appeal of lending crypto assets.
Could we see a greater (capital) efficiency in the cryptospace if risk mitigation was decoupled from individual ‘investment’ protocols, and at the same time underwritten by their trustlessly aggregated reserves under management?
Without seeing patterns in large numbered events with the aid of an advanced sensemaking apparatus, distilling information from unexpected losses was precluded. Modern science exploited quantitative methods for empirical sentience. Statistics provided a logical lens for discerning uncertainty and deducing foresight.
Modern statistics provides a quantitative technology for empirical science; it is a logic and methodology for the measurement of uncertainty and for an examination of the consequences of that uncertainty …
— Stephen Stigler, The History of Statistics: The Measurement of Uncertainty Before 1900
Because insurance prices risk, two essential measurements are needed for the construction of modern-day insurance policies: measurement of uncertainty (risk) and consequences of uncertainty (cost of failure). The burgeoning field of statistics created the tools necessary to create sophisticated insurance policies.
The birth of modern-day insurance arose in parallel with statistics in the 17th and 18th centuries. First manifesting as a subset of mathematics, statistics quickly divorced into its own field of study — with entirely new formalisms and laws.
I outline five major statistical innovations that enabled modern insurance.
- 1650s: Prominent mathematicians Blaise Pascal and Pierre de Fermat exchange letters regarding “games of chance”, and thus inadvertently laid the foundations for a theory of probability, “thereby changing the way scientists and mathematicians viewed uncertainty and risk”¹⁵.
- 1693: Edmund Halley constructed the first life table. He showed how it could be used to calculate the premium amount for (what is now considered) life insurance based on the age of the insuree. This allowed the British government to sell life annuities (with appropriate pricing) for the first time, giving birth to life insurance.¹⁶
- 1713: Although informally written about prior to the 18th century, Jacob Bernoulli proved the Law of Large Numbers in his work Ars Conjectandi (trans: “The Art of Conjecturing”) published in 1713, eight years after his death. From a 2013 article revisiting his work: “From an actuarial point of view, Bernoulli’s law of large numbers is considered to be the cornerstone and explains why and how insurance works… This is the foundation of the functioning of insurance; it is the aim of every insurance company to build sufficiently large and sufficiently homogeneous portfolios which makes the claim ‘predictable’ up to a small shortfall probability.”¹⁷
- 1762: Edward Rowe Mores — considered the father of modern actuarial science — established the Society for Equitable Assurances on Lives and Survivorship, effectively the world’s first mutual insurer. It laid “the framework for scientific insurance practice and development”.¹⁹
- 1812: Pierre-Simon Laplace formalized the Central Limit Theorem — “considered to be the unofficial sovereign of probability theory”²⁰. The theorem roughly states that: in many cases, the average of independent random events tends toward a normal distribution, regardless of whether or not the original events are normally distributed themselves. That is, predictable regularity can emerge even in the most chaotic of randomness.
“I know of scarcely anything so apt to impress the imagination as the wonderful form of cosmic order expressed by [the Central Limit Theorem]. The law would have been personified by the Greeks and deified, if they had known of it […] The huger the mob, and the greater the apparent anarchy, the more perfect is its sway. It is the supreme law of Unreason. Whenever a large sample of chaotic elements are taken in hand and marshalled in the order of their magnitude, an unsuspected and most beautiful form of regularity proves to have been latent all along.”
— Francis Galton
Interestingly, although data had long been collected— e.g. censuses, trading records — the idea to analyze data via probability theory didn’t really emerge in a sophisticated form until the 19th century.
Although builders in the blockchain space are fully equipped with the statistical tools developed in the past few centuries (and in particular, the relatively recent modern advances arising from large data sets and powerful computing machines — i.e. “Big Data”), the space is statistically immature in several ways.
The cryptospace (for lack of a better word) is dominated by programmers and computer scientists. However, what separates cryptocurrency and blockchain from other fields of computer science is game theory. That is, we make assumptions about human incentives and behaviour and extrapolate game theoretic predictions. This results in otherwise vanilla network protocols being much more powerful. This idea is touched on in this presentation at Devcon 5 where Vitalik Buterin asks and answers the question: “What did Satoshi invent?”. (Short answer: cryptoeconomics.)
As cryptoeconomic models become increasingly complex, the need for analyzing the underlying assumptions of such models also increases. This is something I’ve talked about previously, in relation to the recent EIP-1559 proposal:
Just as the emergence of life insurance necessitated life tables (“real” data),the cryptospace needs data to properly assess on-chain risks. And perhaps we can excuse the nonexistence of insurance in the crypto/blockchain space to the lack of data to accurately assess on-chain risks.
We shouldn’t forget to mention the recent emergence of projects precisely looking to use on-chain data to assess risks. Gauntlet Network — and their Risk Score — notably comes to mind.
Utilizing data from centralized and decentralized exchanges combined with on-chain user data, we are able to run simulations directly against protocol smart contracts to estimate market risk.
— from the October release article on the Gauntlet Network Risk Score
Chainalysis — advertised as “The Blockchain Data Platform” — also comes to mind. However, like a lot of analyses of “on-chain” data, their primary focus has been inspecting, identifying, and modelling various activities — primarily illegal activity and general trends— with transactions being the fundamental information source. Risk analysis isn’t and hasn’t been the primary focus of data science efforts in the blockchain space.
Insurance reached a sophisticated form in Enlightenment-era Europe.
The Age of Reason or Enlightenment of the 17th and 18th centuries provided the grounds for accepting actuarial science as a rational means to conduct better business. Insurance, and especially life insurance, resonated with the search for laws, the statistical recording of natural events and the calculation of future developments. Behind this innovation was the conviction that the world, and its possible future states, could be predicted and computed.²⁰
In the late 17th century, Lloyd’s also pioneered the concept of reinsurance, the means by which risk is fractionalized for distribution between multiple insurers. This process was (and is) carried out by a network of accredited brokers. They assimilate the specification of the risk to be underwritten, and match it against a roster of known providers that specialize in the relevant categories of risk. (Usually sub-insurers propose their respective rates basing off that of the “lead insurer”, similar to valuations in venture capital rounds.)
Along with mathematical & statistical formalisms and the industrialization of the industry, the beginnings of modern-day insurance were also demarcated by the increasing specialization of insurance companies — namely into property, business, life, and accident insurance.
What began as a community effort to handle risk effectively has turned into a hostile industry. Incentives — between insurers and those insured — have become misaligned due to the adversely evolving practices of centralized insurance companies.
Disconcertingly, insurance has moved away from community guilds and donation-based remedies for risk management (from early history) to a very adversarial industrial insurance model.
There is a large investment/expenditure in traditional defense of insurance claims (internal and external counsel), which is a cost ultimately passed on to consumers in the form of higher premiums or fewer claim payouts. Specifically, “roughly 35% of insurance premiums are lost due to frictional costs in the system. Only 65% of premiums are returned to customers via claims, the rest is lost in distribution, operational expenses (including regulatory), capital costs and profit.”²²
Because insurance companies resell most of the risk they underwrite, there are often layers of reinsurers who spend significant resources distributing payouts. Many small claims are never brought due to the burdens of going through disputes with insurers. Information is asymmetrically distributed between insurer and insured. The insured aren’t aware of the formulas used to compute insurance premiums and are unable to tell, in most cases, whether they are overpaying for premiums.
The insurance industry has developed over time from a community-based model to an adversarial one where large institutions dominate. It is also inefficient in many areas leading to large frictional costs being borne by customers …
— from the Nexus Mutual whitepaper
In a well-publicized case, after the wreckage of many homes by Hurricane Katrina in 2007, many home owners were unable to receive insurance claims on (what they believed were their insured) homes. It was impossible to determine whether the damage was caused by “water or by wind or by both”²⁰. It was determined, by the insurance companies, that the home damages were mostly caused by water, thus excluding home owners from full coverage.
Prominent trial lawyer, Richard Scruggs, filed a number of lawsuits against these insurance companies shortly afterwards (resulting in numerous favourable settlements but he ultimately dropped his remaining cases following an apparent indictment in a bribery scheme). In response to an interview question asking whether there is a way to make insurance work, Scruggs responded: “There is. And it’s disclosure of what you’re buying […] there [should be] a black box warning on there: ‘this is what it does, this is what you should watch out for’. As opposed to this device which is called the modern insurance policy which no one can interpret or understand.”²⁴
In a more recent example, Hiscox — a business insurance provider — made news for refusing to pay out a large claim for coronavirus-related business interruption losses, something they had previously confirmed would be covered.
Blockchain technology can offer relief to frustrated and disenchanted insurees. Smart contracts can remove administrative inefficiencies and regulatory related costs by providing trust in a more cost-effective way. Since blockchain-based insurance products must, by necessity, provide open and transparent incentive mechanisms, they can create a mutually beneficial service for both insurers and the insured.
Armed with a century of progress in statistics and actuarial science, we can transition back to the original spirit of insurance, where aligned incentives foster community spirit rather than the existing adversarial and unbalanced relationship between individual and large institution.
Inducing social networks to unite around the goal of claiming autonomy over forces of common interest is as much the goal of DAOs as it was that of early insurance societies. Power over probability is well within the purview of power through crypto-economic disintermediation.
If we can collectively better handle risk and uncertainty, the crypto/blockchain space can gain more adoption by way of safety expanding its reaches.
From a very recent article by Pantera Capital:
DeFi [I will add: the crypto/blockchain space as a whole] has not traditionally attracted the risk-averse, so it may come as no surprise that the space has been slow to offer better risk-mitigation solutions. But with the growing popularity of DeFi protocols — and the ever-present bugs and vulnerabilities in their code — the time has come for protection providers to fill the existing gap in coverage paving the way for mainstream adoption.
Risk consumes progress — the “battle of innovation and entrepreneurship against uncertainty still rages.”²⁵
In his paper The History of Insurance: Risk, Uncertainty and Entrepreneurship, professor and economist Pietro Masci shows that, historically, insurance policies and contracts have “evolved over time and over space” tending to respond to the needs of entrepreneurs. Importantly, this need — and the satisfication of this need (via insurance) — is critical to economic activity.
Although there are a handful of insurance-related projects in the space²⁶, the cryptospace as a whole has yet to fully embrace the mitigation of risk with the well-developed tools of insurance. Perhaps it’s time.
- &
- 000
- 7
- 9
- Account
- activities
- Adoption
- ADvantage
- All
- Allowing
- analysis
- appeal
- APT
- around
- Art
- article
- asset
- Assets
- Biggest
- Billion
- Bitcoin
- Black
- black thursday
- blockchain
- bots
- Box
- breaches
- British
- BTC/USD
- bugs
- build
- business
- Buterin
- buy
- Buying
- capital
- cases
- caused
- chainalysis
- charging
- chinese
- church
- claims
- closer
- code
- coming
- Commerce
- Common
- Communities
- community
- Companies
- company
- Compensation
- Compound
- Compute
- computer science
- computing
- construction
- Consumers
- continue
- contract
- contracts
- Costs
- Creating
- credit
- crypto
- crypto traders
- cryptocurrency
- Customers
- data
- Dates
- day
- Deals
- Debt
- decentralized
- Defense
- DeFi
- Demand
- Design
- destroyed
- detail
- Development
- Devices
- DID
- disaster
- discovered
- diversification
- Diy
- donations
- driven
- dropped
- Early
- Economic
- Economics
- ecosystem
- efficiency
- Eggs
- energy
- England
- entrepreneurs
- entrepreneurship
- estimates
- ethereum
- Ethereum Classic
- Europe
- Event
- events
- exchange
- Exchanges
- expanding
- expenses
- exploration
- Face
- Failure
- families
- Fees
- Fields
- financial
- Fire
- First
- first time
- Focus
- food
- form
- Framework
- fraud
- full
- funds
- future
- Gambling
- game
- gap
- ge
- Gemini
- General
- GitHub
- Giving
- GM
- goods
- Government
- GP
- great
- Group
- Grow
- Growing
- Growth
- GV
- harvest
- High
- history
- Home
- How
- hr
- HTTPS
- idea
- Illegal
- Including
- Income
- industrial
- industry
- information
- Infrastructure
- Innovation
- insights
- Institution
- institutions
- insurance
- interest
- Interest Rates
- Interview
- investor
- IT
- knowledge
- large
- Law
- Laws
- Lawsuits
- leading
- learned
- Led
- Legal
- lending
- Leverage
- Limited
- Loans
- London
- Long
- Machines
- Macro
- Mainstream
- mainstream adoption
- major
- Making
- management
- Market
- Markets
- Match
- mathematics
- medium
- Members
- Merchants
- million
- mirror
- MIT
- model
- money
- namely
- names
- network
- networks
- news
- NFT
- numbers
- offer
- offering
- open
- order
- Other
- owners
- pantera capital
- Paper
- Pattern
- Pay
- payment
- Platforms
- Point of View
- policies
- policy
- pool
- power
- prediction
- Predictions
- Premium
- price
- pricing
- Products
- Profit
- projects
- property
- proposal
- propose
- protect
- protection
- proves
- public
- quantitative
- Rates
- records
- relief
- Rent
- Requirements
- Resources
- response
- REST
- Results
- Risk
- risk management
- Roll
- rounds
- Run
- Satoshi
- Science
- scientists
- SEA
- Search
- secondary
- sell
- set
- shops
- Short
- shortages
- shortfall
- small
- smart
- smart contract
- Smart Contracts
- So
- Social
- social networks
- Society
- Solutions
- son
- Space
- spend
- spread
- State
- States
- statistics
- Storm
- Study
- support
- Supreme
- surprise
- system
- Systems
- Tablet
- Technology
- time
- touch
- Traders
- Trading
- Transactions
- trial
- Trust
- us
- USD
- users
- Valuations
- value
- venture
- venture capital
- View
- visualization
- vitalik
- vitalik buterin
- Volatility
- Vulnerabilities
- Watch
- Water
- What is
- WHO
- Wikipedia
- wind
- within
- Work
- world
- year
- years
- youtube