Description

“Buffett had Geico. I pick Metromile”

Chamath Palihapitiya, Nov-2020 via Twitter

I pick Linea Directa.

As it turns out, Chamath’s idea could have worked: well-run insurance businesses CAN BE great investments. In P&C insurance, there are certain basic qualities that travel well: disciplined and profitable underwriting through cycles; low running costs; customer service above direct peers; a culture that incentivises and reinforces the previous points.

It would have worked better for Chamath to look for Geico’s clone abroad. There will be local nuances of market structure and regulation to keep in mind, but if you are familiar with “your” Geico, finding a clone elsewhere can offer geographical diversification and (mostly) uncorrelated risks while keeping you exposed to a familiar business model.

A few “clones” come to mind:

-              GEICO. From 1951 (when Buffett first invested in the business) to 1995 (when Berkshire Hathaway took the company private) the stock delivered a 21% annualized return.

-              Progressive, founded around the same time as GEICO: a shareholder who bought shares in their 1971 IPO (the first time they came to market) would have compounded at >20.5% (assuming dividend reinvestment) as of the end of 2021, (the S&P did ~10% in the same period)

-              Berkshire Hathaway: since Warren Buffett took over in 1965, the stock has compounded at 20% annualized. The S&P 500 has done ~10% over the same period.

-              Admiral: since its 2004 IPO in the UK, annualized returns have been 18.5% (the All Share did 6.5% in the same period) 

-              Qualitas Controladora de Seguros: a Mexican motor insurance specialist, listed in 2012 in the local exchange. It has delivered almost 19.4% annualized returns (in MXN) since 2015 (16% in USD) vs. a MEXBOL return of 3.75%!

-              Sampo: the Finnish holding company whose largest business is car insurance in the Nordics, has compounded at >15% annualized since 2000, vs. its local index at 2%.

-              Tryg, Gjensidige Forksikring and Topdanmark: a group of Scandinavian insurers, have compounded at ~17%, 18% and ~18% respectively since late 2010. The OMX 25 index has done 10% over the same period.

And of course, Linea Directa Aseguradora (LDA), a Spanish motor and home insurer with a short history in the public market but 26 years of organic and profitable growth under the umbrella of Bankinter, who executed a Spin-off and IPO of the business in March 2021.

Peculiarities of the motor insurance model:

In most developed markets, car insurance is a compulsory purchase. It’s the ultimate subscription business with a few caveats: it is impossible to know the costs of goods sold upon the sale of a premium; barriers to entry are relatively low: as capital flows into the sector, premiums decrease. The opposite happens when it withdraws; the key inputs to the model are based on estimates: the insurer agrees to cover a customer against certain risks for an agreed period in exchange for a premium that it believes will cover the cost of any potential claims, the cost of running the business, and leave a margin.

Scale can be a friend: a larger and relevant database can help in deciding which customers to underwrite and at what premium; running a lean operation is always helpful, and amortising a relatively fixed cost base on a larger amount of premiums offers an extra cushion against the volatility of the top-line; having a larger balance sheet protects you from an unlucky clustering of large claims, for instance; and so on.

Inflation is an acute problem for insurers. The cost of fixing bodily or metal parts increases each year. Insurers incorporate their inflation views in the premiums they charge. But claims can take years to settle and some, once settled, require ongoing payments for years, so a large mismatch between expected and realized inflation is dangerous.

Frequency of claims collapsed 2020 and 2021 due to Covid restrictions. Motor insurers around the world had the most profitable years ever. Underlying severity of claims kept creeping up, however: record used car prices, shortages of labour, garage bottlenecks, supply chain issues affecting availability of parts, etc.

Insurers lowered premiums in these years, either to compete for market share or to retain customers, and have been stunned by the inflationary shock that came with the normalisation of driving activity after the lifting of Covid restrictions. In rare synchronized fashion, Combined Ratios in the industry, in different markets, have increased, in many cases past 100%.

The playbook from here is simple: premiums need to increase to the point where industry profitability is restored. Efficient and prudent underwriters should benefit disproportionately: they need lower premium increases to revert to their targeted loss ratios compared to less able peers. They should be able to take market share.

The timing of premium increases takes a different pace in each market, depending on local characteristics: intensity of competition, regulations, balance sheet strength of market participants, etc. In Spain, premiums are increasing at a slower pace than some participants, or their investors, would like. But like in other markets, inflation for drivers is coming.

A bit of history

Direct Line launched its direct insurance model (telephone-only) in the UK in 1985, backed by the Royal Bank of Scotland. Cutting out the intermediary from the sale process turned out to be a good idea, with the business dominating car insurance in the UK during the 90s and early 2000s, peaking in 2006 at close to 38% market share.

In 1995, Alfonso Botin (of Santander and Bankinter fame) and Peter Wood, founder of Direct Line, partnered to launch Linea Directa. The idea was simple: apply what was working in the UK for Direct Line to the Spanish market: no brokers/intermediaries to keep distribution costs as low as possible, invest in building a strong brand, offer a better service, and be selective with your risks.

Grow and repeat.

It worked. There were no monoline car insurers at the time in Spain and multi-line incumbents didn’t seem to care much about turning a profit in a line of business they had little interest in but helped sell other types of coverage and brought in investment income from the premiums. They also had higher distribution costs and therefore offered higher priced premiums.

By 1999 they were breaking even. In 2009, Bankinter bought Direct Line out and continued growing the business. They began offering Home Insurance in 2007 and in 2018 entered the private health insurance market. They remain focused 100% on the Spanish market and under the umbrella of Bankinter became the 5^th largest car insurer in Spain (2.6 million customers, 7% share), and the 7^th Home insurer (750k customers, ~3% share).

In April 2021, the business was listed in the Spanish market and Bankinter spun-off most of its holdings to its shareholders. I estimate Bankinter realized over 20% annualized returns from their early investment in Linea Directa in 1995, dividends and capital appreciation combined. They remain shareholders with 17% of the business. Alfonso Botin’s family office retains 19% and the Masaveu family holds close to 5%.

Spanish motor insurance market:

There are about 32 million vehicles in Spain and ca. €11.5bn in premiums. 43 companies underwrite car insurance, with the top 5 accounting for 58% of the market as of June 2022. This group had 57% share in 2010 and 62% in 2020.

The stability at the top reflects 1) a profitable sector: with combined ratios averaging 96% since 2010; 2) entrenched distribution channels and purchasing habits: with over 80% of new business conducted via brokers, retention rates of around 84% (pre-Covid) and automatic policy renewal; and 3) low average premiums compared to other european and global markets, which peaked in 2004 at €457/policy and have fallen at ~2% annually since.

The incidence of large bodily injury claims is lower in Spain than in a market like the UK, for example, as are typical bodily injury costs. A large claim that may cost €17 million to settle in the UK may settle for €3 million in Spain. Baremo, a rate table by type of claim introduced in 2016, increased overall bodily injury costs, but levelled the playing for all participants, reinforcing scale benefits of incumbents.

LDA will write ca. €785m in motor premiums in 2022 at around 94% Combined Ratio.

Spanish Home insurance market:

There are 26 million homes, 21% are uninsured, €5bn in premiums. Agents and bancassurance control 80% of new business distribution. Premiums are low at around €226/policy. The market is profitable, with an average combined ratio of 92% since 2010, and it’s less concentrated than motor, with the top 5 insurers controlling 41% of premium in 2022, essentially the same share as in 2014.

LDA will write ca. €143m in home premiums in 2022 at around 94% Combined Ratio.

Additionally, LDA will write around €32m of GWP in Health at around 141% Combined Ratios.

What Linea Directa do well

In a modestly concentrated market Linea Directa still has a long runway for growth as its share in each of its markets is below 7%.

While higher profitability and market share cannot be guaranteed, some qualities of LDAs business increase the likelihood of both in my view:

1. Underwriting advantage: demonstrated by its sustained outperformance in claims ratios vs the market as a whole.

In Motor, LDA has achieved claim ratios 5pp better than the market on average in the 5 years to 2021 and 7pp better in the last 10. It has also been consistently profitable for the last 21 years.

In Home, despite higher growth rates, LDA has achieved 9pp better claim ratios compared to the market in the last 6 years. Home turned profitable in 2013 and has remained so ever since.

Finally in Health, a market where it has only been competing for three years, its claim ratio was 89% in 2020 vs. the market’s 75.5%.

Being able to underwrite profitably is key to the long term success of an insurance business (ask Buffett?), more so when there are (have been?) limited gains to be made from investing premiums in high grade, short duration bonds.

2. Cost advantage:  Since 2010, the cost advantage in Motor has been 3-4pp better than the market. In addition to direct distribution, LDA also does a very good job handling claims, tackling fraud effectively and keeping repair and service costs under tight control. Over the five years to 2020, the average claim repair costs were €632 for LDA, €269 lower than the market as a whole.  

3. Customer service: LDA can afford to spend a little more acquiring customers if they want to as they get to amortize them over a longer timeframe, at a higher margin than most competitors. Up until the IPO and year-end 2021, customer service was embedded as an element of management’s variable compensation (for example by aiming for a minimum NPS of 38, while the market was around 18 as of 2018).

Good customer outcomes help with higher retention rates: in Auto, these remain close to 92% compared to the market’s 88%. In Home, they’re 86%. There’s room to improve in Health as they currently have a lower (~79%) retention than the market (according to industry contacts).

Multi-product bundling further improves these figures. For example, in Home, a customer who is also a Motor policyholder improves retention by 7pp.

61% of Home customers are also Motor customers, and 71% of existing Health customers are also Home or Motor customers. The introduction of the Home product in 2008 led to an overall improvement of the aggregate retention ratio in both Home and Auto, from 77.5% and 83.3% respectively in 2009 to the figures quoted above (86% and 92%).

High retention creates interesting economic dynamics: for example, Discovery, a South African life insurer that also offers motor and home, mentioned in one of their investor days (pre-Covid) that a customer with more than 2 products has retention rates above 96% (LDAs Vivaz’ health product is partially modelled after Discovery’s Vitality offering).

Beyond insurance but within the financials universe, Bank of America first showed in an old investor deck (from around 2010) how a customer with more than two products would increase retention to 99% or close (this is now regularly quoted in quarterly calls). Financial product bundling is the best retention strategy in what otherwise becomes a commoditized offering.

4. The right incentives: salaries have a fixed (70%) and variable (30%) component. The variable component varies according to achieving group-wide as well as division specific goals.

These goals align with Management’s own variable incentives (there was a change in CEO earlier this year, and new compensation plan details have not been disclosed, the following correspond to plan details under the previous CEO and for the period ended December 2021): for Executive Directors, 40% was based on Net Premiums Written targets (+4% for 2021 vs. 2020 for example) AND 60% based on Profit before tax (below 2020s figure but aiming for a ~83% Combined ratio).

For Senior Management, variable remuneration targets were weighted as follows: 30% subject to achieving Net Premiums Written, 30% subject to these premiums delivering a minimum profitability, 10% subject to the business NPS being 37.4 as a minimum and 30% subject to each area’s strategic objectives. There were no policy count targets or specific market share goals. The company reiterated recently that new incentives revolve around growth WITH profits AND customer satisfaction.

One of the reasons the long-term CEO left was the fact that COVID made it impossible to achieve the Gross Written Premium goals set for the 2019-2021 LTIP period, which were set at the end of 2018 in a pre-pandemic world.

Finally, over 95% of LDAs employees have an indefinite term contract, compared to an average of 75% for Spain as a whole.

5. Good stewards of shareholders’ capital: LDA has grown the number of policyholders organically over many years while delivering ~30% Return on Equity (on an unlevered balance sheet) and paying the majority of its earnings as dividends. Going forward, the dividend floor is set at 70% of net income. The company typically reserves for future claims well above its actuarial best estimate, recognizing profits as underwriting years develop.  It’s not clear what the excess figure will be after 2022 given claims inflation during the year, but as of the 1H21, they had over €260m of reserves over best estimates.

What about the risks?

-  Price Comparison Websites (PCWs) penetration: more than a decade after the introduction of the PCW channel, their penetration remains low. While LDA and other incumbents use it for a small portion of their new business, their practices tend to distort the search results and slow down the channel’s growth potential. For example, both LDA and Mutua Madrileña may offer a low price to appear high in PCW rankings, but then quote a different (typically higher) price when the customer calls them to finalize the purchase. Rastreator.com, the leading PCW, and Acierto, the #2, have in the last couple of years penalized those insurers who engage in said practice (i.e. lowering their position in rankings) unless they show a demonstrably “final” premium to customers shopping for insurance on PCWs.

The practice has been detrimental to conversion and PCWs customer satisfaction, but LDA and Mutua rightfully fear the UK experience and this has been their defence strategy so far. The diversification of PCWs into other verticals in recent years is an indication that their initial goal of disrupting the car and home insurance sector has not yet played out.

Admiral’s Spanish brands, Balumba and Qualitas, have been exploring alternative routes to market besides PCWs and announced a bancassurance distribution partnership with ING Direct earlier this year (they competed with LDA, interestingly, who said they are not interested in volume but profits and the terms of the partnership were not attractive in that regard). Admiral’s disposal of their Penguin Portals business, which included the Spanish, French, Italian and UK PCWs could be interpreted as a sign of diminishing confidence in the channel’s longer term prospects in some or all of these markets.

However, PCWs are not going away. They remain profitable and cash generative and can continue investing even if growth prospects have changed. If the channel manages to disrupt insurance distribution eventually, it could pose a challenge to incumbents like LDA and Mutua. This risk was somewhat muted during COVID given the downward pressure on premiums eliminating incentives for policyholders to shop around at renewal, but I would expect a hardening cycle to lead more Spaniards to Rastreator and peers looking for a better deal.

The structure of the Spanish market (in Motor) provides some natural hedges against the rise of PCWs. Admiral has experienced this. Spaniards are very conscious of brands when it comes to financial products (Admiral had to launch a new brand after Balumba’s funky caterpillar logo failed to be taken seriously. To position the new one, Qualitas, they launched an ad campaign featuring actor Pierce Brosnan, a display of “class and elegance”)

It’s no surprise that financial service businesses are big advertisers. Car insurers, for example, spent €226 million in advertising in 2018, €165 million in 2020 and a similar amount in 2021. Linea Directa and Mutua Madrileña, the #5 and #2 car insurers respectively, represent almost 60% of the advertising spend for this group. For reference, Rastreator.com, the main PCW, spent €6.5 million in 2018.

This is very different to the UK experience, where PCWs long ago replaced insurers as the biggest advertisers of financial products. Arguably, this didn’t happen overnight and Rastreator.com is a well-known brand, so perhaps they don’t need to spend a lot more to attract customers. Rastreator attracts around 900k monthly visitors, compared to 1.1 million and 1.1 million for LDA and Mutua Madrileña respectively. Still, assuming an aggressive 10% conversion rate of PCW visitors into policy holders would translate into around 1.1m customers in a market of over 55 million (cars and homes) still a low penetration of PCWs as a channel.

- Competitors: Mapfre operates more like a charity than a profit driven corporation. In fact, it is majority owned by a Fundación, itself a non-profit, which lives comfortably with the ca. €400m in annual dividends it receives, and exerts no pressure on management to run a profitable underwriting operation. The shares have underperformed the local index forever, compounding at 4% since 1989. It can be hard to compete with a business like this in any market. However, when Mapfre decides to finally increase premiums, efficient competitors like LDA and Mutua will benefit.

Mutua Madrileña: Mutua has no shareholders expecting dividends. Instead, its policyholders are members continuously benefiting from lower premiums. As a non-profit driven organization, they pass on efficiencies to members, which makes them hard to compete against. And they are good operators, which makes it even harder.

 -  Health is a more concentrated market than Auto and Home, with the top 5 players controlling over 70% of the market. Growing in health will be very challenging. Management is aware of this. Health insurance penetration amongst the Spanish population is close to 20%. There’s increased awareness after Covid of the potential benefits of having private cover, despite Spain providing decent public health services. LDA is catering to those who have not had private cover in the past rather than trying to capture share from incumbent players.

They’re doing it creatively by sharing reduction in claims arising from the adoption of healthier lifestyles with customers in the form of rewards, vouchers and lower premiums (hence my reference to Vitality above, although Vitality is a life insurance product, and Vivaz is not).

They’re also looking at cross-selling Health amongst their existing customer base. They’re doing this in a capital light way: rather than having their own hospital and health professional network, they are partnering with DKV who have built a comprehensive health-professionals network in Spain.

Lastly, they’re sharing 50% of the underwriting risk with reinsurers to mitigate the impact of early losses. The evolution of claims and expense ratios is very promising and they have reached almost 100,000 customers from zero in 2017. The combination of secular market growth, innovative products and 3.5 million existing customers is attractive and worth pursuing in my view.

They are currently operating at around 141% Combined Ratios. Management expects to break even at around 180-190k policies, which they were aiming for in 2023-24 at the time of the IPO, but will prove challenging given the difficult macro environment in Spain at the moment. Policy growth in health has stalled in the last few quarters. Lower profitability in the core businesses will make it harder to accelerate growth in health in the near term.

-  Management changes: as mentioned, long-time CEO left early in 2022 and was replaced by Patricia Ayuela. She has also been with LDA for some time, but most communication with investors is done by the CFO  for example, which can be frustrating some times. The head of the health insurance business also left earlier this year.

-   Premiums can remain depressed for longer than expected and irrationality may become the norm in the market in response to higher inflation as competitors hunger for premiums increase in a higher rates environment, preventing LDA from achieving growth and profit goals. I would argue the cycle remains embedded in the sector. Discipline at LDA has been good historically and I think incentives align management behaviour with it.

-  Upside risks are many and this is part of the attractiveness of the shares today.

Thoughts on valuation

Motor generates around €45/customer operating profit; Home close to €17/customer and  €50/average customer. With close to 90% weighted average retention and ca. €200 acquisition cost per policy, I estimate the value of a customer at around €470-500. The business today trades at €330/policy.

Following the introduction of Baremo in 2016, market premiums increased by around 3% and LDA was able to grow customers by 7% on average over the next three years. At the time, Baremo was the only inflationary pressure felt by the market. This time around, Baremo has already been adjusted twice in 2022 to reflect higher inflation, cost of repairs is creeping up and renewal premiums have yet to respond.

In Home, greater economies of scale will come with more customers. LDA finally managed to operate at a lower expense ratio vs. the market in 3Q22. This should continue.

The market appears to be extrapolating the 40% expected decline in operating profits in 2022 in the medium term.

Looking out to 2025, assuming:

-              Motor customers grow at 4% on average per annum with 2% premium growth,

-              Home customers grow at 7% with premium growth of 6% on average and

-              Health customers numbers grow at 24% with no premium growth

-              Investment yield is ~3.5% of portfolio (around €40m/year)

-              Combined ratios in Motor and Home returning to historical average by 2025 (86% in Motor, 90% in Home)

LDA would be generating close to €1.2bn in written premiums and €197 million of operating earnings, compared to €175 million in 2020. At a multiple of Net Income of 10x (vs. today’s 18x) and accounting for the €84 million in cash, I get to around €1.6bn of equity value, compared to €1.2bn market cap today. In the meantime, the company will be pay out around €300 million in dividends over the coming three years (26% of mkt cap)

The potential IRR is quite compelling at around 20% from current levels. More importantly the upside risk outweighs the downside risk should premiums began accelerating at a higher rate 2023 in Motor for example.

Alternatively, the business is trading at a 6.3% dividend yield today. A similar yield on 2025 earnings with a 90% pay out would lead to about €2.1bn of equity value, €1.85/share plus around €0.28/share in cumulative dividends.

Catalyst

Evidence of higher motor renewal premiums during 2023