Yesterday, my colleagues Mark Geoghegan and Bernard Goyder outlined the current state of play in the InsurTech space. It’s an excellent read, and I encourage you to make time if you haven’t yet read it.
Their essential thesis is that as the InsurTech space evolves, it has become less based around 'full-stack' companies looking to disrupt the industry and moved towards a partnership model more akin to SaaS (software as a service).
“One trait all of the [successful InsurTech] models have in common is they pose little or no threat to incumbents. Instead, they promise to reduce the expense, acquisition and loss ratios while at the same time improving customer service and retention,” they said.
“At a time when margins are barely covering the cost of capital – and in some places clearly are not – this must be music to the ears of the average industry CEO,” they added.
I think this is essentially correct. With two minor caveats.
The first is that, while they may well be right in terms of what makes for a successful InsurTech companies, I’d argue the outcome for the “average” insurance company is more mixed.
First, advantages from technology widely available to everyone off the shelf do not accrue to the benefit of providers of capital to insurance companies, it accrues to their customers.
But in actuality I think it is more uneven than that.
If technology does ultimately do anything to affect industry structure that could change return on capital, it is to tip the playing field towards rewarding scale. Higher fixed costs on technology become expensive table stakes, but the lower variable costs for each marginal unit of volume theoretically come at a higher margin.
And here’s the problem. This is largely not the case when you are buying technology in the form of SaaS, where costs typically increase somewhat proportionately to scale.
For some, technology will come not only at an increasingly challenging cost of admission, but as an unchanging and uncontrollable component of variable expenses that leaves margins and returns on capital at lacklustre levels.
All of this is to say is that this should not be music to the ears of the “average” CEO. But it probably is music to the ears of the better-run, larger companies, with scope in their margins to think for long term and invest now for compounding advantages down the line.
The second quibble is for me to reiterate my comments published last week on modalities of change.
I have always thought the risk of an external start-up rapidly displacing incumbents is extremely unlikely. The more likely mode of change would just be that bad systems and technology would act as a tax that slowly separates winners and losers through compounding advantages.
However, the exception to this has always been the combination of a far superior technology-driven mousetrap at an incumbent with a track record of insurance know-how and embedded distribution.
And this has always been most likely from personal lines, transmitting over time through osmosis to bigger ticket risks. The high frequency, low severity and homogeneous risks in personal lines simply lend themselves better to the high-res statistical approaches empowered in the age of big data technologies. And here, Progressive and Geico have been among the most likely change agents.
Just last week Progressive’s executives told investors about their aggressive plans to begin expanding into other commercial lines. This comes just weeks after Berkshire Hathaway announced its own foray into small and mid-market commercial.
Both companies have a proven track record of this very type of disruption, and Progressive in particular has proved it can operate across multiple channels and products. In fact, the firm’s rapid ascent to become the number one commercial auto insurer in the country provides a case study for what this can look like. Ditto its nascent steps in homeowners’ insurance.
If I worked in the industry, these announcements would have sent a chill down my spine. Far more than any start-up InsurTech, these are the two companies I would want to compete against least.