David Pratt
Property General Manager at Progressive
Thanks, Pat. I'll review the results in the property insurance business that's underwritten by Progressive, and then talk about the changes we're implementing to position this business for future growth and consistent profitability. Pat described the huge market opportunity presented by the Robinson bundled customers. Since Progressive acquired a controlling interest in ASI in 2015, we have invested to leverage Progressive's brand and distribution channels to attract and retain Robinson customers. We have rebranded ASI's Progressive Home, expanded the Platinum agent program and built online quoting and buying capabilities, as Pat just described. These investments have helped us more than double the size of this business reaching $2.2 billion in direct written premium last year. We're now the 10th largest homeowners insurer in the United States. While we're happy with the growth of the property business, we are not satisfied with its profitability. We've seen a consistent pattern over the last five years. Our underwriting expenses and non-weather losses have been consistent with our forecasts but weather losses have been much higher than the estimates that we included in our pricing decisions.
In order to diagnose how to address this problem, we'll look at our performance relative to the overall property insurance market. I'll provide an overview of our countrywide performance and then highlight a state case study, focusing on the actions we've taken in Texas. When Progressive acquired a controlling interest in ASI in 2015, 64% of the company's homeowners premium came from Florida, Texas and Louisiana. We have labeled those as legacy CAT states in the map here. Growth in other states has reduced that legacy CAT mix from 64% to 41% in 2021 direct written premium. But that's still a much heavier mix than the 22% that these states represent of the overall U.S. homeowners insurance market. The challenges facing the Florida property market have been especially painful for Progressive since it represents almost 1/4 of our homeowners premium. Our Florida direct loss ratio over the last five years is four points better than the overall industry. But at 74.5%, it's still much higher than our target for the state. We have grouped the remaining states based on their exposure to volatile weather perils. The yellow states in the center of the country experienced significant hail losses.
Our mix of business in those states is consistent with their share of the total industry premium. The red states labeled moderate volatility are exposed to hurricane risk on the East Coast and wildfires in California, but there are less risky areas inland and away from the areas of the high wildfire risk. Progressive was a new entrant in the California market in 2016 so our mix of business in the moderate volatility states is less than their share of the overall industry. That leaves the less volatile states in green. As you can see, these more stable markets represented only 27% of our direct written premium last year, much lower than their 40% of total U.S. homeowners premium. In order to build a more stable homeowners book of business, we will limit growth in the volatile states and work to grow our share of premium in the less volatile markets. You've seen this growth and profitability chart in the past for our auto insurance business. It shows combined ratio on the X-axis and premium growth in the Y-axis. Each dot represents one of the top 10 insurers with the industry total represented by the black dot. We want to be in the upper right quadrant, representing profitable growth. Over the five years from 2017 through 2021, the U.S. homeowners insurance market grew at a 5% annual rate and suffered a combined ratio over 100. As we noted earlier, we're happy with Progressive's growth over that period, but we're certainly not satisfied with our position on the profitability axis.
I talked before about our heavier than industry mix of business in the legacy CAT states. If we adjust Progressive's results by weighting our state level results using the industry's mix of premium by state, our combined ratio would match the industry. Let's take one more look at our results over the last five years before moving on to our plans for shifting the mix to less volatile states. This chart shows our 2017 through 2021 direct homeowners loss ratio for the groups of states described earlier. As you'd expect from the mix adjusted comparison of our results against the industry on the last page, our loss ratios are close to industry results within each group of states. This has been a difficult period for the industry in states with significant exposure to catastrophic weather and our relatively high mix of business in those states has driven our overall property results to be worse than the industry. I should note that California accounts for 45% of total industry homeowners premium in the states that we've labeled moderate volatility. But California is less than 20% of Progressive's premium from those states.
While I'm happy to say that our loss ratio was better than the industry in five of those six states, our relatively low exposure to California's wildfires in 2017 and 2018, exaggerates our relative performance advantage for this group of states. It's encouraging to see that Progressive's loss ratio is very close to the industry average in the less volatile states since we have relatively little experience operating in these markets. We hope to see our relative performance improve as we roll out product model improvements. These regional comparisons simply reinforce the point that we need to shift our mix of business to build a property book that is consistently profitable. I'll talk next about the actions we're taking to accomplish that shift. As Tricia mentioned in her first quarter letter to shareholders, we have begun implementing a plan to shrink our business in Florida and limit growth in the other states with significant exposure to catastrophes. We've taken a number of actions in these states. We've increased rates, particularly for homes with older roofs; we've expanded underwriting eligibility restrictions to ensure that we're writing only business we expect to be profitable; we've introduced targeted nonrenewals in areas with concentrations of risk or segments that are unprofitable; we've introduced cost-sharing initiatives such as higher deductibles and actual cash value coverage on roofs that are nearing the end of their useful lives; and we've improved agency management, including a reduction in our agency force in select markets.
While these changes are important, they're not sufficient. We can't shrink our way to success. So we also need to continue investing in product improvements that will support growth in the less volatile states. Over the last four years, we've made significant investments to develop the organizational capability to build and deliver best-in-class property insurance products. As you can see in the graph on the right, our product teams have grown from just 21 people in September of 2018 to 134 people at the end of this year's second quarter. We have added auto-like capabilities to a team with significant amount of property experience and expertise. I'll briefly describe the new teams that we've added in the product area. In the second quarter of 2018, we introduced an underwriting analytics team created to advance the science of underwriting and improve our risk selection. In the fourth quarter of that year, we introduced a dedicated research and development organization to focus on product model development and segmentation improvements. In the first quarter of 2019, we introduced a dedicated pricing team. And finally, in the second quarter of last year, we introduced a new rate revision and product delivery team to improve our speed to market so that we can take these new model enhancements and roll them out more quickly.
These investments have helped us to improve our product offerings with more improvements still to come. In 2019, I described our 4.0 property product, which added new coverage features and expanded eligibility to meet the expectations of preferred agencies outside of our legacy CAT states. We have since introduced the 4.1 product model, which allowed us to price the water peril more accurately. These two product versions are now live in 41 states that represent 93% of our non-Florida premium. We'll move to full by-peril rating with the 5.0 product model. We've completed the research work for this new model and IT development work is in process now. The first state is scheduled to go live during the first half of 2023. We expect 5.0 to offer more competitive rates for a majority of shoppers and allow us to further expand the range of risks that we can accept. We will need these improvements to achieve our growth ambitions in the less volatile states. Now let's look at a case study for how these product improvements can help us improve results. In 2017, our Texas business was too concentrated in the Dallas-Fort Worth area, with that region representing 40% of our Texas business compared with roughly 20% of the single-family homes in the state.
We also had too many customers with low wind/hail deductibles, especially in the northern half of Texas that experiences frequent hailstorms. Our customers are often encouraged to file claims after these storms by aggressive marketing from roofers. While a free roof initially sounds appealing, we found that customers with higher deductibles are less likely to file a claim for minor damage. In order to address these problems, our Texas product manager increased rates, introduced higher deductible requirements for new customers, and improved price segmentation with the introduction of the 4.0 and then 4.1 product models. We have seen significant shifts in our Texas book of business as a result. The Dallas-Fort Worth mix is now in line with that area's share of homes in the state, our average wind/hail deductible is up 50% and we're attracting more bundled customers and fewer customers with roofs that are nearing the end of their useful life. These changes have resulted in improved performance relative to our competitors in Texas. In 2018 and 2019, our Texas combined ratio was 10 points worse than the industry. As we've shifted the mix of business, we started to see improvement in 2020 and our 2021 combined ratio was 15 points better than the overall industry in Texas. We were able to accomplish these shifts without shrinking the total business in the state.
We now expect to achieve our target profit margins over time or in an average year, but it's important to note that Texas homeowners results are naturally volatile due to the state's exposure to big hurricane and hail events. That's why we have decided to limit growth in Texas as part of our overall plan to shift our mix to less volatile states. Texas is now well positioned for profitable growth but we wanted to represent a smaller share of our total property premium over time. I'll close with two topics that many of you have asked about. The first is the effect of inflation on our Property business. The table on the left on this page shows our filed rate changes over the last 2.5 years in our core homeowners product. But that table doesn't fully explain the changes in ultimate premiums that our renewing customers experience. Let me explain that process in a little more detail. When we sell a new homeowners insurance policy, we estimate the cost to replace the home in the event of a total loss. We insist that the agent insure the property for at least that amount to ensure adequate coverage for the customer. We update this replacement cost at every renewal event. If the estimated replacement cost has gone up, we automatically increase the coverage limit to maintain adequate coverage. Until recently, these renewal adjustments typically increase coverage limits by 2% or 3% at each policy renewal. The blue line in the chart on the right shows the average year-over-year change in these coverage limits by quarter.
And you'll see that inflation-driven cost increases during the second half of last year jumped from the traditional 2% to 3% to double-digit increases in average coverage limits. The orange line shows the average renewal premium increase paid by customers who renewed with us in each quarter. So that orange line is a function not only of our filed rate change, but also the increases in coverage limit to ensure adequate coverage. And then any mix changes that might have resulted from some segments of customers renewing with us at a higher or lower rate as we've changed our product over time. While we would certainly prefer to keep prices more stable for our customers, this automatic adjustment to maintain appropriate coverage helps us to maintain appropriate rate levels to compensate for increases in claims severity. And finally, I'll provide a brief update on our reinsurance program. We use reinsurance to protect Progressive's capital in the event of an unusually severe catastrophic event in addition to protecting against unusually volatile property results. Progressive is a large and consistent purchaser of coverage using both traditional reinsurance and ILS markets. We've developed long-term trading relationships with our reinsurance providers, and we strive to keep them well informed about our business results and plans. We currently retain the first $200 million in loss from a single catastrophe event with reinsurance coverage available above that amount up to a limit of $2.6 billion for a single Florida hurricane.
We also have $175 million in coverage available for aggregate losses from 2022 catastrophe events in excess of $575 million. While the total coverage limit and per event retention will continue to evolve to fit the growth of our business, we expect to remain a consistent purchaser of reinsurance coverage. Consistent with this history, we were able to fully place our desired coverage at both the January one and June one renewal events. So let me summarize briefly before we take your questions. As Pat described earlier, a competitive property insurance product is required for Progressive to attract and retain bundled auto and home customers. And this segment represents a huge growth opportunity for Progressive. In the direct channel, our multi-carrier property agency and market-leading quote volume, are driving success. We're investing to improve the profitability in our underwritten property business to support continued Robinson growth in the independent agency channel. Through rate increases, coverage limitations and underwriting eligibility restrictions, we have improved results relative to our competitors. Best-in-class product design will support growth in the less volatile states, while a reduction in our Florida exposure and growth limits in our other states exposed to volatile catastrophe risk will reduce the volatility of our overall property, results, and we will continue to use reinsurance to protect Progressive's capital from extreme events. Thanks again for joining us this morning.
We'll now begin our live question-and-answer session