PGR: Progressive (1)
Business Model
PGR is mainly an automobile insurer operating in the US market. Founded in 1937 in Ohio, it was listed in 1971 at which point it stated an objective of growing while achieving maximum combined ratio of 96%. This remains true today and is generally achieved or exceeded. The average loss ratio from 1987 to 2022 was 92.8%, and it is worth noting that since 1987 there was only one year that PGR made underwriting losses in 2000. Even during 9-11 (2001) and Hurricane Katrina (2005), PGR was profitable.
The net earned premium (NEP) is split into 3 lines of businesses. Personal (mostly personal cars), Commercial (small business trucks and cars) and Property (homeowners’ insurance). In 2022, proportion of NEP was Personal 77%, Commercial 18%, Property 5%. Historically, Property business line was unprofitable with combined ratio averaging 104%. However the other 2 major lines of businesses generate large underwriting profits to offset.
Alongside with Berkshire’s GEICO, it is the only other auto insurer that has a significant direct auto insurance business. Market share is usually tied with GEICO, with State Farm being the market leader.
Competitive Advantages
PGR primary strength is the ability to more accurately appraise individual risk and so better “match pricing to risk”. This is shown in the impressive multi-decade yearly combined ratio of below 100%. This is in the context of the insurance industry that generally loses money on underwriting and relies on investment income for profitability.
For the last 10 years (excluding the exceptionally good COVID year), the average combined ratio was 95%, comprising of loss ratio 73% and expense ratio 22%. Competitor GEICO achieved slightly worse off at 96%.
GEICO is often referenced as the low-cost operator in the industry. It is true that its expense ratios are lower than PGR. However, PGR wins greatly in loss ratios.
How does PGR manage to price risk better?
It was the first auto insurer to recognise the significance of credit ratings and incorporated this to its risk pricing.
It was early to embrace telematics. Risk pricing is a constantly evolving game of finding small data and analytical edges and incorporating these as tools with marginal gains into the system.
There is a positive feedback loop whereby an insurer with a risk pricing edge is able to sell insurance at an attractive price to customers that competitors over-charge (thereby gaining a profitable customer), while avoiding selling insurance to customers at too low a price (thereby pushing loss making customers over to competitors who price the risk too low).
At the 2019 Berkshire Hathaway Shareholder’s Meeting, Ajit Jain, Vice Chairman of Insurance Operations commented:
GEICO has a significant advantage over Progressive when it comes to the expense ratio to the extent of about 7 points or so. On the loss ratio side, Progressive does a much better job than GEICO does. They have, I think, about a 12 point advantage over GEICO. So net-net, Progressive is ahead by about 5 points.
PGR has a robust loss adjustment operation that is efficient in ensuring they pay out only for genuine loss claims. They have hundreds of claims offices around the country and they often benefit from having local knowledge of how to manage claims.
It is reassuring that the current CEO, Tricia Griffith, started her career at PGR in claims management and not in sales.
The other key source of competitive advantage is that PGR is one of only two scaled direct auto insurance sellers in a market in which there is limited online price comparison via aggregators.
It is interesting that the US market is very different to many others in that price comparison websites have never taken off in the US despite repeated attempts. This is partly from the resistance of the larger insurers and also from complexities in the regulatory landscape which are particular to each State.
Direct selling comprises ~25% of the market, with the rest made up of agents. Direct selling is growing and slowly gaining share (it was 10% in 2000) but unless one has a strong brand and a very large marketing budget, such that customers come directly to your website, it is difficult to do this profitably. Barriers are lower in markets such as the UK where anyone can reach customers via price comparison websites.
The #4 player Allstate tried to develop a direct business by acquiring Esurance, a relatively small direct business, in 2011. But they largely gave up after incurring significant losses from investment in customer acquisition, in an attempt to build the brand.
GEICO and PGR marketing budgets are huge, both spending ~$2b annually on advertising. It seems likely that the direct channel will be a long-term source of growth, as it captures further share and that it will only be enjoyed by these two players.
PGR selling structure is about 50% direct and 50% agents, whereas GEICO is purely direct. PGR have invested to improve their position in the agency market. In 2016 they acquired a Homeowners insurer, so that they could offer bundled Auto & Home Insurance via agents. Those customers who have insurance products bundled with one insurer are more likely to stay with the insurer.
Management
PGR is the only company we are aware of that reports an abbreviated P&L and balance sheet, with operational metrics (customer count, combined ratios), on a monthly basis. We like to think that this suggests a transparent and confident organisation.
PGR incentivises employees through an unique and long-standing incentive plan called “Gainshare”. The performance measure is on growth (policies in-force) and profitability (combined ratio). Note that the “growth” measure is not on new business sales. This creates incentives for underwriters to keep profitable accounts in the books and refuse unprofitable ones. You can see the details of their KPI scorecard here.
Risk
There is a risk for PGR’s terminal value as autonomous cars will in due course reduce need for insurance. In time, autonomous cars will reduce the number of car accidents, and to the extent insurance is provided.
The question is about timing and what this means for value. The barrier to adoption is as much regulatory and psychological as it is technological. For regulatory approval and widespread adoption, self-driving cars need to be significantly safer than human-driven cars because higher standards are demanded from machines.
In terms of valuation, by year 2030, $1 of cashflow is worth $0.4 today (our discount rate is 10.5%). By 2050, when autonomous cars may well have materially impaired the auto insurance industry, $1 dollar of cashflow is worth $0.06 in today’s money. The intrinsic value calculation is more a function of the next 10 years when the autonomous threat is not that material.
Combined ratios may not improve back to the 36 year average. This is due to more people on the roads and hence naturally higher accident rates. However, we think that the general population tends to become more wealthy over time and the cost of auto insurance for people who can afford to own a car is a small fraction of their incomes. Therefore we believe that the industry as a whole is able to raise prices in response to this problem.
Valuation
We assumed in our valuation model that the latest reported year 2022 combined ratio of 96.3% will gradually improve back to long run average 92.8% in year 2032.
Current share price is $139, market cap $81.1b.
We approach valuation with DCF model, deriving the free cash flows from P&L. The main assumptions are:
Combined ratios improving from 96.3% to the long run average 92.8%
Investment income grows at 2.5%
Debt to equity ratio at 40% with interest expenses at 4.4% of debt
Tax rate 24.5% and no fair value movement (FV movement is removed from cash flow)
Discount rate 10.5%
We arrive at intrinsic value of $132 per share, quite close to what the market is offering now.
It’s interesting to note that the price to book is 5.1x, which is high for an insurance company, and the default valuation metric for insurance companies is usually a low price to book.
Believe it or not, there is actually a similar question raised during 1996 Berkshire’s AGM.
Question: “… Berkshire is unique in that equity assets far outweigh fixed income assets… GEICO upon entering the merger looked like a typical insurance company with four times fixed income assets compared to equity assets. How are decisions in asset allocation going to be made in GEICO after the merger?”
Warren Buffett: After some words of praise for Lou Simpson (then manager of GEICO), he replies, “We let Lou Simpson do whatever he wants… Lou did not have that ability before GEICO became part of Berkshire… Now after merging it allows GEICO to have more freedom in asset allocation, because Berkshire can guarantee the liabilities of GEICO…”
Charlie Munger then goes on to add, “that was a shrewd question, you ought to be complimented.”
That meant a price to book valuation made sense because GEICO’s ability to generate returns became relatable to book value after the merger.
On the other hand, PGR has a market cap of $81b and writes $50b of NEP. It won’t be appropriate for PGR to go beyond a certain allocation in equities. Most of PGR invested assets are fixed income backing the loss reserve liabilities. When we buy PGR, we are buying it’s ability to underwrite profitable business and produce cash flow; book value in this case does not relate to PGR’s ability to produce cash flow.
We will be buying PGR at current price of $139/share.