The chief executive officer of an InsurTech insurance company responded to recent reports about his company’s unfavorable 2017 loss ratio result on a social news platform last Friday.

“We want to make the right long-term economic decisions, not optimize a quarterly P&L,” wrote Alex Timm, the chief executive officer of Root, an InsurTech carrier offering auto insurance to good drivers, also noting the conservatism in loss reserving to help explain why Root’s incurred losses exceeded premiums collected last year.

The statement came during an “Ask Me Anything” session on Reddit, which also featured Root’s Lead Data Scientist Joe Plattenburg, the company’s Social Media Manager Justin McIntosh and Chief Product Officer Lauren Gruenebaum, responding to questions from customers and insurance industry participants.

“We are Root, a car insurance company that’s actually fair. We use AI to provide fair and affordable rates,” the introduction to the Reddit AMA said. “We use artificial intelligence and an app on your phone to give you a rate based primarily on how you actually drive,” it said, before listing the Root leadership team participants.

The back-and-forth Q&A that generated Timm’s comment about long-term thinking started with a question about underwriting losses at startups like Root. “Similar startups (Lemonade, Metromile) are already experiencing significant underwriting losses, which they can offset for a while with the tech startup capitalization, but it doesn’t look great for the long haul. Where do you see Root doing something differently?” the questioner wrote.

Timm’s initial response stressed that his company puts pricing accuracy at the forefront of its business model. “Over the long term, prioritizing underwriting and pricing accuracy will lead to superior performance,” he wrote. He noted, however, that during a startup’s growth phase, “there will be substantial investments into technology costs and customer acquisition,” explaining that an improved steady state of financial performance is reached as the book of business matures.

That response drew two follow-up questions from different Reddit users—one asking what Root would do differently from similar companies experiencing high loss ratios, and another specifically asking Timm why his company had such a high loss ratio in 2017.

Carrier Management recently republished a LinkedIn article by Matteo Carbone and Adrian Jones, which revealed that Root’s 2017 statutory net loss ratio (including loss adjustment expenses) was 168. The questioner revealed the Root loss ratio figure, linking to the Carrier Management article “Dispatches From InsurTech Survival Island: Five Takeaways From Statutory Financials,” in a back-and-forth with other non-Root participants who wanted to understand the genesis of the question. The article by Carbone, the founder of The Connected Insurance Observatory, and Jones, a reinsurance executive, revealed 2017 net loss ratios of InsurTech’s Lemonade and Metromile of 75 and 107, respectively.

Responding to the more general question about the loss ratios of the trio, Timm said: “High loss ratios are a function of either (1) inaccurate pricing and adverse selection, in which case data and pricing segmentation need to be addressed, or (2) a willful decision to lose money to gain market share.”

Pointing to the prospect of having “superior pricing segmentation” from smartphone data, he said superior loss ratios would result from Root’s approach.

“We also have no intention of underpricing insurance to gain market share,” he added, using boldface type to emphasize the point.

Specifically responding to the question of why Root had a high loss ratio in 2017, he distinguished between paid losses, case reserves and incurred-but-not-reported losses for new insurers. “Early on, any new line of business does not have reserving or loss history. So at Root, we take extremely conservative reserves on our statutory financials, which supports our state expansion. If you pull the statutory financials you will see exactly this,” he said.

Timm also pointed out that loss ratios on new business run higher “before the book has had a chance to see the impacts of underwriting.”

“So when comparing a book of business you need to tenure adjust the comparison, otherwise you run the risk of hitting short-term loss ratio goals while ignoring customer lifetime value,” he continued, ending his response with his comment about his company’s desire to make the right long-term economic decisions.

The Reddit session lasted three hours, with 61 comments appearing in total while the chat was live (and more than a dozen others after the Root management team signed off).

Root’s Twitter followers, including Carrier Management, were alerted to the Reddit AMA in a Tweet a day earlier, offering a rare opportunity to engage directly with the management team of an insurance company.

InsurTech Lemonade engages with the public about its financial results through the “Transparency Chronicles” feature of the company’s blog and messages on social media platforms but does not specifically comment on statutory financial results.

How Root Distinguished Itself

During the Reddit Q&A, the loss ratio question was the most popular (based on Reddit user point values), but Timm and his team also addressed questions about Root’s competitive advantage vs. other insurers (including traditional insurers like Progressive and State Auto), about dealing with potential fraud, and about the two- or three-week test drives that customers take for underwriting purposes.

“Other companies aren’t leveraging data to [the same] extent” as Root, Timm said in another part of his response to the loss ratio question, simultaneously introducing remarks about the company’s competitive advantage. “For example, only using mileage leaves over 80 percent of the predictive value of this data on the table,” he wrote, seemingly drawing a comparison to Metromile. “Also, ‘flat’ pricing schemes that are currently seen at other new market competitors in other lines of business leave a company completely open to adverse selection,” he wrote.

Plattenburg started to draw distinctions between Root and existing auto insurers in response to a question from a 21-year-old BMW driver from the U.K., unhappy with rates charged by a traditional insurer based on age and not experience. (Editor’s Note: Root only writes in the United States currently and said it has no immediate plans for expanding to Europe.)

Elaborating in answer to a follow-up question about telematics devices, he said, “We collect all data through cellphone sensors…We don’t make any assumptions about our drivers before the test drive.”

An agent representing traditional insurers in the United States using data from telematics devices was more direct in asking Root executives why they believe they price more fairly than existing competitors with more experience and higher net worth.

InsurTech carriers offer glimpses into their business models through social media postings. Root took to Reddit last week to answer questions from potential customers and other interested parties.

In the introduction to the AMA, Root said: “Because we only insure good drivers, Root members save up to 52 percent.”

The agent later countered that his clients had been surcharged 10-20 percent by the existing competitors. Timm distinguished between underwriting and pricing decisions using accident history and violations and surcharges based on telematics data to respond.

“No other large carriers have surcharges on their UBI programs. Note that the actuarial indication on this data is closer to around 100 percent—so a 10 percent surcharge is nowhere close to what should be happening,” he wrote. He went on to offer distinctions between Root and specific carriers when pressed further in the Q&A, pointing out, for example, that existing insurers use telematics to price renewals—not for initial underwriting—and noting the use of third-party pricing models by competitors to assign their discounts, rather than proprietary technology.

Claims adjusters asking questions were more interested in how Root stays ahead of potential fraud schemes that might be possible when using an app and AI to underwrite and set rates—and to pay claims.

“We believe that AI and increased data capture actually enables us to reduce fraud. With this additional data, eventually we will be able to automatically detect crashes and damages—which will both cut down on fraud as well as provide a better consumer experience during the claim,” Timm wrote.

Plattenberg noted that if a customer claims to have been hit during a trip “when we can show they were not driving at all, this would be an easy win.” Instances of erratic and distracted driving prior to an accident could also be revealed, he wrote.

While some commenters chatted amongst themselves about ways to beat the underwriting and rating system—including driving carefully during the underwriting test period or leaving the phone at home when driving. “It is a big ask for someone to change their behavior for multiple weeks at a time for the hypothetical benefit of reduced insurance cost. In fact, if someone is able to change their behavior for that long of a time, it is probably just as likely that they may actually become a better driver through the process,” Plattenburg said, responding to the careful test driver scenario.

“We have logic that requires people to look like a ‘realistic human’ in order to become eligible,” he added.

If “I end up having to file a claim, will my rate go up?” one happy customer asked the executives, reporting that she saved a “ton of money” over a prior insurer by going through a test drive with Root.

“The answer is yes. The fact of the matter is, getting into an accident is a strong indicator of future accident probability, and there’s a large difference between people that can avoid those situations and those who can’t/don’t,” Gruenebaum wrote.

In answers to other questions, Root executives said that insuring motorcycles, writing lines other than auto and expanding to Europe are not in the company’s immediate plans.