Rooting (cautiously) for ROOT

I took a deeper look at ROOT after seeing it in @ryshab’s port overview. I remember looking into them in 2020, so was interested to find what happened in the meantime. Quite a lot! The stock price went from $431 in Oct 2020 to $4 in April 2023. And it is now at $39.71.

This is a very high-risk company, make no mistake. Just based on that share price trajectory alone, it won’t get much more than a small-ish position in my port.

Links:

IR site: https://ir.joinroot.com/financials-filings/quarterly-results
Latest Shareholders letter: https://ir.joinroot.com/static-files/9cba0cdf-4502-40ee-b98b-0b8bd550fcd0

Who leads the company?

Co-founder Alex Timm, CEO and Megan Binkley, CFO.

Alex is an actuary, Megan is a CPA, ex KPMG.

Says Alex:

I founded Root Insurance in 2015 on the idea that car insurance rates should be based primarily on driving behaviors, not demographics. I’m proud to lead Root in revolutionizing this outdated industry, using mobile technology and data science to give good drivers personalized, fair rates through an easy-to-use app.

What do they do?

Per Dan Manges, ex CTO and co-founder (since left the company):

Root is a car insurance company founded on the simple idea that people who drive well are less likely to get in accidents, and that means they should pay less for insurance. With that in mind, Root set out to reinvent a broken industry—an industry that was still assigning rates based primarily on demographics, had a complicated paperwork process, and determined pricing based on the driving records of all drivers. Using technology in smartphones to measure driving behavior, Root determines who’s a safe driver and who isn’t. Since Root doesn’t insure bad drivers, they can save good drivers money. And it’s all done in a simple, easy-to-use app.

Or as per their “About” page:

With a vision of bettering lives through better insurance, Root was founded in 2015 on the principle that car insurance rates should be based primarily on driving behaviors, not demographics. The company has revolutionized the archaic auto insurance industry using mobile technology and data science to offer fair, personalized rates to good drivers. The company has collected 20 billion miles of mobile telematics data to enhance and inform pricing models. With more than 12 million downloads, the modern app experience allows drivers to get their policy in less than 1 minute, manage their policy, and file a claim. A proprietary claims system, paired with advanced machine learning techniques, allows for automated loss expense, fast claims payout, and higher customer satisfaction. Root is the nation’s first licensed insurance carrier powered entirely by mobile and is currently offered in 34 states.

→ So in a nutshell, they use machine learning and a modern app to provide individualised underwriting for auto insurance. Something like the Upstart (shudder!) of car insurance.

The app got very good ratings online in a variety of places that I looked:

  • Trustpilot: 4.4 (2.3k reviews)
  • Apple App Store: 4.7 (69k reviews)

On to the numbers:

Revenue

Rev $m Q1 Q2 Q3 Q4 yoy Q1 Q2 Q3 Q4 Yr
2021 69 90 94 93 -45% -26% 86% 83% 0%
2022 85 80 74 71 24% -10% -21% -23% -10%
2023 70 75 115 195 -18% -7% 56% 173% 46%
2024 255 289 264% 287%

Something bad happened in 2021 and 2022. P&C insurance had a terrible couple of years. This was due mainly to inflation which put the insurers in a bend: they have to fix insured cars at higher prices than they thought when they issued the policies. Policy income stays stable, but costs go up. Not good. Because ROOT was not yet profitable by a long shot when this happened they were doubly in trouble: they were losing cash operationally, on their policies too, and they weren’t in a position to grow out of that hole. Hence they were priced as if dead. That explains the freefall part of the share price movement above.

So what did they do? They reinsured a lot of the risk, which led to revenues dropping and they cut costs - 17% of the workforce. They increased prices and prioritised surviving = making money/moving to profitability. That meant issuing fewer policies too. But toward the end of last year things seem to have turned and they started growing again.

It seems to have worked out, as revenues were up 264% and 287% in the last two quarters.

How did they manage to grow revenue?

Firstly, they took back some of the reinsured policies onto their own books. Secondly, they started ramping up their sales machine again as you can see from the number of policies they have in force below:

Policies in force '000 Q1 Q2 Q3 Q4
2021 369 383 390 354
2022 343 306 262 220
2023 200 204 260 342
2024 401 406

→ The number of policies in force almost doubled yoy in the last two quarters, albeit off a low base.

The higher number of policies written translated into higher new written premiums - which more than doubled last Q (and which will translate into revenue in future quarters):

Gross Written Premium $m Q1 Q2 Q3 Q4
2021 202.5 177.1 204.6 158.4
2022 187.2 140.1 150.7 122.0
2023 134.7 145.0 224.2 279.2
2024 330.7 308.2

Gross margins

Gross margins have been a bit all over the place but has recently grown to the highest level in the last 4 years, and holding steady for 3 quarters now:

GP% Q1 Q2 Q3 Q4
2021 9% -21% -17% -25%
2022 -14% -9% -11% -6%
2023 8% 17% 10% 24%
2024 25% 24%

EBITDA margins

Adjusted EBITDA margin has seen more improvements as the cost reductions of the past, price increases and efficiency in cost to acquire new policies all started translating into better profitability. They turned EBITDA positive for the first time in the last two quarters.

Adj EBITDA % Q1 Q2 Q3 Q4
2021 -131% -184% -125% -79%
2022 -61% -79% -63% -34%
2023 -16% -16% -17% 0%
2024 6% 4%

Net income margins

And of course the litmus test Net Income has trended in the right direction, with the company now on the cusp of (gaap) profitability after some truly impressive margin improvements over the last several quarters. In fact, the CFO said in a recent ER that the company would be profitable already should it stop new customer acquisition. They are basically at break-even, and with some additional scale should break into positive net income territory in the next quarter or two.

NI % Q1 Q2 Q3 Q4
2021 -145% -199% -142% -118%
2022 -91% -119% -90% -82%
2023 -58% -49% -40% -12%
2024 -2% -3%

Loss ratio

Another key ratio drove the improved economics: their loss ratio. Or how much they need to spend on claims from the premiums they earned in a given period. They targeted 65% and have exceeded that in the last q or two (lower is better):

Gross accident
period loss ratio
Q1 Q2 Q3 Q4
2021 76% 90% 93% 93%
2022 81% 85% 79% 78%
2023 66% 65% 64% 66%
2024 60% 62%

They are seeing the opposite from what hit them in 2021/22. They are now seeing inflation come down vs their expectations when pricing the policy, resulting in lower costs on claims. I think this dynamic has some way to run still.

→ Clearly there has been a lot of operational improvement as well as tailwinds over the last couple of quarters. That again explains the share price ramp from the lows of early 2023 (when it was priced for death) to now. Of course that is all in the past, what does the future hold is the ultimate question for us.

Going forward

They need to maintain policy growth or at least not decelerate too much to maintain momentum. Management mentioned that they expect to slow down a bit due to seasonal and competitive reasons. I’m fine with that as 250%+ growth is clearly not sustainable, and they have capital adequacy requirements too, which puts a damper on this type of explosive growth.

They also need to hang on to that loss ratio of theirs. On that score things are looking good, as policyholders that renew have a consistently lower loss ratio vs their total loss ratio in all quarters. Meaning that any slowdown in new policies written should result in a better overall loss ratio and profitability.

Another matter which should play in their favour is an imminent refinancing of a $300m loan which currently has a rate in excess of 10% pa. Management stated reducing the rate on their debt as an objective for the short term.

And then of course lastly we have a lower interest rate environment which should improve the affordability of cars and insurance and lift all boats.

Valuation

The company has a market cap of just under $600m for a ttm p/s of about 0.7x (yes). Annualising last q’s revenue, growing it by 20% for 3 years, assuming they can get to a 7% net income margin by then, and slapping a 15x multiple on the result gives a potential market cap of $2.1bn in 3 years. That would be good for a 3.5x or more from current levels.

Of course it could all go sideways too, and the stock could return to its all time low of $4…But I don’t think so.

Would love to hear opinions from industry insiders or other enthusiasts!

-wsm

(Long ROOT <1% position for now)

51 Likes

Thanks @wsm007 for doing the homework! Here’s how I would summarize:

  1. This is not too dissimilar to my previous investments in ELF and CELH when they were attractive investments. All three companies’ blistering growth was/is the result of taking share in a huge existing market. ELF had similar products, great marketing, and gained share by offering much lower prices. CELH had great marketing and an incredible distribution partner. ROOT is offering lower prices to a subgroup of consumers of auto insurance. It is using an approach similar to UPST, but it doesn’t have the main problem that plagued UPST, which was convincing the big banks to abandon their underwriting approach with something that while it seems to work better might have horrible outcomes if situations arise that might be unforeseen. ROOT is targeting the consumers directly so it doesn’t have about the bureaucracy and the conservative nature that big banks have. But all of these share grabbing companies (ELF, CELH and ROOT) will run out of share to grab. So if ROOT is targeting the 10% best drivers of the $300B market, they are targeting a $30B opportunity, and they have already grabbed about 3% of that. Some grabbing left but share gains won’t last forever.
  2. Their secret sauce is using mobile phone data to predict which drivers will have the fewest accidents. It’s not unlike what Tesla is also doing…using personalized driving data to determine each driver’s insurance premium, which can be dynamic if driving behavior changes. How strong is this competitive advantage? I think it’s ok but not the best. This could be replicated but how many other companies will replicate it before we run into a growth slowdown as described in #1 above? This is probably not an issue in the timeframe that would be the holding period of this investment.
  3. The auto insurance market will get disrupted with the increasing adoption of self driving cars be they robotaxis or cars with FSD which are individually owned. This disruption will take a number of years, maybe 5-10 years before it affects ROOT.
  4. I have to agree with @wsm007 that ROOT seems worthy of an investment.
  5. Maybe the holding period for this company could be 1-3 years provided the results continue to come in strong.

I’ve taken an initial 1.5% allocation starting position, and I may consider adding more, but it won’t become a large holding.

GauchoRico

36 Likes

The 7% profit margin is the key. I was burned by ROOT once and thankfully exited LMND with a profit. I see dissimilarities between the two and that is positive for ROOT. I especially like that they have gotten their LR down and seem to have a handle on underwriting their book. The question is whether they can scale that UW process because the TAM is ripe for disruption. LMND, OTOH, shows signs of good UW results, but not consistently. Plus with all the new lines of business they have added, it’s tough to figure things out. Plus, they have a self-proclaimed smartest guy in the room running the show and I shouldn’t let that affect my investment decision, but it does. If LMND’s process was as good as he says, amnopther carrier would just buy the company as its value has shrunk so much.

Life actuary by training, but am familiar with P&C. That disqualifies any opinions.

Vince

9 Likes

Anecdote:. I went to the website and entered my information. My insurance rates came back five times more expensive than my current insurer, USAA. Driver with a clean record.

12 Likes

It looks like you have to do a test drive period according to AI and the Root website.

From Perplexity AI these are the primary factors for determining driver safety,

The app collects data on several driving behaviors, including:

  • Braking habits (smooth vs. hard braking)
  • Speed of turns
  • Total miles driven
  • Driving routes
  • Phone usage while driving
  • Time of day on the road
  • Acceleration patterns
  • Overall smoothness of ride

Root utilizes the following smartphone sensors to gather driving data:

  • GPS
  • Accelerometer
  • Gyroscope
  • Magnetometer

These instruments work together to accurately determine the motion of your phone and, by extension, your driving behavior.


I am bit skeptical that this couldn’t be manipulated by driving cautiously for the week or two of the trial run. Also thought that GPS data from the phone wasn’t quite that accurate to be able to determine precise acceleration and speed of turns, but could be wrong.

5 Likes

State Farm offers a device and an app on the phone for lower rate to monitor your phone usage plus uses accelerometers to gauge acceleration and deceleration, Cornering etc.

I don’t see enough market differences

4 Likes

This is not a product review site, but I think it worth knowing the potential potatoes of a product being sold. As I understand, I would need to sign up for the insurance and drive around for a couple weeks? If, the price is not competitive, then I would have given up my old policy, which might contain some discounts, which might not be granted to a new policy?

2 Likes

That is why it is great to have experts on this board. I believe that life actuaries can also do quite a bit of P&C (imho :wink: )

@Fool4ZTribe there is a ton to unpack in those couple of sentences and I would like to pick your brain on them:

The 7 profit margin is key.
and
The question is whether they can scale that UW process

I agree on both counts. It would depend on them being able to keep their loss ratios down below that 65% target of theirs as well as scaling their book. They said in one of the ER’s that their incidence rates and severity rates didn’t really change that much, which means it is pricing both on the revenue and cost side that is mainly driving this. In your view, is this sustainable, or do they also need to improve severity and incidence rates? Their gross combined ratio has been improving over time too, such that they are now at breakeven:

Gross combined ratio Q1 Q2 Q3 Q4
2021 156% 197% 173% 144%
2022 136% 152% 130% 132%
2023 123% 118% 119% 110%
2024 100% 100%

It seems that they are on a path to profitability and if they can scale they should get there. On that front they seem to have been doing well, by reinsuring less and less, if I were to use Revenue as a % of GEP as a crude measure of gauging this:

Revenue / GEP Q1 Q2 Q3 Q4
2000 86% 80% 33% 33%
2021 43% 50% 50% 49%
2022 49% 47% 47% 50%
2023 54% 57% 72% 91%
2024 93% 94%

There is one other matter which I have not been able to handicap, and that is what is the limit of their growth (and the options open to them to increase that) imposed by their available capital and capital adequacy requirements; perhaps you can give a guesstimate? What do you see as the risks to them reaching that 7% net margin? Would love to hear your thoughts.

A couple of data points to some of the other posters.

Target market
If you’re not below 35 you’re probably not the target market and your anecdote of your experience with them is largely irrelevant. Like if I were to opine on ELF (I am not gen Z, no).

Manipulation to get a good rate
This happens in all types of insurance and is why new policies are generally less profitable than older ones. I’m not too worried about this part as, as they build their book, they root out the cheaters as you can see from this graph in their latest shareholders’ letter:

Notice the percentage in a bubble at the bottom of the bars. That is the gross period accident loss ratio for renewals - i.e. customers that are not new (lower is better). This is lower in all quarters than the total - i.e. blended for renewals and new policies. So new policies will be even higher than the average.

How the product works and differentiation

Suggest you view this video from a comparison site. It’s 9 minutes and you can watch it at 1.75 speed:

11 Likes

Bumping this up as ROOT had a monster quarter. All underlying metrics moved in a strong direction its first profitable quarter. Shareholder letter here: https://ir.joinroot.com/static-files/6f21bcfa-1ca8-47bb-9dbf-9f850e9e4ee1.

Taking a look, it does have a bit of an UPST vibe: using technology to identify safer customers than traditional underwriting methods. However, ROOT already passed through it’s moon and crash, which means it can valued under more reasonable considerations. To @wsm007 's points, it does look like promised efficiencies are taking hold. To @GauchoRico 's points, the question becomes just how much runway is left for new policy growth (which again gives me an UPST “Loans Transacted” vibe).

Anyone following closely enough to chime in? @ryshab?

10 Likes

A bit of real-life experience from this morning on ROOT and EVER.

My homeowners and auto policies are coming due shortly. Because of the conversations here, I decided to investigate ROOT and EVER, and elected to engage ROOT first.

I began to fill out the on-line data on the ROOT site and within 1 minute I began getting calls from EVER representatives. In the next 3-4 minutes, I had received 4 calls from EVER. The first three were voice calls from foreign representatives of EVER soliciting info on my insurance needs, but who could not be fully understood because of the degree of their accents. I asked each how they got my number, and they responded with something like, “Didn’t you just request insurance information on-line?”. I ended each call without providing any info, and the fourth call asked if I wished to opt out of all future calls.

With the interruptions from calls, I never completed the on-line application, subsequently elected to abandon its use.

Not a positive experience.

Gray

27 Likes