Trupanion (TRUP)
Trupanion (TRUP), a story of marrying "math with love”. Taking over medical insurance for pets with moats set by foundations of a unique organizational culture and management.
For all intents and purposes, TRUP is a medical insurance company for pets. But unlike most insurance companies (i.e. P&C and Life) TRUP educates its shareholders in its unit economics the way a SaaS company would. This is probably the classic market view of TRUP….. SaaS for pet insurance (i.e. high growth coolness in boring industry #tech #disruptive).
If I consider an ‘attraction’ to insurance companies, it’s either because of the companies low-cost operator starts (i.e. Progressive, GEICO) and/or the leaders running the ship (i.e. Gayner at Markel, Sagild at Topdanmark, Watsa at Fairfax, Buffett at Berkshire…. A dual threat with GEICO) who allocate capital/float in a different manner.
Much of my attraction to TRUP was the latter. Surprise, surprise. I learned a few investors I aligned with in investment philosophy speak highly of Darryl Rawlings and his shareholder letters so I thought I’d dig in. It was an added bonus that the company was doing something weird/obscure like medical insurance for pets. Talk about unsexy. Right up my alley.
History
TRUP started in Vancouver, BC in 2000 and IPO’d in 2014. Rawlings started the business with funds he received from selling his cigar business and seed money of $25K from each of his 8 early investors. Rawlings’ story is that he had a family dog who could not receive medical treatment because the bill was too high (I think some ~$20K or something… it was high) for his parents to afford and he’s wanted to start TRUP to address that problem ever since. He notes the cigar business was a means to an end to start TRUP.
Unit Economics
So what does it do? Well, like most insurance companies, it underwrites a policy for the pet (client) and pays out for claims when the medical need arises.
Rawlings says there are lucky, average and unlucky pets. But owners won’t know which of the three their pet is until death. Isn’t that how life works? So, because you don’t want to weather the uncertainty and risk of your pet being the unlucky one, you buy insurance.
Though pet insurance is not ‘mandatory’ like most insurance products like auto, property etc.. So it’s a discretionary product for pet owners.
Per the 2019 shareholder letter, the above is a look at how the unit economics looks per pet. Very much like a SaaS product.
I quite appreciate this point of view (POV) and it makes sense to me. The insurance is a monthly subscription with the average pet subscribed for about 6 years. Given the short lifespan of pets, I’d imagine the various medical conditions happen relatively faster than people. This is probably why most pets will have a claim paid out in the same year of enrolment.
Wait so how is this a business?
It’s a relatively simple business model. TRUP calculates the average expected medical cost of a pet (specifically dogs and cats) and charges a 30% markup. A cost-plus model.
If my husky’s average cost is $70 per month, I’ll get charged $100. This is how the lucky and average pets can help cover costs for the unlucky pets who might have to undergo $30K procedures. Like all insurance, it’s nice to have it when you need it but you don’t ever want to need it.
TRUP aims to make an IRR of 30-40% per average pet. The typical LTV/CAC (or in TRUP’s case CAC = PAC) used in SaaS companies ignore overhead costs so the IRR may be a better indication of the profitability of the operation. Previously, the IRR target was 40-50% at ’scale’ but management noted that they lowered the target to 30-40% to dedicate more reinvestment for growth.
Who gets pet insurance?
Not many people. As of 2019, the market penetration in North America (TRUP operates in Canada and the US.. it’s based in Seattle instead of Vancouver) is 1-2%. Compared to areas in Europe that have a 5-20% penetration, this is meagre.
North Americans spend some $72B on their 200M pets annually. Halloween accounts for $500M of spend on costumes alone. Yet, in 2018, we spent about $1.4B on insurance for 2.4M pets per North American Pet Health Insurance Association (NAPHIA). So, people do spend lot’s of money on their pets, just that medical insurance isn’t a popular thing yet.
Apparently, pet insurance in the past hasn’t been great with most only covering 50% of the total cost and reimbursements taking weeks. Not to mention that vets in different regions charge different prices. Different prices also apply to the breed, age and other factors of the pet. Oh, and there are all kinds of restrictions on various kinds of pet insurance that make it relatively poor value for the cost.
Who provides pet insurance? Market composition.
NAPHIA and TRUP say there are about 20 brands offering general ‘pet insurance’. Though TRUP says they are the only ones who provide ‘medical insurance for pets’. I’ll dig into this a little later. But there generally are 2 major players.
It’s Nationwide with ~40% market share and TRUP with 20%. Then the 18 brands have some piece of the remaining 40%.
If we looked at the number of veterinary hospitals, the providers of the service that people are making claims for, there are 28K in North America. Of which, ~21K are independent. This number has been declining as consolidation is happening in the market. As vets consolidate, I’d imagine that consortium may streamline any medical insurance they use to one provider for ease. Probably choosing one of the few large players.
TRUP has ~10K active hospitals and enrolled ~646K pets in 2019.
What makes TRUP’s product better?
It seems to be 3 factors: direct pay, automation and pricing.
TRUP covers 90% of the cost compared to 50%. So it alleviates much of the financial burden. Most insurers don’t cover for preexisting conditions but TRUP is the only one that covers for future illnesses that arise from preexisting conditions. From what I hear per dog owners, the fact that preexisting conditions aren’t covered is one of the barriers for why they don’t bother with pet insurance because so much of what happens to the dog after gets attributed to the preexisting condition so they don’t feel there is much point.
But let’s say TRUP can cover the cost and you get 90% instead of 50% covered. But that’s probably also because TRUP isn’t a ‘low cost product’. They are a low-cost operator but their products are not the cheapest on the market.
On top of covering more (something competitors can easily adapt to), they also pay for the cost directly. Since the big $10k+ one-time costs can make a dent in most pet owner’s bank accounts, waiting for a few weeks for reimbursement can be difficult. TRUP will pay the vets directly to save the pet owner from the financial burden. Approximately 95% of invoices are paid within 3 months from the invoice date.
On top of the direct pay there is the automated claim with TRUP Express, their direct pay software that is implemented into the vet’s computer and within 16 seconds, the vet is paid immediately for the claim. In 2019, some 32% of claims were automated compared to 2018 when it was only 4.7%. As TRUP Express is installed to more and more of their active hospitals (vets with 1+ pets enrolled every 3 months). It makes the process simple and smooth for the pet owner and vets. Competitors don’t offer this solution… yet?
Then there is pricing. TRUP’s been collecting 20 years of data on all pets in their system. Data collected from processing 4M+ claims and 15M+ pet months. Unlike other pet insurance products that have wider price bands, TRUP gets granular with specific parameters like breed, age, regions, postal code, preexisting conditions etc… to help the owners get as accurate as possible on the pricing.
This means that a Bulldog in downtown Toronto may pay more than one in rural Winnipeg given various factors influencing the price of the same treatment. But it also means that each owner can take comfort that they are both receiving the same value. This can’t be achieved with broader pricing ranges because what commonly happens is pet owners in lower cost areas/breeds/situations subsidizing for higher-cost areas/breeds/situations.
What’s the Advantage?
I think a moat exists.
I’ll first touch upon the low-cost operator factor. TRUP is a vertically-integrated insurance company that underwrites its own policies, owns the data, owns the brand, owns their office building, and is the managing general agent of the entire operation. They also don’t use a reinsurer. Per management, this reduces operating costs by about 20%. I believe Nationwide is the only other pet insurance provider that underwrites their own policies and the other ~18 brands use third parties in various functions of the operation.
There is also the proprietary data piece. In 2015, Rawlings said it would take him 13 years to replicate the 15 years of data they had collected. I imagine there isn’t a purchasable repository of detailed pet data (general and medical) available that a non-insurer would’ve collected. I don’t think even Google would have all that data on the pets in North America. It’s one where the first-mover probably has the advantage.. where as you get bigger, you accumulate more data and it gets harder to copy. Though, I imagine there is a point of diminishing returns where the incremental $0.01 accuracy in pricing doesn’t matter. So, it may be a barrier for any new entrants but if a number of the incumbents can keep up… it’s an advantage that may weaken with time.
But in the meantime, the data allows TRUP to provide a higher value product on the pricing side of things. This may be the reason why the monthly retention rate is ~98% for pets over a 6-year cycle.
Though the retention is high, I don’t think the switching cost is high. A pet owner can end a subscription and that’s it. It’s not like an ERP software where the ongoing operation of your business is impacted. I’d say it’s a more emotional decision where you are choosing to risk the health of your pet going forward and having to be comfortable with not taking ‘option A’ if it’s too expensive. If I think about substitutes… given the data/pricing advantage and direct pay option, there is only a cheaper/lower-valued product as an alternative.
In regards to pricing power, vet services increase some 5-6% annually and TRUP has been able to pass that cost over to the policyholder. However, a rate change of >20% from new data resulted in 1.69% monthly churn compared to 0.93% monthly churn for a rate change <20%. This is a discretionary good that still needs to be a value product and has price-sensitive customers. It ain’t a Ferrari. I don’t think having pet insurance signifies higher social status you’d want to show off for.
TRUP is closer to Costco and Amazon. At least, those are the Seattle-based companies Rawlings admires and is emulating to some degree it seems. TRUP is focused on increasing profitability with scale but also sharing the benefits of scale with customers. What Nick Sleep calls Economies of Scale Shared. Costco and Amazon are both low cost operators and they also provide low cost products to customers. TRUP doesn’t provide the lowest cost products like Amazon does. However, it does payout more than its competitors do. Because it is a low cost operator and it is was willing to forgo immediate profit in the first 15 years, it was able to continuously run the model of spending 70% of the average pet owner’s monthly costs for vet care until it eventually built up enough scale to hit positive operating margins in 2014.
Rawlings said TRUP would hit scale at 640-750K enrolled pets and they are close to that number now and we are seeing operating margins getting closer to the mid-teens level management hinted to in the past. With incremental pets, I’d expect to see more cashflow generation to further reinvest in the growth of the business.
So how would scale become a moat? It seems that each incremental pet provides positive returns. With each incremental pet you obtain more data that can result in more accurate pricing. More accurate pricing will result in better value to future pet owners. The profit from each new pet will be used to acquire more pets. Incrementally more dollars can be spent to acquire pets, potentially expediting the growth of pets enrolled.
TRUP is dedicated to providing the best value product on the market at the best affordable price where they take a 30% margin off of the cost of an average pet’s vet expense. So the customer wins. But the other side of the equation are the vets who treat the pets. TRUP partners up with vets through its Territory Partners (TPs). TRUP has about 130 TPs throughout North America. They are independent contractors who visit veterinary hospitals to educate the vets on TRUP and its benefits to pet owners.
A vet referring TRUP to pet owners has been the most effective way to acquiring clients and that turns them into ‘active hospitals’. Soon, active hospitals get TRUP express installed, get a dedicated account manager and it contributed heavily to increasing same-store sales growth as more pet owners from that hospital sign up for TRUP. Vets don’t receive any referral fees or kickbacks for recommending TRUP. It’s all tied to the ‘win-win’ relationship.
Vets want to provide the best care for pet owners. This means recommending ‘Plan A’… the best course of action for any treatment. Usually, this costs most but its also the best option for the pet (assuming the vet takes their fiduciary responsibility seriously… but I mean who becomes a vet purely for the money.. it’s not like law or human doctors…). The vet wants what’s best for the pet so if the pet owner had medical insurance, they might be able to take on the best treatment. This would also allow the vet to generate higher sales. But, a relationship between the pet owner and vet is built on trust. Hence, if a vet were to recommend pet insurance and it did not do what it promised… if it did not provide any value… well then.. that jeopardizes the vet’s future relationship with the pet owner. This is all a long winded way of presenting the scenario for why its important for vets to only recommend products they truly believe in and why the best product for the pet owner is the best for the vet and how if every stakeholder involved does well, then the company offering the product will do well in the process.
A point to add on the power of a salesforce is that I think there are really two ways to go about it. You either have minimal/none like Atlassian and allow for bottom-up organic adoption or you go for the highly engaged relationship-building way that takes years of knowing someone. Most companies do something in between because of resource constraints (people/money) and get some half-assed results. TPs build relationships with vets over years and the amount of resources it takes to build up such a team is not something that can be easily replicated by any of the small competitors. Maybe Nationwide.
But the difference with Nationwide and TRUP is that for Nationwide, their pet insurance division of IVP is a small part of a much larger insurance company. For TRUP, this is the only thing they do. So, if survival is the name of the game, then I’m willing to bet a company with a homogeneous product will win over one that is distracted and unfocused.
When reading the shareholder letters of the last 5 years, it’s evident that TRUP had to grind its way to squeeze out its operating margins from 1% in 2014 to 10% in 2020. This is not an easy journey for most companies. Even if it’s considered the fashionable thing for companies to blow through VC money in the name of growth to hit scale later. TRUP has actually managed to achieve this while self-funding its reinvestment flywheel.
This leads to what I consider to be TRUP’s sustainable moat. That’s its culture. TRUP started with a mission and purpose from Rawlings’ own experience.
The Culture and Leaders
When I read shareholder letters I have a bias to look for mentions of how the company thinks about employees and investing in people. Shareholder letters can be used by sales-y CEOs to go on about how important shareholders are and most of the time, I’m not too convinced by those letters. TRUP’s letters never fail to refer to how important the culture of the organization is to the company’s success. Rawlings continuously refers to how he not only looks for top talent but those who align with their mission.
Of the 400+ employees at TRUP, 70% are from the animal health world. They also have 200+ dogs and cats in the office as well. Bringing in people who are all obsessed with the same mission. I guarantee you 70% of people in Facebook are not passionate about making people addicted for advertisers.
There are small signs like how the office wall has the company motto: "Be yourself, everyone else is taken” by Oscar Wilde. This is my favourite quote by Wilde too so I may be biased here. In his 2018 letter, Rawlings reflects on the past 5 years and in each year there is a reference to how he mismanaged people. The process he created to stop himself from being a roadblock to people, to help individuals be more accountable from the shortfalls of a traditionally ‘flat’ org. structure.
“We know that execution - our biggest risk - comes down to people and culture.”
Rawlings seems to have a similar belief with me as well. There are other examples like how every desk in the company is the same size regardless of the tenure of an employee or their status as full-time or part-time to signify equality. Maybe this is just show… like a ping pong table. But I think it shows Rawlings’ character.
Even after their IPO, the executive team’s celebratory meal was Shake Shack takeout and a picnic at Central Park. Simple and modest.
When looking at compensation and promotions, the proxy statement mentioned how 123 people were promoted internally within the company. I haven’t seen a proxy statement reference internal promotions in the past. Maybe it’s a greater indication I haven’t read enough proxies but this is different. TPs, the front-line face of the company, made up 9 of the top 25 highest compensated employees in the company (including execs). In relation to acquiring talent, Rawlings noted: “.. Our side-by-side comparisons confirmed that our internally developed candidates were still the best people for the jobs.” Reflecting how he learned that developing people internally was much more powerful than just buying executives. Something I believe in.
There is also the added bonus that he references Tobi Lutke in the 2019 shareholder letter. As a big fan of Lutke and how he thinks about Shopify’s culture, I believe this is another signal for me.
Looking at Incentives.
Nothing like looking at who owns stock in the business to get indications on alignment. I was rather disappointed actually. Though the insiders collectively own ~11% of the outstanding shares with Rawlings owning ~5%, I would’ve hoped to see a higher number. Dan Levitan is the early VC investor for TRUP and both him and his fund, Maveron, have been lowering their ownership from some 20% since the IPO to current single digit levels.
With 5% of his net worth in TRUP Rawlings has ~$50M tied up in the company at a $1bn market cap. It’s good to see the insider ownership but not great is all but not everyone can be like Buffett or Bezos I guess.
Noteworthy are a number of large institutional ownership in the company, making up ~46% of outstanding ownership. With a large short interest of ~30% of outstanding float…. It makes me iffy on whether this will be a battleground ’stock’. Not company. Though I see the involvement of Brian Bares through Nine Ten quite favourably as I’ve grown to like his qualitative and management focused investment style.
Okay, so how do people get paid? It’s not some chicken shit metric like sales or stock price. Thank god. Would’ve been a slap in the face after reading all his letters that quote Malone, Schultz, Bezos, and all the other ‘Outsider’ CEOs.
The exec team has most of the total comp allocated to long-term incentives. As far as salaries go, the C-suites all range between $200-$300 with everyone set to a flat $300 as of 2020. All in, with the equity comp, all execs float in the $1.1M to $1.2M range with chief of people & general counsel receiving the most. Rawlings determined that because he already had much of his wealth tied to the stock price, he did not need to take more of the equity compensation.
The short term incentives were quite modest with the target set at 20% of salaries. Yeah, $40K isn’t a small amount and maybe I’m getting lenient but I felt it was modest and reasonable… even Canadian… compared to the outrageous range of average public companies.
The way long-term incentives are measured is by the 2-year CAGR of intrinsic value per share of the company. Different from what I’d seen in the past.
The proxy didn’t provide details to how they calculate the intrinsic value (IV). Odd given that’s the measurement metric. However, the 2019 letter shed light to the 15 year DCF TRUP uses internally and it provides a view into how they think about the growth of their intrinsic value. The way IV is calculated looks at the FCF generated, accounting for the pet acquisition cost, all in line with the unit economics equation shown earlier. Since they do it on a per share basis, it would limit use of serious dilution and it’s also worth looking for silly buybacks in the future.
The current way of calculating IV assumes that capital will be reinvested as part of PAC to acquire more pets. Not much accounting for acquisitions since it doesn’t use traditional return on capital metrics like acquisitive companies tend to do. I think with the current strategy of reinvesting the cash back into acquiring new pets, the IV calculation method seems reasonable.
Rawlings speaks at length to why they took the cash-based approach (referencing Malone and TCI no less) and how EBITDA doesn’t probably reflect well on a company that reinvests through SG&A compared to acquisitive companies. Reasonable enough. But it’s this continuous communication that I appreciate and I do have faith that if they were to change the method to align themselves better or if the method doesn’t probably reflect recent transactions, I believe Rawlings will make note of that. In 2018, TRUP raised equity to purchase their office building. It diluted the shareholders and he made note of that in the shareholder letters where he showed business metrics in a per share basis and brought attention that revenue growth was misleading because it was some 25% for the company but was 17% on a per share basis. It’s shedding light to such information that he did not have to and sharing something that doesn’t make the company look good (financial result wise) that I respect him for.
Some investors have been critical of the 2018 decision to buy the building. And yeah I don’t know if that was the best use of cash. But, they are playing a game of thin margins… I mean, he is talking about shaving 0.3% somewhere and 0.7% somewhere… and if the entire business has to survive on 15% of the sale… I can understand the purchase if it will shave off 1% off of operating costs. If I was a founder of a company, I could imagine myself pulling that trigger. Not saying it’s right or wrong but I can understand why he might’ve done it.
Touching back on the equity compensation, so the equity pool incorporated some 479K shares available for distribution to the employees. Not just executives. Rawlings is clear on numerous occasions that the increase in IV will allow more shares to be distributed to employees and in 2019, ~38% of the shares were awarded to the 6 C-suite executives. That means the remaining 62% was given out to employees. Stock based compensation accounted for 1.8% of total revenue and with a majority split out to non-executives, I see this as an indication of responsibly incentivizing their own employees and instilling the ‘owners mindset’.
What Does the Market Think?
When you go on the Corner of Berkshire and Fairfax forum for TRUP you see the divergence in opinion. You have the group that says what kind of idiot buys pet insurance and how pets aren’t people and if one dies off you can get another one… and the other group that says they have pets and their pets are like family members etc… There are also a slew of arguments from the short side on regulation concerns because it is a new space and TRUP has had to pay fines in the past (Rawlings notes he anticipates he will have to pay fines in the future as well and it was a mistake not working closer with the regulators and it’s something he is set on improving for the future).
People have questioned TRUP’s ability to deploy capital because they bought a building in 2018, there have been concerns on whether NA will ever hit the penetration rate of EU given different cultures etc… Rawlings has also been heard saying he wants to move onto being an executive chairman in 2025 so such a prematurely intended departure from the CEO role is causing concerns for the long term. I guess the large short interest on the company encompasses the concerns and what the market thinks.
It’s divisive to say the least.
My 2 Cents.
I don’t own a dog. But my partner does. A Siberian Husky named Milo. My partner and I used to scoff at people who were so obsessed with their pets. I thought it was ridiculous my family friend got pet insurance for their aging dog for his eye surgery and I thought they had too much money to toss around. But, Milo had to get surgery in his first years and there was a chance he might not make it.
Thankfully, Milo came out well but it was not cheap. Right after that, I continued to remind my partner to look into pet insurance for Milo. This was all before I learned TRUP existed. But in a post-Milo world, I definitely understand why people treat their pets like a member of the family. Because they do become one. So I do see value in a product as medical insurance for pets. If I were to make the commitment to get a pet for myself, I’d probably get it.
I also think the market for pets will increase. Whether it’s me reading up on developments in South Korea or looking at what my peer group does in Canada.. pet adoption seems to have increased. As my friends start coupling up, getting married etc.. many are thinking about getting a dog before children. I think the recent COVID incident will create a period of economic recession that will make people think again on starting a family. I also think that humans want love and to feel someone depend on them so many will get pets before having children. I have no data on whether economic hardship results to more pet adoption. But with quality of life improving as a whole and people having more ’spare' time now than ever before in human history… pet ownership will increase. Now, will they get insurance? I’m sure a decent chunk will.
Per TRUP’s own 15-year DCF example, they showcased an IV of $1.2bn at a 10% discount rate and 5% terminal growth rate (vet prices rise about 5% annually).
The market cap today is ~$1bn. TRUP’s own example assumes that by year 2030 (10 years from now) they will have ~1.7M pets enrolled. If they maintain a 20% market share, that assumes a market penetration of 4%.. assuming NA stays at 200M pets for the next 10 years. Do I think we will have more pets in 2030 than in 2020? Yes, just no idea by how much. Their model also assumes a continuous annual growth of ~20% in cash generation until 2032. Hmmm..
At current adjusted operating income levels of $44M, the market is showing me a ~4.5% yield. With improving unit economics of 30-40% IRR on every additional pet… a #2 position… 2% market penetration… a discretionary product that does provide a win-win value… as the low-cost operator…. a data pool that can’t be bought… a sales team that builds relationships over years…. That can’t be implemented without sufficient scale in the business…
A culture driven with purpose… where I am seeing signs of management that is willing to invest in people to build a competitive advantage… I see it being a favourable valuation. It’s not screaming at me as cheap. But at sub-scale.. with a pretty clear execution plan… solid cultural advantage… as the dominant homogenous company.. I like what I’m seeing.
Disclaimer: Not investment advice. My opinions, observations and thoughts. Heed at your own risk.