Insurance tech Lemonade (LMND -4.70%) posted its quarterly report on Nov. 8 and hosted an investor day a week later. The company had plenty to say, but what management said likely will only change a few investors' minds.
Here is one big reason bears are remaining cautious, and one reason that bulls were excited about what the company said.
1. One big reason to be cautious
The loss ratio -- losses paid out divided by premiums earned -- is one of the most crucial measurements for assessing an insurance company's health and profitability. One of management's goals is to maintain its gross loss ratio below 75%, giving the company enough money to pay expenses and set aside reserves.
The company went public with a loss ratio of 67%. But it was not long before the problems began. The Texas freeze in February 2021 sent the gross loss ratio skyrocketing to 121%. Since the freeze, Lemonade has only been able to bring its loss ratio below 75% one time, which concerns investors.
Its third-quarter 2022 report showed a gross loss ratio of 94%, up 17% over the previous year. Management blamed this quarter's high loss ratio on the acquisition of car insurance company Metromile, which it expects to add 3 to 5 percentage points to its loss ratios for the next few quarters. And the company suffered losses from Hurricane Ian blasting Florida. It has also blamed its substandard loss ratio on inflation in the past.
But investors tire of these excuses, a huge reason that the stock is down 53% year to date, and the bears remain wary of this company.
2. Reasons to be optimistic
During Lemonade's investor day, CEO Daniel Schreiber explained how the company uses artificial intelligence (AI) to bring the gross loss ratio below 75% and reduce the expense ratio to achieve profitability.
Image source: Lemonade.
The company has several forms of AI, which it uses to effectively replace humans, reducing costs within the loss and expense ratios. For instance, the company relies upon AI Cooper to handle most DevOps engineering tasks. Cooper can do almost everything except write code. It monitors servers, spins up virtual environments, and puts more than 12,000 software builds into production yearly, saving around 10,000 human-developer hours.
AI Maya performs marketing functions, selling 98% of its policies. Maya removes humans from most of the sales process so Lemonade doesn't have to pay commissioned sales agents. Instead, its machine-learning models generate predictions for how likely a customer is to churn, the potential for the customer to make claims, the severity of claims, and how likely they will be to buy additional products. These models also make predictions on every marketing campaign, every product, and every geography. Today, Maya governs 86% of the company's marketing spending.
CX.ai handles about one-third of all customer requests, including scheduling items, adding spouses, changing addresses, adding coverages, and more, with a customer satisfaction score of 4.5 out of 5.
AI Jim is a massive part of lowering the loss ratio by handling 98% of initial claims reports. And it can process 45% of those claims from start to finish without human assistance. AI Jim also taps into algorithms that can detect fraud better than humans. And fraud is a significant chunk of the loss ratio.
Image source: Lemonade.
Lemonade enjoys two competitive advantages from building a digital-first AI in the cloud.
First, it will be difficult for large legacy insurers to re-create what Lemonade does because the company's technology was built from the ground up as one integrated brain that shares data -- a complex, lengthy, and expensive process to build. And even when a company manages to create a platform similar to Lemonade's, it will need to catch up in collecting data sets for that platform. The legacy data sets that older insurance companies have collected for decades are useless with modern AI.
Image source: Lemonade.
Second, the more data the AI has to process, the more accurate the predictive models are. And these models should produce more-accurate projections than competitors do.
According to management, its AI model forecast that the policies it wrote in the first quarter of 2021 will produce an 85.8% loss ratio. But the policies that Lemonade wrote in the second quarter of 2022 are predicted to have a 60.6% loss ratio. As a result, it is now confident that the loss ratio will go below 75% and the company will achieve profitability because management sees those positive trends in its AI forecasts.
Investors have yet to see those AI models proved correct with actual results. So if you choose to invest in this company, it requires faith that Lemonade's AI models can predict the future. Once the loss ratio goes below 75% and investors see the company move toward profitability for several quarters, the stock should take off. If not, it will remain in the doldrums.
