Blockchain: It’s All About Trust
Blockchain is cool technology because it eliminates the need for trust.
Hand me a $5 bill and I can instantly confirm receipt of $5. Nice and simple.
But that changed with the shift to digital payments.
You want to pay me the $5 with a credit card? Ok. But how do I know I’m going to get that $5?
Trusted Third Parties Enable Digital Payments
Trusted third parties – companies like Visa, Chase, and Paypal – will ensure you get that $5.
But it will cost you.
Paypal’s transaction fee illustrates this cost. It charges “2.9% plus $0.30 USD.”
So of that $5, $0.45 goes to Paypal and its partners, leaving you with only $4.55.
Blockchain Eliminates the Need for Trust
Blockchain is cool technology because it eliminates the need for trust.
You don’t need to trust the buyer or pay a trusted third party to ensure you get paid the $5.
How? You can consult the blockchain, a public ledger listing all account balances.
Before a transaction you can use the blockchain will verify that the buyer has the needed funds.
The buyer sends you the funds directly eliminating the need for an (expensive) trusted third party.
And, once the fund have been sent, the blockchain will show updated account balances reflecting the successful transfer or funds.
Trust In the Insurance Industry
The insurance industry is built on trust.
When you purchase insurance, you trust your insurer will be both able and willing to pay valid claims that may arise.
So surely there is a place for blockchain, right?
A Simple Blockchain Insurance Example
Let’s take a look how blockchain could reduce the need for trust in an insurance transaction.
You buy a house. You want to insure it for 100 BTC.
You find an insurer that will offer a blockchain insurance contract. It offers 100 BTC of coverage for 1 BTC a year.
Every day the smart contract checks the ledger:
- If the insurer has 100 BTC set aside to cover potential damage then 1/365 BTC is transferred from your account to the insurer’s.
- If the insurer does not have 100 BTC in its account then the contract is canceled immediately. The insurer keeps the funds it collected for the days it provided insurance. But will no longer be paid as it no longer has the funds required to pay a full loss.
- If you don’t have at least 1/365 BTC in the account each day when the smart contract is scheduled to collect then the contract is void and coverage is canceled.
- If your home is destroyed while the contract is in force, the smart contract automatically takes 100 BTC from the insurer’s account and places it in yours.
Problems With The Insurance Example
There are two major problems with this example.
The first is that the insurance company has to hold 100 BTC capital to cover an expected loss of less than 1 BTC (note: if the expected loss were greater than 1 BTC the insurer would charge more than 1 BTC of premium).
This is much more capital that insurers typically hold and, if insurance companies were forced to hold so much capital against so little premium, insurance rates would have to increase dramatically to allow insurers to earn a reasonable return on their capital.
The second problem is the mechanism for triggering the smart contract to pay a claim. How does the smart contract know a loss has been sustained? How does it know a claim is legitimate? How does it value that claim? The blockchain has no good answers to these questions.
The Capital Requirements Problem
The example shows how the ledger can guarantee the insured gets paid in the event of a loss – the smart contract verifies that 100 BTC are earmarked for my home to ensure it can pay for a full loss.
But setting aside funds to cover the full limit of each policy would come at a very high cost.
In practice insurers set aside only a small fraction of the policy limit.
They don’t need to hold the full limit of each policy because it is highly unlikely that every home they insure will be destroyed in a given year.
What they do is forecast their expected losses, add a little extra to cover a fair amount of unexpected loss, and then set aside enough money to cover both expected and unexpected loss.
For example, in 2016, State Farm collected $39.593 billion in premiums from its policyholders. It ended the year with $87.573 of policyholders’ surplus (policyholders’ surplus is the difference between an insurance company’s assets and liabilities and is a measure of its ability to pay unexpected claims).
This equates to a ratio of 2.21 dollars of policyholders’ surplus for every dollar of premium it collected. Far from the 100:1 ratio in the blockchain insurance example.
By trusting that State Farm will set aside enough money to cover all losses that could reasonably be expected – but not all losses – consumers benefit from lower premiums.
Of course, in theory, an insurer could hold funds needed to pay the full limit of each and every policy.
But, in practice, sitting on capital costs money and those capital costs would be passed along to the customer in the form of dramatically higher rates.
Claim Handling Problem
Let’s say your home is damaged. You submit a claim.
Who determines if the claim is valid? Assuming the claim is valid, who determines the value of the payout?
Smart contracts don’t have great answers for these questions.
Even with smart contracts trust is needed.
You can trust that your insurance company. Or you can trust a third party (e.g. an independent claims adjuster, a mediator, a judge, a jury, etc) to determine the validity and amount of a claim.
In both scenarios trust is needed and the blockchain’s killer feature – the elimination of the need for trust – disappears.
One potential solution is to switch from indemnity triggers (i.e. an insurance company pays you for actual losses sustained) to parametric triggers (i.e. the insurance company pays you if a qualify event occurs). An example of a parametric trigger would be the landfall of a category four hurricane in Florida.
Parametric triggers could more easily be incorporated into smart contracts. But the problem with such triggers is that they introduce basis risk. Basis risk is the risk that the contract’s payout differs from the loss you suffer.
Your home could be unscathed in Orlando and you receive a payout when a category four hurricane makes landfall in Miami. Conversely, your home could be completely destroyed by a category three hurricane and you receive no payout.
Real World Insurance Blockchain Examples
Recently a friend alerted me to two insurance blockchain projects he came across.
I was cautiously optimistic I would see these projects employing novel solutions to overcome this trust issues. Unfortunately, neither replaces the need for trust in insurance transactions.
With EtherRisc you are both the insured and the insurer.
You join a small local group of up to 12 members. Each small group is a member of the larger EtherRisc network to share premiums and losses.
For every $1 per year in premium that you pay in, you are entitled to a payout of $100 if you sustain a covered loss.
If you have a claim, the validity of your claim will be validated by your “local spokesperson,” a designated member of your local group.
But EtherRisc doesn’t solve either of the trust issues.
Insureds cannot trust that there will be sufficient funds to payout valid claims. The model is calibrated to pay out in claims all premiums every year. So if it collects 1000 BTC in premium, one can expect 1000 BTC in losses.
Actual losses sustained are rarely if ever as expected. So, even if one were to assume the EtherRisc risk assessment is accurate, in half of all years the losses would be greater than expected and there would not be enough money to pay claims.
And insureds cannot trust that claim will be handled properly.
A member of your local 12 person group – your “local spokesperson” – validates claims.
It wouldn’t take long before people realized they could collude with members of their local group to enrich themselves at the expense of other EtherRisc groups.
And this would happen. A popular Bedouin proverb highlights this aspect of human nature. It states: “I am against my brother, my brother and I are against my cousin, my cousin and I are against the stranger”
You and your brother and your cousin are members of your EtherRisc local group. Other local groups are the strangers. You will seek to enrich yourself and your family at the expense of strangers.
Here’s how it would work…
You file a fraudulent claim, your local spokesperson approves it, you get your 100x payment, you keep some for yourself and you share some of the proceeds with the local spokesperson and other members of your local group.
Other members of your local group note your success and do the same thing. Other local groups take notice and realize they have two choices:
- they can sit around and do nothing knowing that the system will be drained by fraudulent claim making it unlikely that there will be adequate funds to pay them if they have a legitimate claim.
- They can get in on the action and start filing fraudulent claims.
Some groups will be honest and sit back while other groups raid the system. But some will get in on the action.
This will trigger a “run on the bank” type situation and the system will be drained of all funds.
B3i is a blockchain technology consortium formed by a group of insurance companies. The goal of the initiative is to make reinsurance claims payments occur more quickly.
Per a CoinDesk interview with B3i co-founder Paul Meeusen, B3i “contracts are being designed to settle transactions in the fiat currencies accepted by the members. But in the future, [Meeusen would] like to be able to accept cryptocurrencies as a way to settle balances between counterparties with longer-term contracts.”
In the short term, when contracts are not automatically settled by payment on the ledger, the the trust issue is not solved.
These contracts simply generate invoices when payments are due. They do not ensure that the billed party has the funds. Nor do they guarantee that, if the billed party does have the funds, it is willing to pay up.
Longer term they could switch to smart contracts that automatically settle with crypocurrency payment. But, in this scenario, the capital requirements problem re-emerges making blockchain insurance solutions many times more expensive than traditional solutions.
The Problem: Insurance Companies Are Too Trustworthy
Blockchain’s main benefit is its ability to eliminate the need for trust. This makes its ideal applications low trust scenarios.
The problem with the insurance industry applications is that, for the most part, insurers have demonstrated that they can be trusted to (1) have ample funds to pay claims and (2) to properly handle claims.
The Insurance Information Institute reports the net cost of the 10 largest US insurer insolvencies at $5.03 billion. The data goes back to the mid-1980s meaning the cost of insolvencies has averaged roughly $150 million per year over the past 30 years.
$150 million may seem like a lot but, given the size of the insurance industry – the US insurance premiums exceeded $1.15 trillion in 2015 – it’s not all that much. And one must realize that the cost of these insolvencies was largely shouldered by other insurance companies via payments to insurance guaranty funds. So it is extremely rare that an insurance consumer does not have a claim paid because an insurance company lacks the ability to do so.
Insurance companies have also proven trustworthy when it comes to their willingness to pay valid claims.
For example, in 2016 the National Association of Insurance Commissioners reports that there were 27,670 complaints on $214.5 billion of personal auto insurance premiums. That works out to $7.75 million of premium per complaint. That same year there was one complaint for every $10 million of homeowners insurance premium.
When you’re insured by a name brand insurance company, a company like State Farm, MetLife, Liberty Mutual, Allstate, GEICO, Progressive, Travelers, a company with a nearly 100 year track record of willingness and ability to pay claims you can trust that you will be treated correctly. This trust is incredibly valuable.
Blockchain solutions that don’t required trust are possible. But they destroy the value insurers have created with their long track record of being trustworthy and thus come at a very high cost.