Dear Peter Diamandis:
I was pleased to see your post, “5 Tech Forces That Will Change Insurance for Good,” and your participation with Lemonade. I have followed your newsletter for years and have previously referenced your thoughts in this blog. I am offering a few comments and describing my admittedly foggy view of the future.
A bit of history
As you likely know, ‘modern’ insurance dates back to London in the late 17th century, where Edward Lloyd ran a coffee house that attracted merchants and ship owners. To reduce the individual risk of losing a ship and its cargo, Lloyd’s customers developed a scheme where a larger group, composed primarily by other traders and investors, shared the risk.
Thta sounds like peer-to-peer (P2P) to me.
So, why are we re-inventing P2P?
Lloyd’s coffee house became the Lloyd’s of London insurance market and moved away from the P2P model.
But P2P now has renewed appeal once again. As you noted in your post, Peter:
“Imagine finding a group of peers who you trust and can vouch for and coming together as a group to self-insure.
“You skip the centralized, expensive middleman insurance carrier – instead, a technology stack (app, database, AI-bot) manages a decentralized network of people who pay premiums and file claims that the group approves.”
Lemonade – and others – are developing a P2P model similar in many respects to the coffee house model, albeit with much more interesting and powerful tools.
So why did insurance lose the P2P model and become more complex and confrontational? And is a new version of P2P extensible?
A colleague of mine has a straightforward theory:
“Insurance is dead simple: Money in, money out, and profit coming from the float. But trained professionals have worked for centuries to make it complex.”
There’s some of that, but it goes a bit deeper. Here are a few thoughts.
Bad guys know about technology too …
You noted in your post: “Fraud consumes as much as 38% of all the money in the traditional insurance system, inflating premiums by $1,300 and making the claims process protracted and unpleasant.”
That might be a bit high. I tend to use 10%-15% for North America. The FBI web resource on insurance fraud says: “The total cost of insurance fraud (non-health insurance) is estimated to be more than $40 billion per year. That means Insurance Fraud costs the average U.S. family between $400 and $700 per year in the form of increased premiums.”
Regardless, fraud is a big deal. And I buy your argument that digitizing the process end-to-end and adding machine learning would reduce incidence of fraud and improve the settlement process.
And a number of organizations – insurers, claims experts, and technology suppliers – are working diligently on this to ensure that risks are lowered (using sensors, as you noted in your post) and that insureds and suppliers – health care, repair, replacement, litigation, etc. – follow protocols. This is a constant challenge as the bad guys consistently find new paths to seek subvert protection.
Could P2P manage this complexity?
The elephant in the room: Disrupting auto disrupts more than auto
Peter, you are spot-on in saying that “the car insurance industry is about to get disrupted in a huge way” (e.g., see our post on “The Future of Disruption: Will Telematics Uberize UBI?“).
But the impact will extend beyond this one line of business.
Experts agree that automobile insurance premiums will decrease over time. As insurers (especially publicly traded companies) lose revenue in auto, they are likely to become more competitive in other lines: home insurance, commercial property, commercial liability, cyberrisk, etc., potentially driving these rates lower.
The North American insurance industry has been in a soft market for some time and the competition will only exacerbate the situation.
Will this cause a rationalization? Will this cause a buying spree by insurers for InsureTech start ups? (Or …?)
The point is …
The P2P model worked well in the early days with a limited group and will likely work with tightly connected peer groups. But as we seek growth with larger groups and more complex risk scenarios, it is unlikely that P2P will work effectively in the short- to mid- term. Organizations will seek tech opportunities to reduce costs, but will not radically challenge busienss models.
Of course, this is just my view, Peter. There may be other constructs that can work. I would certainly welcome more of your thoughts and those of other readers.
We are living in interesting times.