Karen Bailo
Commercial Lines President of Progressive Insurance at Progressive
We will be sharing a number of charts and graphs to highlight our business performance during this presentation. While our June results include Protective based on our closing date of June 1, for the purposes of this presentation, unless otherwise stated, the numbers exclude Protective.
I'd like to begin with a look at our performance relative to the industry.
This is a 20-year time series of Progressive commercial auto net written premium growth rate versus the rest of the industry. During this time period, our results have really diverged from the industry. There are a couple of observations around growth that I'd like to highlight. First, commercial auto has some cyclicality to it and tends to move with the larger economy; and second, when the market grows, Progressive has historically grown at a faster pace. In fact, since 2016, we've doubled the business and gained almost 4 points of market share and that equates to more than 20% compounded growth rate and achieving more than 12% market share.
The more significant divergence from the market has been in underwriting profit. We've outperformed the industry by 8 to 10, and as much as 20 points over those 20 years. Our objective is to grow as fast as we can at target combined ratios, and we've had a track record of success and outperformed the industry on both growth and profit over those 20 years.
There is a number of contributing factors. But, perhaps the most important has been the intense focus on commercial auto as a core line of business for the company. We've shared this information in the past and wanted to provide a brief refresher on how we approach our auto business. We segment our business into what we refer to as business market targets and we introduced these business market targets for commercial auto in 2014.
Since introduced, BMTs are now operationalized across virtually every aspect of our business. That's important because we see meaningful and actionable differences between these BMTs. For example, the demand function is different by BMT and how that demand function responds to changes in different economic conditions is different. We also see that losses present differently and how they develop, attorney representation rates and litigation outcomes differ by BMT. Certainly, frequency and severity trends, and other factors change at different rates and at different times by BMT, all of which are critical inputs to determining rate level.
This granular focus allows us to develop insights faster, be more responsive with strategies and tactics to prosper [Phonetic]. I want to talk about what we're observing in shopping, loss trends, and how we're positioned for continued success. We have the benefit of over 10 billion miles of driving data with our Telematics data. This chart shows patterns and driving miles, highway congestion and highway speeding, pre- and post-pandemic. You can see that as stay at home orders were issued, there was a significant decrease in miles driven and congestion on the highways. At the same time, there was an increase in highway speeding.
Now, while the impact from COVID on small businesses was severe, when we look at our business class level data, it's also clear that different businesses were affected differently. For example, in the truck space, fully a third of our Smart Haul customers saw their mileage increase while about 8% were shut down completely. Landscapers and most construction trades were still working and other service businesses were not.
Looking at more recent trends, while we see mileage and congestion back to pre-COVID levels, speeding events haven't dropped back to where they were. This raises the question around whether COVID brought a permanent change to truckers' driving habits and it's something we'll keep a close eye on in the months to come.
In addition to our own data, we look to other macroeconomic trends to develop a deeper understanding of shopping and small business trends, especially during challenging economic conditions. This chart shows our insurance shopping trends; in this case, agency quote growth for businesses that tie to goods and services sectors indexed to 2019. Progressive quote data is in the solid line and consumer spend data is in the dotted line. Our experience tracks closely to the rate of consumer spending on goods and services. So, this is data we have and will continue to monitor.
The macroeconomic data shows spending on goods recovered more quickly, while the services sector has lagged and has been more depressed relative to goods sectors. Some of that is because services sectors were more affected by stay-at-home orders and goods spending has been supported by federal government stimulus. That intuitively makes sense and we see that in our underlying data.
Businesses that were considered more essential, plumbing or sanitation services or those that tied to the transportation of goods like agriculture hauling or livestock hauling, responded differently versus those that are tied to services industries like airports shuttles, food trucks and entertainers. The positive news is that as the service businesses reopen, we're seeing a recovery in spending towards 2019 levels. This increase in spending on services should drive a rebound in insurance shopping for businesses related to that sector, again food trucks, restaurant delivery, airports shuttles, just to name a few.
In summary, different businesses have been affected differently and this is important in terms of how we think about trends, the implications for frequency and severity, and ultimately rate level going forward.
In our last update, in 2019, we shared a historical perspective that provides a good illustration of how we will approach today's environment. Back in 2016, we saw a marked increase in frequency between May and November. At the same time, some prior year reserve development was contributing to an already positive severity trend. We responded quickly to address those trends, raising rates by more than 10% and made a series of underwriting changes in about three months and we've continued with a series of changes and adjustments since.
Fast forward to recent trends, you can clearly see the impact of COVID on frequency. We saw a sharp decline in frequency as stay-at-home orders went into effect. And as I shared earlier, driving miles from our usage-based insurance data shows driving miles and congestion levels are back to pre-pandemic levels and the services sectors are also rebounding. Given all of those conditions, we're seeing lost frequency rising relative to COVID lows, when frequency fell dramatically. And at the same time, we continue to see a steady increase in severity trends that we've been accounting for in our rate level indications.
Given the variation and how businesses were affected by pandemic, some slowing down and others seeing an increase in business, we have been and plan to continue to be cautious in our actions. We haven't lowered rates and we've been conservative in our pricing and underwriting decisions. The trends we're observing now are lining up with what's forecasted in our rate level indications. We have planned for frequency recovery and have continued steep severity trends and have kept pace with net trends with the combined rate increase of 29% from the beginning of 2016 through the first half of this year.
The additional segmentation we've built into the product over the last five years has proven effective and driven better than industry underwriting results and growth. And I would suggest having a granular approach to the business and reacting decisively to what we see, while much of the market is slower to react, has been an important part of maintaining strong underwriting margins and growing the business over the years. So, we continue to advance our product segmentation and underwriting capabilities, and we plan to continue to respond appropriately to loss trends going forward.
I'd like to shift to a discussion on expenses and efficiency.
We know from experience that companies that can achieve a lower cost structure and gain share at a greater rate than the overall industry. We've seen that in the personal lines market and believe it matters in the commercial lines market as well. The correlation of efficiency aiding in growth isn't lost on us and we're well positioned on this chart.
2020 results show that we are nearly 11 points lower than the industry average in LAE and expense ratio. In a very price sensitive industry, that 11 point advantage is significance. But, there is a balance. We don't necessarily want to be the lowest because we believe in also investing in what matters, quality outcomes for customers that earn loyalty and investments that foster great work environment for our employees.
To that end, I'd like to highlight where we're making investments to improve on both those fronts. This is a view of our expense ratio track record over a 10-year time period. And over that time period, we've seen a 7 point to 8 point expense advantage compared to the industry. That advantage was 6 points in 2020, due in part because COVID credits flowed through expenses rather than premium, resulting in a slight elevated expense ratio.
While we've had an expense advantage over this time period, we're prioritizing initiatives to extend our leadership position. There are a number of levers that drive expense ratio advantages. And while we don't have enough time for a comprehensive review of all the efforts underway, I wanted to share with you two examples that demonstrate active cost management efforts to drive expense reductions and improve experiences at the same time. These both highlight our focus in managing expenses related to our growth and improving our operational efficiency.
We maintain a disciplined focus on managing overhead and growth in headcount as we grow the business. This chart is designed to represent a few things. The solid blue line represents overall net earned premium growth; the solid orange line represents employee costs; and real estate costs are represented in the solid black line. You can see that while employee costs and real estate costs have gone up, they've grown at a lower rate relative to net earned premium, and there are a couple of reasons for that.
One is being disciplined and judicious in decisions to increase staff. We've added the dotted lines here to represent volume-driven and non-volume driven employee growth. And while we've grown our volume-driven counts in line with our net earned premium pace, we've been very targeted in adding non-volume-driven resources and our non-volume driven employee count has grown at a much lower rate. This discipline has resulted in growing our total employee costs, the orange line, less than net earned premium growth.
The second is related to real estate. Like personal lines, prior to the pandemic, our customer and agent services organization enabled real estate expense savings by implementing a home-based consultant model. This model has a number of benefits. It provides broader access to talents and improved our ability to increase staff to support our growing business, especially last year, and it's also provided flexibility that employees value. In addition, this model allows us to grow our business without a commensurate growth in space. In 2019, almost 40% of our commercial lines, agents and customer services consultants were working from home. Now, we're still working through our plans as we transition back to the office post pandemic.
We expect the portion of our customer and agent services consultants that work from home to grow materially, as people make the decision to remain working from home going forward. This will support our ability to grow without significantly growing our real estate footprint.
The second example of our cost management efforts highlights investments in systems and technology to deliver the products and services our agents and customers value, while focusing on increasing operational efficiency. We've been investing significantly in systems to support our core auto business and in expanding our business with our BOP product and direct small business capabilities. Despite significant investments in technology and systems, we've been able to maintain our expense advantage. And we acknowledged and expected this relatively short-term increase in technology costs to lower our costs in the long run with gains in efficiency and new lines of business.
Now, improving efficiency to push our expense advantage is a key objective. Two examples of progress are shown here. We've shared in previous Annual Report commentary that we're transitioning to a new policy administration system. And this new system is a significant driver of our technology expenses and pulling out the new system is a big part of our ability to drive more efficiency. The new system introduces more modern functionality that enables faster delivery of our products and enhancements, and the ability to improve customer experiences and self-service capabilities.
The initial launch of this new system in a state [Phonetic] brings consumer online buy capabilities. The chart on the left shows the total lift in direct online sales yield with our new system. Now, completely isolating the effects of the online buy capabilities is difficult, but we are seeing an 80% improvement in online sales yield after introducing this new system. And that would translate to more than a 20% increase in direct auto sales.
Now, while direct is still a relatively small portion of our business, we expect this added functionality to bring long-term economic benefit to us in terms of improved sales yield and lower acquisition costs for that part of our business. And that's very important as this business grows.
A second way we're gaining efficiency is via process automation.
On the right is a representation of improvements and we've already made reducing the manual work associated with millions of documents we receive every year via fax, email, or paper mail. Until recently, each document had to be manually reviewed and categorized for action, either to be attached to a policy and sent directly to storage or put in queue for additional processing. Over half of these documents are in that first group. They don't require any action beyond archival.
We have a project underway to automate that workflow by the end of the year using optical character recognition technology to review those documents automatically and attach them to a policy and send them to storage without human intervention. And this frees up resources to focus on other more value-added work. These are just two examples to illustrate how we will focus future investments.
Now, while we're pleased with our early progress, we're just getting started, and we have a robust road map ahead of us that will target more efficiency improvements and self-service capabilities designed to meet customer and agents' expectations. Investments in products, experiences and more efficient workflows combined should enable us to reduce the drivers of cost to service our customers. We anticipate that by lowering our costs, we will be able to further extend our expense advantage and position us well for any market conditions.
Now, moving onto the other part of our expense advantage. We believe our claims organization provides us with a significant competitive advantage on commercial auto vis-a-vis our competition. We've doubled the core auto business, while maintaining very competitive loss adjustment expense ratios and good quality. Our claims advantage comes from a surgical focus on claims accuracy and efficiency.
And back in 2019, we shared how our claims organization leverages the scale of our broader personal lines claims organization, with a focus on specialization for high impact and complex claims. We continue to push commercial lines specialization as we grow our business and we also see significant potential in leveraging technology and analytics to increase productivity and claims handling segmentation, along with improving accuracy.
I wanted to share a brief example to illustrate how we're leveraging data, data science, and advanced analytics to monitor and react to claims activity and exposure changes.
The image on this slide represents a home-grown tool developed to inform leaders of claims that need review. The tool is powered by a file intervention program that uses analysis and data science to populate the tool and is designed to improve a leader's focus on at-risk files. This helps improve claims accuracy through better coaching and supports timely claim handling efforts. We know handling delays can result in unfavorable outcomes and the alerts are prioritized and triggered on the best date for the leader to intervene for the most favorable results.
In this example, you see a code, FS13 with a description potential for delayed total loss resolution. In this example, an alert triggered on March 19 with a message indicating a potential delay in resolving a total loss settlement. The leader intervenes by providing guidance to the claim rep on the open activity on the file. And since the leader intervened, the trigger is programmed to return two days later to make sure the claim rep is following up on the guidance provided. The next alert triggers on March 21. The leader reviews the file and sees that the claim rep followed up on the direction provided, so no additional intervention is needed at this point in time. The leader indicates no to the intervention question in the lower left. And once no intervention is selected, an alert is programmed to return eight days later if the total loss is still not resolved. In this case, the total loss resolved on March 25 and so the alert didn't return for the third time.
This is a powerful tool that helps our claims teams efficiently run the business of claims and contributes to more timely and accurate claims handling.
A final point on our performance relative to the industry is related to reserving where our philosophy is to be as accurate as possible with minimal variation. This chart shows our commercial auto reserve development versus the industry and we have a much tighter variance, and we see two primary reasons for this. One contributing factor is the claims organization we just talked about. Leveraging technology and advanced analytics to increase productivity and get the right claims to the right resource more quickly leads to more timely and accurate claims handling and better outcomes.
The other contributing factor is the highly segmented approach we take to loss reserving for commercial auto as we do with our other products. We've operationalized that BMT structure in addition to the usual loss cost cuts and it allows us to see pattern changes sooner and react appropriately. Having more consistent and predictable loss cost estimates through accurate reserving lets us understand our true ultimate costs faster and be more responsive with pricing and product refinements. So, that's a little background on what we've been seeing in our commercial auto results recently and why we've been able to produce some different outcomes relative to the industry.
I'd now like to turn it over to Jochen Schunter, who will share a little about our plan for extending our leadership in commercial auto, including how and where Protective Insurance fits into those plans.
