Lemonade (Ticker Symbol: LMND) is Manhattan headquatered public-benefit company in the insurance business and is disrupting the insurance business by being able to keep their adminstrative costs to minimal by using AI to evaluate the risk premium for the insurances and claims. They are currently operating in US and Germany. They have offices in New York, Amsterdam and Israel. Lemonade currently offers Homeowners, Renters, and Pet Health Insurance in US, France, Netherlands and Germany.
Management Team: CEO Daniel Schreiber; COO Shai Wininger; Chief Behavioral Officer Prof. Dan Ariely; CFO Tim Bixby; Chief Insurance Officer John Peters,
Lemonade’s Business Model Canvas:
- Incubment Legacy Insurance Companies AXA, Geicko, AIG, Allianz etc.
- New Start-ups in Insurance Business like Root insurance (Fair car insurance for good drivers | Root Insurance (joinroot.com), Metromile( Pay Per Mile Car Insurance – Metromile)
Arguments in favor of Lemonade
Loss Ratio Decreases
Lemonade is doing this in part by changing the normalised policy holders attitude towards the insurance company and a AI system getting better at detecting fraudulent claims. Traditionally the attitude towards Insurance company, because of conflict of interest or perception of conflict of interest, has led to policy holder feeling acceptable to make false or fraudalent claims.
What is Loss Ratio?
Loss ratio is percentage of claims the insurance company is paying out from the total amount of risk premium collected. To know more about the insurance industry read the post: Industry Analysis: Insurance.
Lemonade is trying to improve this attitude towards the insurance company by removing itself from the conflict of interest by not having any incentive in delaying or denying the insurance claims. Lemonade charges a flat fee of 25% on the premium and from the remaining 75% a part is used for reinsurance and rest is added to the premium float. If there is any money left over after the end of the insurance term the money is given to the charity of choice of the policy holder.
Lemonade has been able to offer insurance at price points which has made insurance an impulsive/unplanned buy. The price of entry level insurance is as low as couple of coffee at Starbucks or a monthly fee for a Spotify.
Almost 30% of all the claims to Lemonade are handled by the AI bots. I believe the ease of the claim process is delighting the customer and the fact that leftover premiums is going to the charity of the policy holders’ choice means that they are less inclined to make false claims.
On Clearsurance.com, majority of the excellent reviews seems to be of customers happy with the onboarding process or small ticket claims. Lemonade also tops their list of best renter insurance companies in US for 2020. Almost all of the negative feedback has been from people who have been denied claims. On Trustpilot, the total number of reviews is only 25 reviews have been posted but majority of them have rated Lemonade Poor or Bad. And again most of the bad ratings are from people who have had their claims delayed or rejected.
Scalability of Lemonade’s Business Model
As the CEO Daniel Schrieber has talked about on multiple occassion about Lemonade being built on “Digital Substrate” that is harnessing the power of Disruptive Innovation Technologies including Artificial Intelligence, Mobile Connected Devices and Cloud Computing.
At Lemonade, AI bots replace the Brokers and Actuaries. AI is being used not just on the front end with interactions with the customers but also in the backend, for eg. cancelling of existing insurances, the AI bot interfacing the customer asks if they have an existing insurance and if the answer is yes the task is handed over to a bot in the backend which checks the database to verify and send the cancellation request to the insurance company in the prescribed format and mode for that particular insurance company. This gives Lemonade the possibility to scale geographically as soon as they have the regulatory approvals without any increase in staff and elastic IT infrastructure.
Since most of the activities are automated using AI, the cost for handling small claims is minimal compared that to legacy insurance providers who would incur same level of administrative cost “bureaucracy of the claim” irresptive of the size of the claim.
Traditional Insurance companies does not have the same granularity, resolution or data points about the customers as Lemonade has and that allows them to get to a much more customer specific risk premium or price to risk. During the onboarding process the customer is asked 13 questions from which 1600 data points are created. Compare this to the 20-50 questions that the broker for a legacy insurance company asks a potential insurance customer and calculates the risk premium based on this 20-50 data points.
In this section, I have gathered from the internet some of the counter arguments for not investing in Lemonade and also my views on them.
- Low Moat, Reinsurance, Regulatory Constraints
Ark Invest (34:25) has commented that Lemonade is just an legacy insurance company wrapped up with a nice UI and that Lemonade have very low moat compared to other insurtech companies which have more unique technology stacks.
Low Moat: In my view most technology companies cannot have a lasting moat and the only way to keep themselves ahead of the curve is by innovation or by pivoting. Lemonade has been in the business for only five year now (incorporated April, 2015) and they have managed to grow their customer base to a million (end of 2020). But it remains to be seen how well they can execute going forward in the face of other challengers. Also when it comes to other insurance products it remains to be seen what unique technologies they add to their stack.
Reinsurance in my view is more a hedge tool for a small insurance company like Lemonade and not necessarily a liability. Those reinsurance premiums can be optimized when they have been in operations and have much more granularity on the risks they need to cover.
Regulatory Constraints: As I understood in US market Lemonade needs regulatory approval from every state they would like to operate in and so regulatory constraints are for real. In Europe, they are already selling their insurance products in France, Germany and Netherlands. Dan Schrieber mentioned in one of the interviews that its much more easier to get regulatory approvals from other EU countries as Lemonade already has an approval from more that one EU country already.
- Insurance is mature and extremely competitive. Their profit margin is limited.
In rental and homeowners insurances, it seems Lemonade has been successful in lowering the operating costs by fully automating the onboarding process and handling 33% of the claims by using AI bots. But again it remains to be seen how they execute when they expand their product portfolio to auto or life for example.
- Although, the insurance market is huge, Lemonade is targeting quite a small subset and because Lemonade is focused on a niche part of the market, the policy size is too small; As their model is entirely reliant on the availability and pricing of reinsurance as their source of capital, they are a relatively small buyer and will be at a major disadvantage position.
Lemonade’s CFO Tim Bixby, mentions that Lemonade is planning to expand into Life Insurance product somewhere in Q1, 2021, involved probably in the onboarding experience for the customer and the claims process being handled by the partner (for regulatory reasons and getting faster to market). Tim also talks about Lemonade expanding into Auto Insurance products as well some time in the future. Based on these comments and the fact that they have been expanding to newer markets geographically is, in my opinion hugely expanding TAM for Lemonade products. As the figure from the latest Shareholder Letter shows, the customer growth rates and increase in IFP has been outstanding.
Regarding the Lemonade’s planned expansion in the Auto Insurance, I am not sure how they plan to collect those data points that are needed for calculating the risk premium and also how much time and data points the AI algorithms will need to be able to automatically process the Auto claims. I was listening to the Dan Manges, CTO & Co-Founder of Root Insurance describing how they ask their potential customers to drive around with mobile having their app activated for two weeks before quoting them a premium based on factors like hard braking, accelerations. Dan also talks about not giving a quote at all if they are not able to offer a better deal than competition for riskier drivers. The same remains to be seen with Life and Health Insurance Products. There are two scenarios either Lemonade will have to invest in innovating a way to handle these new insurance products & keep delighting the customer or use the more traditional way of assessing the risk & not have any differentiation.
- Lemonade’s loss ratios still look way to high compare to other insurance companies.
I have not been able to compare Lemonade’s loss ratios to other insurance companies. But looking at the NAICs industry-wide net loss ratios for U.S. Property & Casualty Insurance Industry for the period 2010-2019, it has been varying between 67-79,5%. And in combination with declining trends in net loss ratio for Lemonade over every passing quarter leads me to believe that as the ML stabilizes with newer data points the net loss ratio would be at the least be inline with the industry if not better in future.
- A path to profitability is not visible, and they have scaled their model too quickly (like Uber; burning money). The stock’s P/S is about 77 makes it strict no buy.
It is still early days for Lemonade in my view and scaling in new markets and products has meant they had to spend at this stage. But there is some solace in the fact that trends for Net Loss Ratio and Total revenue is positive.
- Is Lemonade really disruptive in their usage of AI ? What is stopping others to catch up with Lemonade ?
I heard Dan Schriber talking about having a proprierty AI algorithm and data points which gives Lemonade an edge over the other start-ups and also the fact that they have already built up quite a bit of data to leverage from their more than a million customers. But yes, there is nothing stopping the competition from developing better AI and getting more data & refinements of their AI in product areas where Lemonade doesn’t has insurance products yet.
Disclosure: I have an explorative position in the company and may or may not increase my position within the coming weeks and months. See my complete portfolio here. Please read the disclaimer before acting on the information provided on the website. Some of the links in the post may be affiliate links.
- Lemonade Inc. – Hey investors, welcome home!
- Barclays Global Technology, Media and Telecommunications Conference
- Dan Manges, CTO & Co-Founder of Root – YouTube
- Lemonade is 5 Years Old! Here’s Our Strategy and Results to Date – YouTube
- The Future of Insurance | Lemonade @ DIA Munich 2017 – YouTube
- Lemonade, Inc. Q3 2020 Earnings Conference Call Letter to Shareholders
- Lemonade Quarterly Investor Metrics Q2 2020
- Business Model Generation: A Handbook for Visionaries, Game Changers, and Challengers: Osterwalder, Alexander, Pigneur, Yves: Amazon.se: Books
- August mARKet Update Webinar – ARK Invest (ark-invest.com)