Analysis and trends of investment in the fintech insurance subsector
- Investor interest around insurtech has remained strong amid an overall slowdown in venture capital and fintech-specific funding activity. The operational and technological expertise required to operate an insurance-focused tech startup pushed many of these business models later in the cycle. There’s simply a smaller pool of people with the range of chops to execute on these business models in insurance relative to similar challenges in other industries.
- Given the diversity of niches along the value chain, insurtech has come to include diverse applications of nascent technology including artificial intelligence, IoT and drones.
- Corporate VC arms have been more active in insurance than other industries. One driver of this trend has been recognition of the need for innovation, but also the role of supplier played by insurance companies’ large balance sheets, which are already used to invest in VC funds. but also the historical trend of insurance companies utilizing their large balance sheets to supply capital to VC funds.
Investor interest in insurtech has taken longer to build in comparison to general fintech. In spite of the sector accounting for 7% of US GDP, the operational expertise required to launch a platform in the insurance space remains far higher than that required for other applied technology platforms. The insurance industry in general has bought itself time with huge economies of scale and massive balance sheets due to the favorable quirks of their accounting treatment and business models. Lately, insurers have begun to use their balance sheets to fund corporate venture arms to lead the charge on funding potential disrupters. This phenomenon has emerged as corporate VCs serve as a focal point of innovation between traditional VCs partnering with insurance giants to provide capital, and startups providing innovative technology and ideas to the industry.
The slow-building interest in insurtech has meant that private capital flowing into the space remains robust, even as fintech and general VC activity experiences a slowdown. Through the first five-odd months of this year, the $1.4 billion invested in the space will soon eclipse last year’s $1.7 billion in total deal value. The 61 completed deals so far in 2017 stand about on pace to match last year’s 136 completed deals. Not only have early-cycle companies matured and are now raising larger follow-on rounds, but also a number of seed and early-stage opportunities have emerged, integrating more advanced technology including AI. Furthermore, the higher-valuation early unicorns in the health insurance space have yet to achieve profitability and scale beyond specific regions. They may need to raise additional capital in order to fund geographic expansion and navigate continued regulatory complexity, thus greatly increasing total deal value.
A clear trend has emerged with CVCs or insurer-affiliated VCs such as MassMutual Ventures and XL Innovate partnering to fund promising early-stage startups, and then providing expertise in partnership with traditional capital sources in bringing these companies to market in the later stage. Insurers have historically invested in venture vehicles for non-strategic reasons to generate a return from the float of their premium payments. This existing expertise makes insurers natural partners with plenty of capital to fund proofs of concept before bringing on institutional VCs.
Since the passage of the Affordable Care Act, the unit economics of health insurance have proven challenging even for incumbents. In recent years, several companies have raised significant venture funding on the premise that the millions of people looking for insurance online for the first time would demand a better user experience and a more online-optimized, customer-centric business model from their provider. However, these companies have struggled to achieve scale and profitability in part due to regulations including state marketplaces that keep insurers operating at the state and local level.
The greatest success in offering direct-to-consumer healthcare coverage has been found in niche markets that are more profitable for incumbents and startups alike. Clover, backed by the likes of Sequoia, First Round and others, offers Medicare Advantage plans to customers in New Jersey. The company recently joined the unicorn club in May, raising $130 million at a $1.2 billion post-valuation. Clover has emphasized technology by implementing systems to improve outcomes via data, tracking for example whether prescriptions are filled and following up with patients.
The company reported a loss of $34 million on $140 million in revenue for the 12 months ending December 2016.
Oscar, the most richly valued insurtech startup, most recently raised capital at a $2.7 billion valuation in a 2016 round led by Fidelity with participation from a flock of other VC firms. In the 12 months ending December 2016, the company lost $204 million on $425.9 million in revenue. However, more recent financials accessible via the PitchBook Platform indicate that the New York-based unicorn has clawed its way to just shy of profitability by paring back operations to three states after backing out of New Jersey. These strategic changes to limit 1Q 2017 losses to $25.8 million, down from $48.5 million in the same period last year. During the period, the company has improved financial results by increasing premiums and also cutting back some network offerings in New York. Similarly, Bright Health has limited its geographic focus to greater Denver as it attempts to build a scalable platform. The company recently raised $160 million at a $400 million post-valuation to partner with local provider networks and to invest in technology to better connect patients with providers, aiming to make network and coverage information as seamless and transparent as possible. These local partnerships showcase why health insurance is difficult to scale across state lines, which bring on a whole new set of regulators and providers to work with in each new jurisdiction.
Applications for AI
The application of artificial intelligence (AI) to various business problems has received extended attention lately. Fintech and insurtech have been no exception. Quantitative hedge funds have applied machine learning techniques to large datasets of price action and increasingly alternative data in order to price securities and generate returns. One example is the sophomore effort of veterans from Climate Corp. after the firm was acquired by Monsanto for $1.1 billion in 2013. Similarly, these techniques have been applied to allow insurers to price risk and predict claims.
Compared to asset management, insurance offers far richer opportunities up and down the value chain, not only for pricing and risk management, but in enhancing the customer experience and expediting internal processes. On the client-facing side, chatbots and natural language processing (NLP) tools aim to provide quick access to answering questions and filing claims. However, text processing tools remain nascent, as many struggle with non-routine queries that only just recently reached a human standard. Speech processing is still years out from truly functional use, but both text and speech NLP applications will eventually become investable within the decade as they have the potential to solve a huge customer service pain point in the insurance space.
In the near term, deep learning techniques applied to image processing offer the most investable opportunities. The sweet spot for this process requires huge datasets of images tied to a specific business need. In the insurance industry this exists in the auto collision and home damage spaces. The next step will be curating further datasets using drones and other sensors to build a sample of sufficient size. One promising startup in the space, Tractable, recently raised an $8 million Series A from investors including Zetta Venture Partners and Ignition Partners. With previously developed proprietary deep learning techniques, the company is using the recent funding to specifically target the auto insurance claims space. Auto body shops already submit photo documentation for each accident and claim, thus the dataset is already more than complete. Enterprise AI services offer significant first-mover advantage and network effects given the benefits of a large dataset that comes from widespread industry partnerships. There is a limited window to build the key partnerships to develop a robust and scalable dataset that can be applied industry wide.
As technology sector and insurance business models merge, the market must decide to value these companies with either the lower multiples assigned to insurers or the lofty ones given to tech. Further private investment remains predicated on these companies’ ability to grow into their high multiples. Insurance companies have historically performed like utilities, trading at low valuations. Startup firms will need to continue to limit focus on the core technological value add of pricing risk and serving customers by offloading risk to partners. While historically reinsurers have in many cases partnered with startup insurers to purchase their risk, others have begun to work on the problem of securitizing risk at the portfolio level. This will allow the industry to move beyond traditional relationship-driven opaque partnerships into greater transparency and better pricing. Once companies can package risk into portfolios and sell off similarly to catastrophe bonds, traditional insurance partnerships will be turned upside down. This technology will perhaps allow manufacturers of insured products such as automobiles or smartphones to insure their products themselves, which will align incentives around durability and safety, allowing these manufacturers to build brand equity. In turn, such brand equity will also create lower barriers to entry into innovating in the space, allowing startups to attack specific pieces along the value chain.
Location: New York, NY |
Year Founded: 2013 | Capital Raised to Date: $51.05M
First Funding Date: December 2013 |
Latest Funding Date: May 2017 | Latest Funding Amount: $30M | Latest Funding Post-Valuation: $130.0M
PolicyGenius recently raised a $30 million Series C round led by Norwest Ventures. Since each insurance segment has differing unit economics, PolicyGenius’s strategy seems to be to fund break-even segments with higher-margin segments. Only recently have companies begun to attack the entrenched agent distribution network in order to reach the digital consumer. According to an oft-cited McKinsey report, the average insurance agent is 59 years old. Therein lies a huge opportunity to aggregate providers on a digital platform. This has been executed in other sectors such as in travel by Kayak, Expedia and others. PolicyGenius solves the fundamental conflict-of-interest problem of agent-based distribution, as their economics do not depend on swaying consumers toward one provider or the other but rather offering easy comparison of the best options.
Location: New York, NY |
Year Founded: 2015 | Capital Raised to Date: $59.7M
First Funding Date: December 2015 | First Funding Amount:
$13M | Latest Funding Date: December 2016 | Latest Funding
Lemonade provides peer-to-peer property and casualty insurance services designed to reverse the traditional insurance model. The company’s platform utilizes machine learning to improve the customer experience and expedite claims processing. In order to differentiate themselves from other providers, the company hired behavioral economist Dan Ariely to apply his research to the product. The company is backed by notable VCs including XL Innovate, Sequoia, General Catalyst and others. It most recently raised a $33.09 million Series B in December, giving the company a post-valuation of $208 million.
Article by PitchBook