Driving the bull case for Auto-Insurance

The US auto insurance market is nearly the same size as US car sales market ($309bn vs $545bn)[1]. In fact, auto insurance is the second largest auto related market in the US; 14x bigger than the Car Rental Market ($22bn), 17x bigger than Car Auto parts market ($18bn) and 5x bigger than Car repairs market ($66bn). Yet despite this huge size, the market is expected to continue to grow at a not-so-shabby rate of 3%pa over the next decade[2].

Is car ownership at risk?

On the contrary to what has been expected over the last decade, car ownership in the US is continuing to increase. 93% of households last year said they had access to at least one vehicle in the US and that number has been trending higher over the last decade[3]. Surprisingly, a large percentage of this growth has come in cities where ride sharing services such as Lyft and Uber operate, which studies put down to increasing ownership by drivers (more than offsetting decreasing ownership by riders)[4].

Whilst on paper, car ownership could see some headwinds as younger consumers delay buying a car in favor of public transport to work/ride sharing, for now this is being offset by cheaper financing costs for lease vehicles and the very large percentage of the US population which cannot functionally survive without a car[5]. For instance, millennials still account for 12% of all vehicles sold on the road today.

COVID-19[6]

We are seeing the trend away from car ownership being slowed down significantly by COVID19. The percentage of US consumers who think car ownership is necessary has shot up 14% in the last year (with 60% of those preferring to buy from a dealership).

This is despite ridesharing available to more than 20% of the US population (up from 7% in 2015)[7]. In 2020, ridesharing has experienced a massive hit to demand, with 65-70% of users between the ages of 18-24 having stopped their usage of these services due to COVID-19. Resultingly, Uber has had to lay-off 25% of its workforce (Lyft laid off circa 20%)[8].

Public transportation has experienced similar headwinds although it started from a much lower base. US public transportation has historically had very low rates of adoption (outside of a few major cities) with 45% of millennials (and 38% of general public) having reported to have never used public transportation.

US Public Transportation has a very low rate of adoption

But in the cities where public transportation started at higher levels of adoption (as a percentage of total transport) such as in New York, Jersey City and Philadelphia, we are seeing usage down significantly. In fact, across the US (according to Apple Mobility Data), public transportation remains the hardest hit mode of transport due to COVID-19 and is down 61% on average (compared to driving being down only 20%).

The outlook

What this shows is that car ownership is not experiencing a decline and if anything, the stickiness of consumer decisions to move away from public transport and ride sharing, should continue to provide a steady rate of growth for the auto insurance market over the coming years.

Case for disruption in the market

The insurance market has not changed much over the last 50 years. Put simply, the business operates on the law of large numbers – insurers collect smalls bit of data (like age and driving history), set you a premium and then hope for the majority, that they don’t get into an accident (and on the whole they do ok with 64% of premiums collected being handed out in claims)[1].

That being said, there are underlying problems in this market:

  • Inefficient pricing

The current auto insurance market experiences an inherent (economic) inefficiency arising from a lack of data distinguishing between good drivers and bad drivers. 35% of drivers put in more than more than half of total passenger miles and cause more than half of insurance losses[2]. However, a lack of data about which drivers are causing these losses means that premiums increase for all drivers, to offset these losses (and are not targeted at just the bad drivers). This effectively leads to 65% of drivers overpaying for auto insurance in order to subsidize the bad drivers.

  • Incentive Problem

There is an inherent incentive problem when trying to deal with pricing in the auto insurance market.

Firstly, the existing insurance market relies on a linear relationship between miles driven and insurance losses i.e., the less passengers drive, the less losses an insurance company is likely to experience as a result. In a market which has long been predicted (see above) to experience declining passenger car mile usage due to more public transportation and ride sharing, it becomes harder for companies to overcome the internal headaches/investments to change their pricing models.

Secondly, given that no US carrier currently has more than 20% market share, there is a reluctance on behalf of the larger carriers in particular, to fundamentally disrupt their business (via telematics-based pricing) at the detriment of conceding market share. This is a classic innovators dilemma as the larger companies have focused on sustaining technology/small scale disruptions to meet their customer needs today, instead of overhauling their business with large scale disruptive technologies for users in the future.

Both of these are evidenced by the relatively low penetration of finance technology and advanced data analytics in the US auto insurance today[3].

No carrier has more than 20% market share in US Auto-Insurance
Fintech Adoption has room to grow

The future of the market: Data Analytics

According to a comprehensive auto insurance study by Mckinsey, the auto insurance market in 2030 will belong to the insurers who have the best pricing capabilities[1]. In order to achieve this, insurance companies will need to adopt advanced data analytics and data collection methods (so things like: monitoring real time driving habits, integrating this with information about consumer behavior and then delineating risk to price policies). In fact, by 2030, they expect that over 90% of policies will be automatically priced and for the top tier of companies that can get this right, this will result in operational savings of over 30%.

From a compiled risk of disrupters in the insurance space, we can see this is getting some traction. Approximately 20% of new insurance technology companies are focusing on data analytics and Machine Learning models to improve pricing capabilities in the insurance market.

20% of Insure-Techs focused on Data/Machine Learning

Compiling a list of the major insurance technology disrupters, filtered by those with a market cap of minimum $1bn, there are only three companies dedicated to taking on this challenge in the auto insurance market; Lemonade, Root and Metromile. They are all currently publicly traded (Metromile is in the process of being taken public by SPAC).

How do they rank on key Metrics?

At surface level, Lemonade, Root and Metromile are similar. They all focus on data analytics, are rolling out across the US and intend to or have already expanded to insurance verticals (such as Renters and Pet Insurance). The things I like about each are that: Metromile boasts a pay per mile insurance policy, Root uses your smartphone (instead of a dongle) to track driving behaviors and Lemonade has a fully automated claims processing function (which I have personally used before).

At closer look however, Metromile comes out ahead of the competition on the metrics that will be most important. That is, Metromile is better than Lemonade or Root at (1) pricing insurance risk (i.e. lower Loss ratio), (2) retaining customers on its platform (especially important for insurance cross selling which decreases Loss Ratios by close to 15%!) and (3) has significantly higher life time value per customer in relation to the cost of acquiring these customers.

For me the most striking thing about this is that Metromile is the newest of the three businesses (originated in 2019), with the smallest customer base and in a sector where there are natural economies of scale (the more customers you have, the more data you collect, the better your pricing models).

Conclusion

The auto insurance market is a stable market and ironically, this is what will drive investor returns. The relatively sluggish pace of growth in the overall market (of 3%) has left room for tech disrupters to take a slice of the pie, which will likely come at the expense of the bigger and more inefficient providers of auto insurance. Whilst I don’t expect the industry to outperform some of the more exciting sectors that I have covered in other blog posts, I do expect the market share of tech rivals such as Lemonade, Root and Metromile to expand as they beat out these larger rivals. For now, Metromile is my pick.

Verdict: Moderately Bullish
Timeframe: 1-3 Years

[1] https://www.mckinsey.com/industries/financial-services/our-insights/insurance-productivity-2030-reimagining-the-insurer-for-the-future


[1] https://assets.metromile.com/wp-content/uploads/2020/11/24120556/Ext-Investor-Preso-vFinal.pdf

[2] https://assets.metromile.com/wp-content/uploads/2020/12/16214939/12.16.2020-Metromile-Financial-update-supplement.pdf

[3] https://www.mordorintelligence.com/industry-reports/united-sates-motor-insurance-market


[1] https://www.ibisworld.com/united-states/market-research-reports/automobile-insurance-industry/

[2] https://www.mordorintelligence.com/industry-reports/united-sates-motor-insurance-market

[3] https://www.thezebra.com/resources/research/car-ownership-statistics/

[4] https://www.newscientist.com/article/2264144-uber-and-lyft-operating-in-us-cities-linked-to-rises-in-car-ownership/

[5] https://investorplace.com/2019/04/4-charts-car-ownership-over/

[6] https://www.thezebra.com/resources/research/car-ownership-statistics/#own-lease

[7] https://investorplace.com/2019/04/4-charts-car-ownership-over/

[8] https://abc7news.com/ridesharing-apps-covid-19-rideshare-uber-lyft/6197137/

Leave a Comment

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s