Lemonade IPO: A unicorn vomiting a rainbow

On Monday, Lemonade filed its S-1 with the SEC, kicking off a highly anticipated IPO. The move comes after an aborted run at a listing in November 2019, due to perceived weakness for similar “growth/tech” companies such as WeWork.

With markets now (astonishingly) up in the YTD, Lemonade appears to be carpe-ing that diem. Given that sales of securities are typically seen as a negative call on valuations, it should be noted that this serves as another reminder of the perceived surplus of money looking to get into the insurance business.

Lemonade is a company that divides opinion on the best of days. Its champions see it as boldly disrupting a conservative, staid and inertia-prone industry; its critics regard it as a hype factory with little growth runway beyond an impressive renters’ product experience.

Given that its S-1 reads like a unicorn vomited a rainbow into a securities filing, the company seems to be unabashedly targeting the former and making little attempt to convert the latter. One might put it another way and say it is likely targeting non-insurance specialists including generalists, growth and tech investors, rather than investors with specific insurance expertise.

Whether it is successful in that regard remains to be seen. We’d be remiss not to mention the recent public spectacle of the WeWork IPO, a company that shares a key cornerstone investor in SoftBank and a comparable penchant for injecting Silicon Valley buzzwords directly into its veins. However, as a smaller, more niche and less well-known company, Lemonade does not need to convince as many investors to execute its listing.

We certainly found the prospectus… interesting. Heck, we’ll admit it – entertaining. We include some of the most nausea-inducing segments below in our “unicorn walk of shame”.

That said, in a consolidating industry with more public exits than entrants, we are sure the securities industrial complex will be heavily incentivized to pump this thing to high heaven – perhaps even among insurance analysts who should know better. But for our part, despite an extra 203 pages and 135,000 words of disclosure on a company with limited financial transparency to date (its “transparency chronicles” notwithstanding), we find ourselves left with more questions than answers on its key value proposition.

Note, the following makes no judgement on valuation. With no information on expected share prices, we make no pretense to judge the merits of investment. Our thoughts and questions below are focused on questions around the business model as outlined by the company and its expected strategy from here. We see three key questions on its total addressable market (TAM) and unit economics, reinsurance dependence and questionable use of charitable donations.

The business model is premised on expanding its TAM through “graduating” low-cost renters’ policies into bigger-ticket items. Is that a good bet?

The renters’ market is not large – by our understanding, it’s around $3.5bn of 2019 US premiums. Lemonade’s GPW in 2019 of $116mn, the majority of which is renters’ policies, gives it a market share of 3%.

There is clearly scope to grow that. Lemonade’s renters’ product is excellent; it has incredible UI, is millennial and Gen Z friendly, and its use of technology has enabled it to address a low-ticket-price line of business that makes little sense to most traditional carriers given the administration costs relative to available premium. By being a low-cost manufacturer, Lemonade has a long runway of growth in this market. Indeed, it may also grow the size of the pie due to its innovation.

That said, even if Lemonade had a market share in renters comparable to Geico in auto (~14%), it would get to a top line of $500mn.

And that’s unlikely to happen overnight. For context, it has taken the Geico juggernaut a decade to go from 8.5% market share when it launched mobile policies in 2010 to ~14% today, with a well-known consumer brand, competitive advantages and billions of dollars spent on advertising.

Lemonade appears cognizant of this limitation. In its risk factors, it explicitly states something that is more hinted at in its other disclosures.

Our business model is premised on the expectation that a significant number of our users that are renters will continue to retain coverage with us as they move from being renters to homeowners.” [Emphasis added]

This is not a boiler-plate plan to grow through expanding into more lines and product, but an explicit admission that its entire business model depends on sticking the landing from renters to other products.

This is a big bet. In fact, the size of this bet should not be diminished or lost in the otherwise justified praise of the firm’s innovations in the market it has touched to date. There are serious reasons that homeowners’ is a more difficult market to address from a product and competition perspective. We list a few below.

But there are two perhaps more salient points. The first is that Lemonade goes to great pains (the pain was all ours) to describe its missions to “harness technology and social impact to be the world's most loved insurance company”, and that “we believe we are making insurance more delightful”.

More explicitly, it emphatically describes its business model thusly: “At the foundation of our business model is a direct, digital, customer-centric experience that delivers rapid growth and strong retention.

All well and good – with one minor problem. The company’s disclosed retention rates are not that impressive. Its filing stated its year-one and year-two retention rates were 75% and 76% respectively. This does not compare favorably to other personal lines products in auto or home where something in the mid-to-high 80s would be considered good.

And notably, this does not even account for the true full retention once including company-initiated non-renewals – putting the true retention rate at 62% in year one and 71% in year two. These numbers seem more like a high-churn non-standard auto than a book of future “Robinsons” with a lifetime value to deliver a return on that $90mn annual marketing budget.

Another point worth making is its target demographic to date. Some 61% of Lemonade’s premiums come from New York, California and Texas, and much of its early focus was in urban centers. A key issue that does not seem to be addressed in the S-1 is that these urban millennials are not exactly high-propensity home buyers and tend to be long-term renters. In fact, if you were going to target a group of people least likely to buy a home in the near future, these have to be around the top of the list. The company boasts that 70% of its customers are under 35 (close to the median home-buyer age). We wonder what the distribution looks like, and how many have the characteristics of likely graduates in the near term given age, geography and socioeconomics.

If renters’ is the beachhead not the destination, it doesn’t seem a particularly direct route, and one has to wonder at the likely long-term conversion rate. Currently, just 10% of its condo insurance customers “graduated” from renters. If the goal of the business model is to manufacture homeowner PIF, the business has so far organically produced 1,245 of them – or probably around $2mn of annualized premium.

Furthermore, the rate of graduates in its condo business relative to its total customer base appears to be decreasing, not increasing over time as you would expect, given (a) the explicit strategy and (b) the mix of tenants towards longer-term renters from first-timers.

Another way to think about this whole business is that the renters’ product built today is simply an expensive customer-acquisition pipeline – as clearly the goal is to maximize the funnel and worry about profitability later by increased premium per customer.

With 729,000 customers at Q1:19 and an accumulated stockholder deficit to date of $206mn, the company has effectively “spent” $282 of shareholder funds per customer acquired, or $165,000 per “graduated” homeowners policyholder – and was running at a loss per customer in 2019 of ~$170.

You can argue that it also has the intangible assets of its brand and the investment into its technology and capabilities. This could allow it to scale additional volume that improves unit economics over time. But it seems to have an awfully long runway until it can self-fund its growth – by our back-of-the-envelope math, more than doubling its customer base or premium per customer just to break even, assuming all costs are fixed.

For us, the key questions are (a) what level of premium per customer does the company need to get to for positive unit economics, (b) what competitive advantages does the company have outside of its well-targeted niche in renters’ and (c) how much capital does it need to get there given its current cash burn and unprofitable economics? None of this seems super obvious to us at this point.

Second, the firm appears to be building a reinsurance dependence that could be embed risk in its business model.

Lemonade’s S-1 reads a bit as if it were the first teenager to discover sex. The company has long sections around why its use of reinsurance places it in “defiance of industry norms” where underwriting results tend to be volatile and driven by the weather.

“The low cost of capital for reinsurance companies creates an opportunity to share premiums and maintain our gross margins while dramatically reducing our capital requirements through a structure called ‘proportional reinsurance’ (also known as ‘quota share reinsurance’).”

The problem with Lemonade’s model here appears to be reinsurance dependence. The firm’s S-1 explains it uses a 75% Q/S and then manages the risk on the remaining 25% with a combination of per-risk and facultative covers. It estimates that it puts a 3% collar around its target 25% gross margins “in 95 years out of a hundred”.

We’d make two observations.

The first is that reinsurance companies are often willing to “invest” in reinsurance-dependant start-up business models on the premise that, over time, they will have an opportunity to make money. Lemonade has consistently handed loss-making business to reinsurers (inclusive of a 25% ceding commission and underwriting expenses). With this market already tightening more broadly, it is hard not to see this source of capital arbitrage lessening over time as it rolls over.

In fact, we would bet that the entire existence of Lemonade’s operating history to date will come to be seen as a golden age of start-up reinsurance capital through reinsurance. Embedding reinsurance dependency into your business model and feeling smart about it is like buying a one-way ticket for a boat ride to an abandoned desert island – and then negotiating the price of the return leg once you arrive.

The second is likely an issue around increased reinsurance costs if/when it begins to be successful in its efforts to become more of a homeowners’ company.

It should go without saying that this “volatility” Lemonade wants to hedge is not a unique problem, and other companies would undoubtedly have arbitraged reinsurance for fixed 25% no-risk margins if the market provided the capacity at scale.

But as mentioned above, the company’s apparent focus on urban millennials is likely to lead to a lot of risk aggregations by geography – unless it expands its customer base. This is likely to influence cost of reinsurance capital negatively relative to homeowners’ peers with more diversified books of business. (If this is not true, Allstate, if you’re listening, call your reinsurance broker today.)

Third, there seems to be latent risk in Lemonade’s overstated social mission.

Lemonade goes to great lengths to explain its social mission, inclusive of its corporate status, and aims to contribute excess profits above its 25% target margin to customer-chosen charities. The latter is also meant to serve as an incentive against fraud, as the company’s AI-driven claims systems with less human involvement will inevitably invite “testing” of its fraud-detection efficacy and also adverse selection, if not well managed.

We commend this sentiment as an act of writing. It would likely get at least a B+ as an undergraduate moral philosophy essay, and a half-hearted applause at a Davos luncheon. But it seems to have little bearing on the company’s actions to date.

For example, the company has only donated $800,000 to non-profits since its first give-back in 2017. This almost exactly compares to CEO Dan Schreiber’s salary alone over the same period (assuming 2017 and 2018 are roughly in line with 2019). Or put another way, even in 2019 when the bulk of the donations were made ($600,000), each customer contributed just over a dollar each (using average customer) – or less than 1% of their premium paid. This may be in line with the letter of the law due to the product structure, but is likely not what incentivized some customers to choose Lemonade over “greedy insurance companies”.

Ironically, the company has contributed more to its own “Lemonade Foundation” than its customers’ choices for non-profits – accruing a $12.2mn charge in 2019 (albeit in stock rather than cash).

At an ordinary insurance company, this would be no big deal. But at Lemonade, this seems like a business risk. The company’s social mission as a “public benefit corporation”, B-Corp status and use of charity as an incentive against fraud for its automated claims payment, all make it more exposed to criticism of hypocrisy.

This is a danger if a significant part of your value proposition and growth model are based on appealing to your customers’ “values” as a form of marketing and virtue signaling. In an era of “cancel culture”, this seems a non-trivial latent risk related to both growth and potential future margin pressure.

Ultimately the irony may lie in the fact there is inherent charity in Lemonade’s business model. But the charity thus far appears to be from VC’s willingness to subsidize the renters’ insurance of urban millennials. Will public equity markets prove as charitable?

The unicorn walk of shame. We read Lemonade’s S-1 so you don’t have to. Here are the worst parts: