I recently met a friend who works at a traditional UK insurer for a beer. He has spent the last few years telling me that insurtech is overhyped and overvalued. So he was fairly smug that his predictions have been “proven right”.

The share prices of the first wave of insurtechs to go public have fallen by 80–95% from peak. Is this the end for insurtech, and can traditional insurers go back to pen, paper and brokers pressing the flesh?

No prizes for guessing that I believe this would be highly premature.

We are still in the early days of insurtech and there are massive global businesses to be built.

First off — why have the share prices of the public insurtechs been hit so hard? The biggest challenge they have faced is that their loss ratios are too high. Lemonade and Root are still at ~90% gross loss ratios. For every $10 in premiums they pay out $9 in claims. It is impossible to build a good business on that margin. A strong traditional insurer would have loss ratios of more like 60%. To be fair the public insurtechs have been reducing loss ratio significantly over time through scale and customer retention. In 2021 the market bought into this trend, but the more sceptical public market of 2022 demands to see actual data. Matteo Carbone gives a good summary here.

The second problem is unit economics. Customer acquisition costs for these insurers were high, retention not high enough, and so LTV:CAC was low and payback time on CAC was five years or more. See this analysis for more.

The third problem is growth. The insurtechs have pushed hard to improve loss ratio and unit economics but this has been at the expense of growth, which has been mostly flat over the last year. Hard to be valued as a growth stock when you aren’t growing.

Insurance is a trillion dollar industry with many profitable incumbents, and tech can bring dramatic efficiencies

The question to ask is whether these challenges are endemic to insurtech, or whether they are specific to these three companies. My take is much more the latter. Insurance is a hard sector to enter with plenty of pitfalls, but it is a trillion-dollar industry with many profitable incumbents, and tech can bring dramatic efficiencies. I stand by my article from 2015 on the challenges faced by traditional insurers and how tech companies can disrupt them.

There are a number of private insurtech businesses reaching real scale: Zego, ManyPets, Next Insurance, Ethos Life, Alan, Atbay, Coalition etc. Having seen some of their numbers I can say that their loss ratios are good, their unit economics are sensible and they are growing strongly. They are achieving this through great execution, and some combination of the following:

  • Addressing new markets which didn’t exist before, such as cybersecurity
  • Using telematics and novel data to meaningfully reduce losses and fraud
  • Targeting underserved segments such as SMEs and the self-employed
  • Operating in ‘easier’ lines of business where there is less competition and loss ratios aren’t as tight as in home and motor, such as pet insurance

At the same time I have also seen many insurtech startups which are either struggling to find product market fit, or are growing fast but with deteriorating loss ratios, unclear long term differentiation, or other issues below the surface. I have spent a lot of the last few years sitting on my hands.

Balderton’s only major investment in the sector remains Zego. From the list above they achieve real underwriting improvements with telematics and target underserved growing markets such as self-employed drivers. They also have great loss ratios, retention and unit economics.

So what have we learnt so far in insurtech?

  • Underwriting is crucial. As an insurtech you start with the disadvantage of not having a base of loyal good customers. Where are you using tech to give you a real edge over incumbents? But also do you have enough actuarial talent and data to complement this?
  • Telematics (real time data) works. Whether in motor or health insurance, telematics data has repeatedly demonstrated the ability to reduce loss ratios. As data collection becomes cheaper and universal expect telematics to become widespread across many insurance lines.
  • GWP is not ARR, and insurtechs shouldn’t be valued in this way. Typical margins on GWP (after losses and service) are in the 20% range, and insurance is never going to have 100% or more net revenue retention.
  • An insurance license is not something to be taken lightly, and is not the answer for every insurtech. Cash requirements and regulatory overhead are high.But it does given you the freedom to really improve underwriting.
  • Your unit economics have to work. This starts with a good loss ratio, and requires you to have happy clients who stay with you. It also requires you to have a good customer acquisition cost, which in insurance is never easy.
  • Automation will bring down expense ratios dramatically, but only at scale.
  • Brokers are very hard to dislodge. Many startups have tried to bypass brokers without success. Some of the large private insurtechs above have ended up building a successful broker channel.
  • If the barriers to entry are low expect competition to intensify quickly. See for example French home insurance where Luko were quickly joined by Lemonade, Lovys, Leocare and others.
  • Disrupting a large sector takes time. Disruptors start off being worse than incumbents in all respects other than one really crucial one. Look at the early days of SaaS or Fintech for examples.

I would love to make an investment in insurtech in 2022. I sincerely hope that great founders haven’t been put off the sector by the share price performance of a few companies.

__ 

*This post doesn’t at all cover companies selling tech to insurers — topic for another post some time