Good morning, and thank you for joining us today for Progressive's Second Quarter Investor Event. I am Doug Constantine, Director of Investor Relations, and I'll be the moderator for today's event.

The company will not make detailed comments related to its results in addition to those provided in its annual report on Form 10-K, quarterly reports on Form 10-Q and the letter to shareholders, which have been posted to the company's website.

This quarter includes a presentation on a specific portion of our business followed by a question-and-answer session with members of our leadership team. The introductory comments by our CEO and the presentation were previously recorded.

Upon completion of the previously recorded remarks, we will use the balance of the 90 minutes scheduled for this event for live questions and answers with leaders featured in our recorded remarks as well as other members of our management team. As always, discussions in this event may include forward-looking statements.

These statements are based on management's current expectations and are subject to many risks and uncertainties that could cause actual events and results to differ materially from those discussed during today's event.

Additional information concerning those risks and uncertainties is available in our annual report on Form 10-K for the year ended December 31, 2022, as supplemented by our 10-Q reports for the first and second quarters of 2023, where you will find discussions of the risk factors affecting our businesses, safe harbor statements related to forward-looking statements and other discussions of the challenges we face.

These documents can be found via the Investor Relations section of our website at investors.progressive.com. To begin today, I'm pleased to introduce our CEO, Tricia Griffith, who will kick us off with some introductory comments.

Tricia?.

Florida and fixing vehicle coverages. Florida's prior year adverse development accounts for just over 40% of the 4 points of prior year development.

While we continue to be optimistic that March's new legislation will result in lower loss costs over time, the short-term effect has been a significant increase in attorney representation largely in medical coverages.

Our data has matured each month since the bill passed, which has allowed us to continue to fine-tune reserving in the face of changes to Florida's ultimate loss costs. In fixing vehicle coverages, unforeseen severity trends deepening on previously closed claims continues to be the major driver of prior year adverse development countrywide.

The contributors to the steeper trends come from a variety of sources, including longer vehicle repair times, longer rental times, higher parts prices and labor rates and changes in subrogation trends.

Fixing vehicle coverages are short-tailed, which explains why over 80% of the total year-to-date prior-year development is from the 2022 accident year. In fact, excluding Florida, accident years prior to 2022 have developed favorably. Our current year actuarial adjustments are primarily due to fixing vehicle coverages.

Since fixing vehicle coverages are short-tailed, most of the claims that are affected by emerging trends have happened in the recent past. As we have moved deeper into the year, the new steeper trends in fixing vehicle coverages have increasingly affected claims that occurred in the current accident year.

Florida litigation has been a relatively small part of current year actuarial adjustments since the increase in attorney representation rates largely affected claims that occurred before March of 2023. We continue to monitor loss trends across all states and coverages and will adjust if necessary to ensure we're always adequately reserved.

While both reserving and pricing have been a challenge this year, I have great confidence in the teams and our strategies, which have delivered results better than the industry for a very long time. To showcase these skill sets, this morning, we have 2 senior leaders to speak to these 2 topics for our presentation.

First, Gary Traicoff, our Corporate Actuary business leader and Head of our Reserving Group, will discuss reserving practice at Progressive and how we endeavor to accurately set and adjust reserves. Gary is a 28-year progressive veteran and has led our reserving group for almost 11 of those years.

After Gary, Jim Curtis, our Personal Lines Controller, will discuss cohort pricing. Cohort pricing is our term for managing to both our calendar year and lifetime 96 target when new business and renewal business have different expense and loss characteristics. Jim has been with Progressive for 28 years and in his current role for 4.

Again, thank you for joining us this morning. I will now pass you to Gary.

Gary?.

Thank you, Tricia, and good morning, everyone. I'm excited to talk to you today about loss reserving at Progressive. I will first cover some key definitions and metrics in our general process, then walk through an example on how our overall reserve levels are determined.

And finally, walk through an example of the reserving life cycle of an individual claim and how it would flow through the actuarial categories of the earnings release. Let's start off with the size of our loss and loss adjustment expense or LAE reserves.

The columns represent our carried reserve balance at the end of each calendar year from 2013 through 2022 and at June 30, 2023, for the rightmost column. On the left side of the graph, we can see that in 2013, Progressive had about $8.5 billion in reserves on a gross of reinsurance basis at year-end 2013.

This would be the sum of the orange and blue sections within the bar with our growth as a company, reserves before reinsurance as of June 30, 2023, is now almost $33 billion and almost 4x greater than what it was in 2013, while our premium growth was a little less than 3x greater over the same time period.

Much of this difference in reserves outpacing premium growth relates to our growth in longer-tailed products and commercial lines over this period. This means that we believe for all accidents that have occurred through June 30, 2023, our ultimate future liabilities will be almost $33 billion.

The orange section of the bars represents how much of our reserves are ceded for reinsurance. Some examples of this include state-regulated plans, such as Michigan Catastrophic Claims Association and the Florida Hurricane Cat Fund and private reinsurers. Our ceded reserves now make up about 16% of our gross reserves versus just over 12% in 2013.

Much of this increase in the higher proportion of ceded reserves relates to newer products that have been added over time, including property and protective, which have more reinsurance needs than our historical products.

The blue section of the columns represents how much of our reserves are retained after reinsurance, almost $28 billion as of June 30, 2023. Next, let's look at some metrics on how our reserves are distributed. The chart on the left shows that 67% of our net reserves are in loss case, 17% loss IBNR and the remainder under LAE.

Any open claim will be assigned a case reserve that may be set by a claims adjuster or an actuarially derived estimate set by the actuarial team. We will discuss this in more detail in a future slide. IBNR, which stands for incurred but not reported, makes up about 17% of our net reserves. IBNR may have different meanings across companies.

For most of Progressive's products, one can think of IBNR as the future liabilities to cover claims that have already happened and are currently not recorded as open case reserves. This would include late reports, reopens and salvage and subrogation recoveries.

Anticipated salvage and subrogation recoveries can be thought of as a contra reserve and is a reduction to future liabilities. For most of our products, IBNR does not cover any anticipated development in case reserves.

This is because since our case reserves are a combination of adjuster estimates and actuarially set case reserves for the majority of our products, our IBNR reserve, does not cover expected case development.

LAE reserves make up the remaining 16% and typically cover the anticipated expenses needed to settle claims that have already happened, such as defense counsel costs and claims adjuster salaries. The middle chart shows the distribution of our reserves by product.

As expected, reserves will index more to longer-tailed products with higher limits relative to premium size. Thus, our Commercial Lines business makes up almost 1/3 of our total reserves, but a smaller proportion of premium.

The graph on the right shows that in personal auto, 83% of our reserves are set to cover injury and medical costs and 17% to cover the cost of fixed vehicles.

Again, this is due to the longer-tail nature and higher severity costs to injury and medical coverages versus fixing vehicle coverages and reserves are over-indexed to injury and medical relative to premium. In addition, fixing vehicle coverages have a higher proportion of contra-reserves from salvage and subrogation compared to injury and medical.

One additional point. This data is as of year-end 2022 and is on a net of reinsurance basis. While the previous slide gave a high-level overview of the distribution of our reserve mix by various parts, we review the reserves at a much more granular level.

We typically review reserves by some combination of product, geographic areas such as group of states, individual states and even down to a region or subset within a state. Further breakouts typically include individual line coverages such as bodily injury, uninsured motor bodily injury and property damage and even by limit.

To determine these breakouts, we weigh the credibility of the data, homogeneity in the data and additional value of segmenting the data at a finer level. For example, in personal auto bodily injury, almost all states are reviewed separately with the majority of states reviewed quarterly.

State-level data is credible and many states have different development patterns, which warrant the data to be reviewed separately. In personal auto collision, while state-level data is credible, many states exhibit similar data patterns and several clusters of states are reviewed together.

About 25% to 30% of our reserves are reviewed at this in-depth level monthly and about 85% of our reserves quarterly. About 700 reviews, which is some combination of product, state, line, et cetera, are reviewed over the course of the year.

By reviewing reserves at this detailed level, we are able to adjust reserves more accurately at a granular level, which helps our pricing teams better match price to risk. We also have a robust roll-forward process that we will discuss later on, which allows up to 100% of the reserves to be adjusted monthly.

The roll-forward process allows us the time to do a deep analysis on a portion of the reserves monthly while still feeling comfortable that all of our reserves are being updated monthly. Here is an example of how an in-depth review schedule will look over 3 months. In the first month of the quarter, 57 reserve reviews were completed.

Each reserve review is either a loss or LAE review, which is further broken down between defense cost containment or DCC and adjusting and other expenses or A&O.

Some combination of products, such as personal auto, CL or one of our core commercial auto products or TNC, which represents transportation network companies, line coverages such as bodily injury and some combination of state is reviewed. These 57 reviews would make up about 25% to 30% of our total reserves.

In the following month, 60 reviews were completed. Again, there is some combination of type, product, coverage and state, but all 60 reviews are independent of reviews completed during the first month and represent another 25% to 30% of our total reserves.

Finally, the third month of the quarter had 68 reviews completed, which are all different from any reviews completed during the previous 2 months and represent an additional 25% to 30% of the total reserves. Over the course of the quarter, about 85% of the reserves were reviewed. Typically, between 50 to 70 reviews are completed per month.

Now let's talk about how the overall reserve level is determined. We use several different methods. One of the methods that we use is called the accident period chain ladder method in a standard industry practice.

And the example above, each row represents a 6-month period when the accident occurred and each column represents how much in paid losses had been paid out for those accidents at certain time periods after the accident. For example, the top row represents all accidents that occurred from July 1, 2021 through December 31, 2021.

The $100 in Column 1 means that the $100 was paid out as of December 31, 2021 for those accidents. Moving to the right of the same row. Column 2 of $120 means that for accidents that occurred during the last 6 months of 2021, as of June 2022, cumulatively that $120 had been paid out. Thus, an additional $20 was paid out during the first half of 2022.

The increase of $20 could be due to payments on known claims, late reports or reopen activity. The increase of 20% from $100 to $120 is shown in the loss development factors section in the December 2021 row and column 1.

Similarly, the loss development factor of 1.08 in Column 2 of the December 2021 row is calculated by taking the paid loss of $130 in Column 3, divided by the paid loss of $120 in Column 2 for the December 2021 row.

Next, the actuarial analyst selected a 1.27 in the first column as the predicted Column 1 loss development factor for the accidents happening in the 6-month period ending June of 2023. We can see that in the June 2023 row, $120 had been paid out by the end of June 2023.

The analyst is predicting that paid losses will increase by 27% or $33 during the second half of 2023 and cumulatively be at $153 by December 2023. Multiplying the $153 by an additional 8% predicts that an additional $12 will get paid out by the third column for a cumulative of $165.

Thus, the shaded numbers in blue are predictions of what will be paid out in the future. The indicated reserve is then calculated by taking the ultimate estimated paid losses minus what has already been paid as of June 2023 since reserves are to cover future liabilities.

The indicated reserve is $45 for accidents happening in the 6-month period ending June 2023 and $13 for accidents happening during the second half of 2022. In total, the indicated reserve is the sum of all accident periods or rows and adds up to $58. This example illustrates an accident period chain ladder approach using paid data.

Triangle, similar to this byproduct and grouped by line coverage can be found in our annual report. Other approaches that we typically employ include using case incurred data, developing severity and frequency separately and stratifying the data by size of loss layer.

In addition, data is also segmented by late report and reopen lag time from the date of loss to the date of late report and reopen and hindsight testing to prior reserve amounts are considered as well. Additional ad hoc analysis is typically completed to understand changes in the underlying data and loss development factors.

Based on the indications from the multiple reviews completed during the month, the actuarial team will then decide how much of an overall reserve change is needed and adjust reserve factors accordingly.

Those changes will show up in the earnings release under the actuarial adjustment section and be reported showing the breakout for the current accident year and all prior accident years.

For example, in June of 2023, based off the scheduled reserve reviews, which represented between 25% to 30% of reserves, the actuarial team increased reserves $130.9 million for accident years 2022 and prior and increased reserves $160.3 million for the 2023 accident year for a total change of $291.2 million.

The $130.9 million increase to prior accident years directly impacted the prior accident year development. In addition, due to other changes not related to actuarial reviews, prior accident years developed an additional $6.9 million unfavorably for a total amount of unfavorable development of $137.8 million.

Next, we will discuss what are some of the other items that are in all other development. Now let's go through an example of the pathway of a case reserve over time and the impact on our monthly earnings release. Remember, case reserves make up about 67% of our total reserves.

In this example, a claim happens on December 15 and is reported and opened in our claims system on December 19. Since the claim was reported in the same month as the occurrence of the accident, this would not be considered an IBNR claim.

The adjuster does not have much information about the claim at this point and it sets a case reserve of $5,000, noted by the orange dot. In parallel, a tabular case reserve is determined for this claim from an algorithm used by the actuarial team.

The actuarial team uses predictive modeling techniques based off of certain characteristics of the claim and policy to predict what the expected cost would be for all claims that meet the criteria. In personal auto, there are over 200,000 unique combinations of tabular case reserves.

It is important to note that the actuarial algorithm does not expect this one claim to pay exactly the tabular case reserve. But on average, all claims that meet the similar criteria should pay close to the tabular case reserve. This is similar to a pricing rate order of calculation formula. In this example, the tabular case reserve is $12,125.

Finally, the adjuster case reserve as compared to a reserve threshold set at $25,000 in this example. If the adjuster estimate is below the reserve threshold, the tabular case reserve is booked to the general leisure as the financial case reserve for that claim and is reflected on the balance sheet.

If the adjuster case reserve is greater than or equal to the threshold, the adjuster case reserve becomes the financial case reserve booked to the general ledger. The logic here is that the predictive modeling approach is fairly accurate for lower dollar claims.

For more severe claims, the adjuster has received more specific information pertinent to the claim and will predict better than the model for higher dollar claims. This combination of adjuster and tabular reserves provides more consistency and accuracy on a monthly basis.

The threshold varies by several variables, including product, line coverage and limit, just to name a few. Since the adjuster case reserve is $5,000 and below the threshold, the tabular case reserve of $12,125 is used in the general ledger.

In addition, $1,000 of IBNR is set to cover claims that happened in December and at some point after December, would either be late reported or reopened. The IBNR reserve is not attached to any one claim and instead, set as a factor related to earned premium to cover all potential late reported or reopened claims.

The total reserve liability on the balance sheet is $12,125 per case, plus $1,000 for IBNR, equaling a total of $13,125 denoted as the dark blue dot and will be the reserve booked at the end of December. In January, the adjuster has kept their initial estimate of $5,000.

Since this is below the $25,000 threshold, the tabular case reserve will again be booked to the general ledger. As part of our roll-forward process, all tabular case reserves can be adjusted monthly by an aging factor and inflation factor.

As the claim is 1 month older, the actuarial algorithm predicted that all claims remaining open would have a higher ultimate cost. In addition, all tabular case reserves that are not part of an in-depth actuarial review during the month gets inflated as well. The new tabular case reserve is 12,376, which is an increase of $251.

A portion of this $251 increase is aging and a portion is inflation. This will show up in the earnings release as unfavorable prior accident year all other reserve development in January.

It is important to note that other claims that settled in January may have settled for more or less than the December tabular case reserve, which would also show up in the all other development category of the earnings release. In February, the adjuster has received new information and increased their estimate to $15,000.

This amount is still below the threshold and thus, the tabular case reserve is still used to set the case reserve liability. In February, this claim was part of the 25% to 30% of reserves reviewed at an in-depth level by the actuarial team.

The actuarial team increased the tabular case reserve for all claims that have a similar criteria as this claim to $13,500, which is an increase of $1,124 from January. This increase will show up as unfavorable prior accident year actuarial adjustments in the earnings release. In March, the adjuster increased their estimate again this time to $20,000.

Since $20,000 is still below the threshold of $25,000, the tabular case reserve is still used as the case reserve book to the general ledger.

The claim becomes plaintiff attorney rep in March, Thus, while the tabular case reserve will be inflated and aged again, it will also change due to becoming plaintiff attorney rep as this is one of the variables used in the actuarial algorithm. The tabular case reserve is now $20,050 which is $6,550 greater than the prior month.

A portion of this $6,550 increase is due to inflation, a portion due to aging and a portion due to the claim becoming plaintiff attorney rep. The $6,550 will show up as unfavorable prior accident year all other development in the monthly earnings release.

In April, the adjuster has increased the reserve again to equal to the reserve threshold, the adjustor's estimate is booked to the general ledger as the case reserve for this claim. This is an increase of $4,950 in the case reserve on the balance sheet, moving from the tabular reserve at the end of March to the adjusted reserve at the end of April.

The 4,950 will show up as unfavorable prior accident year all other development in the monthly earnings release. The adjuster made no change to their estimate in May. Case reserves set equal to or over the threshold can only be changed by the adjuster revising their estimates.

Thus, there are no additional changes for inflation, aging or any other factors used in the actuarial algorithm. In June, the claim closed out and settled for $20,000. This will show as $5,000 favorable in the prior accident year, all other development section of the monthly earnings release.

Finally, in July, the claim was reopened an additional payment occurred for $1,500. A claim could be reopened for several reasons. For example, a claimant may have initially been undecided about a vehicle repair and received a cash settlement based upon the initial estimate.

Several months later, they may have ultimately decided to repair the vehicle, and additional damage may have been found. Another example is when a claimant may initially feel that they are not injured in an accident, the claim is closed and subsequently reopened when the person realizes that they were injured in the accident.

While the case reserve had closed out in June, we were still carrying $1,000 of IBNR reserves for the potential of a December claim being late reported or reopening.

For this example, the IBNR reserve was released in July, which means that there is an expectation that there will be no future costs from late reports or reopens from any accident that occurred in 2022 and is currently not an open case reserve as of the end of July. The development is unfavorable $500.

This would show up in the prior accident year all other category of the earnings release. In practice, the IBNR reserve for prior accident years tends to get released monthly as part of another actuarial algorithm as the probability of late reports and reopens decreases with time.

An in-depth review of IBNR reserves are completed in tandem with case reserves. Remember, any changes made to factors by the actuarial team as a result of those reviews will show up under the prior accident year actuarial category. For the year, the starting reserve was $13,125 and the claim was eventually paid for $21,500.

The year-to-date development is unfavorable by this difference, which is $8,375. I as the Chief Actuary, have complete decision-making authority on our reserves. And while the actuarial team has complete independence in determining the reserves to be booked, there are many checks and balances in place.

Internally, we have a close partnership with our claims, pricing and product management business partners. Meetings are held throughout the month at both a localized and national level discussing trends, changes in the mix of business, claims process changes, rate activity and methodologies.

Quarterly, I meet with the Audit Committee members of the Board of Directors to discuss results and trends for the quarter. Furthermore, our external auditors perform an annual audit of the company as required by the SEC.

My actuarial team meets with the audit firm to discuss our actuarial process, current trends and results for the respective quarters and year. I hope that you have found this presentation helpful. For more information on Progressive's loss reserving practices, we posted a report under the investor site of progressive.com.

This report is updated every 2 years. Thank you for your time today. And I will next turn it over to Jim to discuss cohort pricing..

loss in LAE, acquisition expense and operational expense. Loss and LAE represent our indemnity experience. We break out our expense ratio into acquisition and operational expenses. And in all cases, renewal business operates at a total lower cost structure than the first term or our new business.

While we price to the lifetime profitability of our policies at a 96 or better, we also manage this approach with consideration to our calendar year goal of meeting a 96 or better-combined ratio. Next will explore each individual category or line item and finish with our management of cohort pricing within our calendar year 96 goal.

Let's start with loss and LAE expense. Loss ratio improves from new business to renewal and is primarily a function of mix of business and policy life expectancy. And we priced to that expected lifetime loss ratio performance based on the characteristics at new business, which ensures the business written is priced to a lifetime target.

Starting in the upper left in this illustrative example, you can see that our first-term business runs at a loss and LAE ratio of 85% with our renewal book of business at 77%. With an expected policy life of 6 policy terms, the lifetime loss and LAE ratio is at 78%.

When combined with our other expense categories, the lifetime CR is at 96%, thereby combining new and renewal performance and achieving our lifetime target.

However, within the new-to-renewal loss ratio performance, there are segmentation opportunities as not all policy characteristics at new business have the same expected lifetime loss ratio performance. Building on this with an illustrative example, consider car color starting in the upper right.

When looking at the expected renewal improvement from new business in loss ratio on red cars, we see a larger change than we expect to see in green cars. The implication for pricing are twofold. First, we observed that red cars last on average 5 policy terms and green cars last on average 7 policy terms.

The second consideration is that the loss behavior also varies. We priced at this differential in expected policy life expectancy and the different improvement in indemnity performance between the 2 segments, in this illustrative case car color.

Applying this to a real-life opportunity, we frequently discuss key consumer segments as Sams, Dianes, Wrights and Robinsons and our continued opportunity for growth in the Robinsons segment. As a refresher, Robinsons, our consumers who bundle their home and auto needs and typically have a longer policy life expectancy.

We can recognize this policy extension and build it into our pricing, which ultimately leads to lower rates with more lifetime renewal premium at a lower cost structure. Finally, while I focused on the indemnity piece of cohort pricing, you'll also note that the policy life expectancy can affect the other expense categories.

We'll move into the distribution channel and acquisition costs next. Before we dig deeper into acquisition expenses, we need to first review the fundamental differences between the direct channel and the independent agency channel.

While we price both channels to a lifetime 96, the economics differ in both how we account for the acquisition cost and relatedly, the timing of those costs. Starting on the left, Direct has an upfront acquisition expense by virtue of our marketing spend.

We spend consumer marketing dollars, media and nonmedia alike and account for the entire acquisition expense in the new business term. We can see that in the illustrative example with a 60% acquisition expense ratio on the first term and 0% on subsequent renewal terms. Again, this reinforces the lifetime value of extending policy life expectancy.

With the independent agency channel, policies are acquired at the cost of commission paid to the agent. The commission expense is paid to the agent and ultimately reflects a percentage of premium earned by policy term.

Unlike Direct, we see a much smaller acquisition expense in the first term, which reflects the commission paid on that term, and we see additional expense in renewal policies also reflecting a percentage of premium earned.

In both cases, in this dialyzed example, lifetime acquisition expense is 10%, and we achieved a 96% combined ratio, but we have a much larger first-term expense in Direct and a higher renewal expense in Agency. Let's look at each distribution channel in a little more detail.

We'll start by spending more time on how policy life expectancy is a key component to our direct acquisition economics. Consider 2 policies with different PLEs. In policy #1, we see a total of 6 policy terms with a target acquisition expense of 60% on that first term.

With an average policy premium of about $750, we can assume our rate level considers about a $450 acquisition expense, which is about 10% across the entire life of the policy. In other words, 750 x 6 policy terms is 4,500 and the $450 policy acquisition expense is 10% over that lifetime premium. Now let's move to policy #2.

Holding all else constant, but the policy life expectancy, you can see the first term assumptions produces the same 450 permissible or 60% of 750. However, with the addition of 2 policy terms, the acquisition expense of $450 million is now 8% of total policy lifetime premium.

Said another way, we invested the same $450 in policy acquisition as we did in policy #1, but we see the benefit the PLE extension gives us as a lower acquisition expense as a percentage of lifetime premium. We can segment our pricing to capture differences in policy life expectancy.

Depending on market conditions, we can either increase the allowable to drive additional demand or we can lower our rate level to increase our new business conversion. In both cases, we will still achieve a lifetime 96.

A real-life example would be our continued segmentation efforts, reflecting our consumer marketing segments and correspondingly, our investment to acquire Robinsons. Again, those who bundle home and auto, which carry a longer policy life expectancy compared to those who don't bundle their home and auto with Progressive.

Moving on to the Agency channel. This distribution business model takes a different form, as I mentioned, we pay commission to agents. In the example above, the policy achieves a lifetime 96, but you can see that the acquisition expense ratio is the same between the first policy term and the remaining 5 renewal terms.

This is possible because we pay commission by policy term and ultimately unearned premium. In practice, it is possible to vary commission by new and renewal. However, even in this scenario, the combined ratio implications will likely be much flatter than what we see in our direct economics.

Because we pay a percent of commission is earned, acquisition costs and dollars will vary based on policy premium, but as a percentage of premium, we'll always match. This is a key element in cohort pricing as the dollars may vary, but the expense as a percentage of the premium will be the same.

We still see cohort performance that has an overall higher cost structure in new business, but it is proportionately more in an indemnity and operational expense relative to our direct business model. Finally, we finished with operational expenses, and we see a modest new renewal differential.

The operational cost for servicing new business policies is greater than the operational cost for renewal terms. The remaining other category includes things like bad debt or unpaid balances on earned premiums. We extend coverage on a canceled policy and did not receive payment. This happens with greater frequency on new business.

We see these expense differentials consistently and can also price them into our product. Calendar year earnings have an implicit new and renewal mix driving results, and we manage that mix to ensure target combined ratios are achieved for the calendar year-end question.

While managing our new renewal mix, we consider both the mix of our cohorts and the rate revisions earning in. Remember that rate from revisions take time to earn in as policies have to renew into the new rate level. Typically, this happens in our new business faster than renewal as our rates hit the marketplace before our renewal book.

This will affect the combined ratio performance of our new business and renewal cohorts.

Setting rate changes aside, periods of accelerated new business application growth can apply upward pressure on our combined ratio, as we now know, new business has a higher expected combined ratio and in the case of Direct has additional pressure based on our accounting of consumer marketing expense.

Using this stylized example, you can see that in a period of high growth, new business makes up a larger portion of our total earned premium. And while we may be within our lifetime targets, we see our calendar combined ratio pierce the 96.

To manage this, we can slow growth and thereby reduce the mix of new business earned premiums, which results in a combined ratio less than our 96 goals. And in both cases, we are within our cohort targets. This is a great example of how we develop our media budget.

We consider both our lifetime economics and our margin position considering each simultaneously with calendar year performance taking priority, and we then adjust our budget accordingly.

Upward pressure from new business can be dampened by the lengthening of policy life expectancy as our renewal book becomes a larger portion of the overall book of business. With the expected combined ratio improvement, we can use that to manage new and renewal earned premium mix.

As you can see, as it relates to our expected cohort performance, we price across the lifetime of a policy and balance our near-term and long-term growth with our 96 calendar year goal. Thank you for your time..

This concludes the previously recorded portion of today's event. We now have members of our management team available live to answer questions, including presenters, Gary Traicoff and Jim Curtis, who can answer questions about the presentation. [Operator Instructions]. We will now take our first question..

The first question comes from Elyse Greenspan with Wells Fargo..

My first question, you guys highlighted that you need 6 points of additional price, right, in Personal Auto over the balance of the year.

Can you just help us get a sense, what are you assuming for severity and even frequency, just overall from a loss trend perspective when you think about additional rate need from here?.

Thanks, Elyse. We really don't share our trend selects normally. But obviously, we've been seeing the trends in shorter-tail business increase as just the repairs to vehicles have slowed down. And when we look at that in overall kind of both frequency and severity, we think we're at 6 right now. And just like I said in the last call that could range.

But at this juncture, we think we need another 6, so about 17-ish for the full year..

And that will get you -- you believe that will get you to the 96?.

It will certainly get close. Again, all the caveats I said last time with weather. We feel much better about our reserving because of those losses being short-tailed, and we started seeing those at the end of last year. So I believe Gary really got out in front of that. It will certainly get us closer.

But again, all the caveats and the uncertainty that we've felt for the last 3 years kind of are underlying that assumption..

And then my follow-up, you guys saw a significant amount of adverse development for accident year 2022. So I'm just trying to square the fact that there was a good amount of adverse development for accident year '22 with the improvement that you're seeing in the accident year ex-cat loss ratio so far this year..

Yes. What I would say to that is as those 2022 rates earned in that I put on one of the slides that I had, we believe we're putting on the books both new and renewal business that is closer to or below our targets, our accident year targets, I should say, in both channels on a year-to-date basis.

So we feel good about where we're at in the business we're putting on the book in 2023..

Our next question comes from Mike Zaremski with BMO..

I guess my first question is on loss cost expenses. And you talked about, Tricia, coming from a variety of sources. And you also talked in the letter about some of those sources looking like they're maybe mean reverting or decelerating a bit.

But I guess just stepping back, I know there's -- I know this is a complicated issue because there's lots of different sources. But we saw, I feel like over the last decade or so, a bit of a secular trend in terms of higher bodily injury expense inflation levels. And maybe you disagree with that.

But -- and the industry seems like they've been doing a decent job of appreciating that over time in pricing for it. It seems like it's coming more -- or we can see the data is coming more from the non-BI side today.

And so do you feel like there's like a stickiness to this that you and the industry are appreciating? And ultimately, does this kind of cause Progressive to not be able to maybe play as much often if there's uncertainty on the trend over the coming year or so?.

Yes. I think we have our arms around it. I mean for BI, obviously, we try to keep ahead of that trend. When we see it, we react to it, and we're in line with the industry. This has been a very unusual time. As you know, if you've looked over the last couple of years, and you've seen the Manheim used car index and how that's flowed.

Even though you're seeing it tick down a little bit, it still is a very big difference in what you saw in 2018 and 2019. So when we look at that, especially as we looked at the second half of last year, we really started to see the trend, specifically in property damage, but also in collision increase.

And it's really -- and we're watching it flow through in our claims organization. So you have -- let's take the property damage example. Somebody is taking care -- another one of our carriers -- competitors taking care of their insured, but they're going to subrogate us. So that accident happens today.

And there's no shop capacity for that particular hit. So they do an estimate as best they can with the sheet metal still in the car. Let's say it's drivable, for this example. And they continue to drive. The shop can get it in 45 days from now, maybe 60 days from now.

It comes in, they get the sheet metal off of it, oh, lo and behold, there's more damage. Those supplements are more frequent, and they're more expensive. So you're having capacities, longer cycle times, longer rental times. I think I said in my letter, the rental time, we're seeing it go down a little bit, and that's one data point.

So we'll continue to watch that. But you've got that, and then you've got labor rates and parts increase. So in terms of your question, the stickiness, will likely be based on wages for workers in the body shop.

So there -- recently a Wall Street Journal article that talked about automotive repair workers, their salaries increased 24% in quarter 4 of 2022 compared to 2018. So 20% over the last year, and that was from the Bureau of Labor Statistics.

And making cars be in the same position, the indemnification of the vehicles has also soared since 2018, about 36%. So I do think we might see some trends like in used car prices as the supply chain issues came through and new, et cetera. That might continue to go down. But I think some of the other things will be part of our new normal..

Okay. That's helpful. My final question is on the expense ratio. It's been -- the expense ratio has come down a lot. You've called out. We can see some of the math behind it. It looks like ad expenses has been the biggest lever, which is a great lever to have.

But kind of curious directionally is -- we're looking at the ad expense ratio over a decade plus, is this -- are we kind of at the trough? Or is there more room for either ad and/or the expense ratio to come down in the coming quarters?.

Well, as we attempt to get to our calendar year 96 where there's always levers we can pull, clearly, I'd rather be in a different position where we're spending more money on media and continuing to grow as long as, of course, that would be at our target profit margins.

We have a lot of discipline within the company as well in terms of expenses in times such as these. So we -- obviously, we'll continue to look at that for this next 6 months to see if we can get closer to that 96..

The next question comes from Jimmy Bhullar with JPMorgan Securities..

So you went through a lot of details on your whole process on reserving.

But maybe just stepping back, can you talk about your confidence in where your reserves sit now? And given that you've had, I think, 6 straight months of adverse development, are you much more comfortable in reserves overall? Or is there still a lot of uncertainty, and it's hard to say that you've got up in terms of loss trends?.

I'll let Gary weigh in on that. I would say from where I sit, I'm much more comfortable. But again, Gary is independent and does his reserve reviews without my involvement. I hear about that just a little bit before you do. So Gary, why don't you weigh in a little bit..

Sure. Thanks, Tricia. Great question, Jimmy. So overall, yes, I feel really good about our overall reserve position and where we're at right now. To your point, we have had quite a bit of unfavorable development each of the last 6 months.

As Tricia pointed out, if you exclude Florida, all of the development would really be coming from accident year 2022 were actually favorable on accident year '21 and prior. And within accident year '22, over 85% is from accidents that happened in the last 6 months of the year.

And a lot of that relates to the fixing cars and the reopen severity that Tricia was talking about. That's fairly short-tailed. We've seen that come through the data, and we have reacted very quickly to it. And so that gives me a lot of confidence because of the short tail nature of -- or most of the issue was that we're in a good position right now.

At the same time, right, I can't obviously guarantee and the data could change going forward, but I feel really good of our current position..

Thanks, Gary..

And along the same lines on Florida, I think it was somewhat expected that the number of lawsuits would go up given total form, but it's lasted longer in your results.

So just wondering how the number of lawsuits that are coming in have trended through the last 3, 4 months or so? And was there an element of Florida in June in terms of adverse development and how that's changed versus maybe April and May?.

Yes. So that's a good question. So with respect to Florida overall, obviously, the big hit occurred March, April, May. And we've started to see it level out in what's been coming through. Most of June really related to the actuarial adjustments that we took from the reserve reviews primarily related to property damage, third-party fixing cars.

So Florida had a very little impact from what we had from the June development..

Yes. And all I would say in addition to that is as soon as the House Bill 837 went through, and then we had the subsequent lawsuits that I talked about, we went in and were pretty aggressive and conservative because we were trying to figure out what was estimable and probable. And that will develop over time as those also get settled.

So that's not something that we'll know right away because those take a little bit longer to settle..

The next question comes from David Motemaden with Evercore..

I just had a follow-up on the reserves. Gary, you had mentioned that you can adjust up to 100% of your reserve balance monthly. I'm just wondering if that happened in June.

Or was it a normal review where you maybe did 60 to 70 reviews, but not a complete review? And we have to wait for potential further movement upwards or downwards as we move forward over the next few months?.

Yes. Good question. So in June, it was the typical 25% to 30% of reserves that were looked at. The comment related to the 100% kind of goes to the part on the reviews that we do not look at in the month. They still will be inflated or unless -- particularly if they're below the threshold, they will be inflated. And they also will be aged.

They also can change. For example, they become plaintiff attorney rep, et cetera. So the potential every month to have 100% of the reserves adjusted somewhat whether it's by an adjuster or whether it's by inflation or aging exists. In terms of our normal process, we typically stay pretty consistent with that 25% to 30%.

I would say the only couple of times we might deviate, in December, in particular, because I'm the opining actuary for the statutory reserves. If we are seeing additional information, we may add a review or 2 to make sure that the statutory companies are in the correct position as of year-end.

And then, obviously, for something like Florida when the House Bill passes, we would go in immediately with new information from that and do an adjustment..

Got it. So I guess -- so in the June review, it sounds -- so I guess you just confirmed that was just 25% to 30% and maybe some other adjustments around the edges, but it wasn't one of those full comprehensive reviews..

Yes, that's correct..

Great. And then just my second question, so I believe there was a stylized example just given about the mix of business and how having less new business, more renewal business can have an impact.

I guess how quickly can that have an impact on the combined ratio, particularly considering the growth that you guys had in the first quarter? Is that something where that mix can change significantly through the end of this year?.

I'll let Jim talk about that a little bit. Obviously, there's a lot more that goes into it than just the acquisition expense. A lot has to do with losses, and there could be uncertainty around catastrophes, et cetera.

But Jim, do you want to talk a little bit about that?.

Sure. Also a good question. A couple of ways to think about that. There's several dimensions to consider when thinking about mix. So to your point about any type of lag function when we write business, it's in written premium, but doesn't necessarily hit our earned premium mix until it starts to age or get into the policy term.

So we would see on a 6-month basis, roughly 3 to 4 months later, where you start to see your earned premium mix shift. That would work in both ways, periods of growth on new business apps and periods of slowing our new business apps..

The one thing I would add to that, just for clarification, we expensed advertising in the period incurred. So all -- on a direct basis, especially -- and you're seeing that come through the expense ratio very quickly..

Yes, I would -- and I would build on that, that the expense does hit immediately. The earned premium mix would -- really would start to see some measure of reaction in our indemnity performance..

The next question comes from Alex Scott with Goldman Sachs..

First one I had is on the commentary that you guys gave on the lifetime combined ratios. And it just strikes me is a lot of that is very geared towards what the lifetime ends up being. And appreciate that you guys probably have the best data and best capability to analyze it, too.

But are we at a period of time where just given how unprecedented it is, with inflation and all the pricing, I mean how much room is there around the estimates that you guys are putting into the lifetime? I mean, are you being conservative around potentially significantly higher churn as we kind of get through this period where everything is being taken out to market on a much more consistent basis? And how much is that being incorporated into your decision as to whether to start growing quickly or not?.

Well, I think we always -- each data point. And as we get more confidence, of course, we were able to do that. We saw an increase in PLE on both the 12 months and 3 months. I'd like to see a couple more data points because there's a lot of shopping in the industry.

And obviously, you're looking at the same data we are from the industry, and there's a lot from movement from rates. So I would say there's hundreds of variables that we look at when we think about that. And we'll continue to do so, again, in this uncertain environment.

Jim, do you want to add anything?.

Sure. We also consider different time periods and just typical when looking at data set, longer time period, more credibility and more confidence we have. We'll look at recent time period as just a leading indicator adjust slowly. My team set sort of our permissible acquisition as an example, and we just go through that exact process..

I would just quickly add is that we're not writing a 4-year policy at once, right? We constantly adjust rates to ensure we're hitting our lifetime targets.

So to the extent that we have those targets, every renewal of every policy, and we have an advantage with a mix of 6 months policies that we can adjust as we see another card or see another data point in loss costs, et cetera..

Have you seen that with how nimble we've been with our pricing over the year? As we've seen things we've made changes. We are as anxious as likely our shareholders to grow. But again, we have a very specific core value of profit on a calendar year 96. And so that's a priority..

Got it. That's helpful. The follow-up I had is on the loss adjustment side of things.

I mean some of the anecdotal chatter, I've heard out there is that just given where the labor markets are and so forth that the claims adjusting process might be challenged by that and being able to get the right amount of hires and talent and to be able to execute that process like it's done normally.

And are you -- have you experienced any of that? Is that at all a challenge that you face in trying to get the reserves in the right spot consistently? Or is that something that's less of an issue at this point?.

I think it's a little bit less of an issue. I think anytime you have the growth we've had and with the labor environment that has been around for a couple of years, it's a challenge. I will say at this juncture, we are fully staffed in our claims organization. So we feel really great about that and turnover is continuing to go down.

That said, our tenure is still low. And so with that comes just the learning curve from that. And our Claims President is having the newer people definitely are in the office with supervisors because that's really where you learn a lot and you get up to speed very quickly, and we have a lot of processes in order to do that.

But I would say, if you asked me that a year ago, it would be a different answer than today. I feel confident in our claims organization, the action they're taking. And as tenure continues to grow, I think we'll be in a great position for accuracy..

The next question comes from Josh Shanker with Bank of America..

Yes. I guess my question is for Jim.

Looking at the new cohort acquisition cost ratio, can you talk a little bit about how the current period with ambient shopping in the last meeting to reach new customers with advertising to get them onboarding affects the way you look at that new cohort margin?.

Sure. We measure that by our permissible compared to actual. Actual is defined as our media spend or consumer marketing spend and income and volume apps, with what we believe to be as a sort of heightened dislocation in the marketplace, a lot of shopping, which we hold our conversion level that will drive our cost per sale down.

And so in the current period, we'd be pretty efficient. And then it's a matter of managing our media budget to our calendar year position..

And I realize how valuable the 96 calendar year is to the company.

Is there a risk that right now, there's an opportunity to acquire new customers at a cost far below the long-term average and by focusing on the 96, there's a lot of growth being forgone that ultimately will be more expensive to achieve in the future?.

2022 was among the worst years for industry Personal Auto profitability and so far, 2023 has been worse. While we can't know the future, current market conditions suggest there'll be opportunity to increase market share in the future. And then just let me wrap up by saying growth can't happen without growing surplus.

Our capital position today is strong, and we believe we have plenty of capital to support the premium growth we expect. We have a structured growth business that is largely self-funding internally through underwriting profits and investment returns. So to bring us back full circle, our profit core value is the key to sustained growth.

I know that was a long answer, but I think it's really important to be super clear and unambiguous about how we think about the business. I hope that answered your question, and hopefully, others..

The next question comes from Greg Peters with Raymond James..

I'm going to start building upon the comments you just had about the growth results and Jim's cohort pricing. I'm wondering if you can provide us a perspective of how the limits profile of your consolidated order book have changed. I would imagine if you're focused more on Robinsons, the profile is moving up.

But maybe you could give us an updated perspective on how that looks and where you think it might be heading..

Yes, I think you're spot on. We are trying to have more and more of the Dianes, Wrights and Robinsons. We love Sams as long as we can make our target profit margin, but we continue to increase our bundle. And as we think about both auto and property, those bundles are more important. So we've increased our Robinsons over time.

And then, of course, we've talked about this in the past, and it's probably worth an update at some point in the next couple of quarters. The soon-to-be Robinsons. So it could be a Diane that's buying a house, et cetera. So we'll continue to try to be a place for every cohort, but we have increased our Robinsons profile in the last several years.

And of course, during times like this, Sams are going to shop more because they're very incented by price..

Right. I guess -- well, I guess what I'm wondering is, in the cohort pricing, there's an assumption around retention.

And given that there's all these price changes happening, not only at Progressive, but with a lot of other companies in the auto sector, I'm wondering how with elevated shopping, how those retention characteristics are holding out at the Robinson level, at the right level? Is it matching what your expectations are? Or is that causing some hiccups along your pricing assumptions?.

Well, I think -- we don't talk about retention with each of the profiles of our consumers, the marketing tiers. But we do look over time. And historically, Robinsons have fared better. And we find that the more products you have, the stickier you're going to be.

And Sams, especially if there's other characteristics to go, can -- last about the same months and haven't varied too much over the years. But although we don't share it, I would say Robinsons continue to be stickier..

Great. I just -- we are constantly looking at the assumption to go into our acquisition model. As Jim was mentioning, we use PLE to understand the permissible costs for acquiring customers, and we do that at a very granular level. So you're right. There have been significant changes in policy life expectancy.

Obviously, lately, it has been going back up, and we're looking at that at a very granular level and pricing that in over the lifetime of those customers. And to pass point, we have many shots at pricing -- repricing those customers, if necessary..

The next question comes from Tracy Bunguigui with Barclays..

Just a quick follow-up on your 96% combined ratio target. I appreciate all the commentary so far.

But I'm just wondering, do you feel like there are secular changes going on in auto that you may need to do better than 96%? Like it feels like the tail is lengthening, which is happening for property damage despite it being a short tail line, which could add to volatility..

Well, good question, Tracy. You remember our 96 is an aggregate. So we look at it differently for new business, renewal business. We've taken to the assumptions of our PLE products, state, although it all goes up to 96, which I don't think will change.

One, thing, we do look at a really granular level; and two, I think we want to watch this playthrough because some of these changes came about pretty quickly in the delays. And a lot of it is really the shop capacity.

And as that starts to open up with the exception of what I talked about with labor rates and probably prices, I think we could be priced adequately and continue to keep our 96 calendar year and lifetime goal. That would be my approach..

Okay. Your year-to-date current accident year actuarial adjustments of $424 million was primarily driven by fixing vehicle coverages. So how do I reconcile that update with your property damage severity loss trend going down when you come up with your loss picks? Like you reported yesterday, property damage severity is 11%.

But if I compare that to 1Q, it was 15% and then 20% in the full year '22..

Well, we had a lot of rate increases come in, so that was one of them. And a lot of our property, if you take -- if you look at the property and the cat load on top of it or the -- I should say, the cats we've had, you can take that into account.

But it is -- that 80% of that, I think you did the quarterly number, but 80% of that $1.1 billion on the year-to-date was auto, 50% was cars, about 35-ish percent was Florida, glass, PIP, BI. And then the rest was IBNR and BI across the country, but property did not -- did develop favorably.

Gary, do you want to add anything?.

So I think Tricia's spot on there. The one thing I would add, Tracy, to your point, the property damage severity that you're showing, that's for all claims, if we stratified and only looked at the reopen severity, which is really where the reserve actual adjustments have come from.

The reopen severity is well above and has been increasing over those quarters relative to the 13%, 14% you're seeing..

Okay. Yes. Sorry, I meant in your current loss pick, your current -- your actuarial adjustment, not the $22 million adjustments.

Are you considering a higher loss pick for property damage or not given that -- it feels like that loss trend is trending down?.

Yes. We are. So I think to your point, even though the overall is trending down some, the reopen severity, we believe, on the current accident year has been trending up. And so a lot of those actuarial adjustments, both on prior and current year are reflecting the elevated reopen severity.

Even though, to your point, the overall has come down, that is a different trend if you just look at only the reopen severity, which is really what's driving a lot of the reserves..

The next question comes from Ryan Tunis with Autonomous Research..

So Tricia, I guess just going back to the comments you were making about how in the past, you've had tons of remediation that have led to tons of outsized growth. And you referenced that you see opportunity for that in the future.

Just curious from where you sit today, is there any reason to think your appetite will grow, will be different going into 2024 than it was going into 2023?.

As long as we look at the data and feel good about where we're at with our combined ratio on overall calendar year and accident year, we will grow as fast as we can. So that has been our objective for a long time, 96, grow as fast as we can. We want to do that. We're a growth company. We feel great about it. But again, profits over growth.

But as soon as we feel confident with the little stability and less uncertainty, we are very prepared to grow..

And just to be clear, like let's just say they're about -- you guys executed, but there's a bad hurricane and you ended up coming in at That wouldn't lead you to punish yourselves and not grow in January, right? You'd be looking what you think the prospective returns are..

Yes. Yes. We don't know how the next 6 months will go. We're going to do the best we can. But again, like you said, if something happens, that is above and beyond that would be different. But yes, we will continue to grow.

And in 2024 would be a new year as long as we felt that we had the policies on the book that we believe are at or below our new and renewal targets..

Got it. And then just lastly, we've talked about the impact of the [indiscernible] stuff on prior year.

Is there any way to quantify the loss ratio impact did that happen in the first half of 2023 -- on the current accident year?.

From Florida, you mean?.

Correct..

The current accident year, the impact was virtually 0 because those are lawsuits that are opening on claims that occurred pre-HB837. It can't be now because HB837 was in March. So there's some, but predominantly, those were prior year claims. We're also looking closely at what we expect to be some benefit in the environment from HB837 as well.

We're looking at the data closely. It will take a while to assess if indeed, we are seeing that. But we personally -- we believe that there is going to be a benefit, and we look forward to pricing in once we see it..

The next question comes from Yaron Kinar with Jefferies..

A lot of questions on the 96, so I'll pile on here. So if I understand correctly, you're still within line of sight, being close to 96 for this calendar year. One of the ways you can achieve that is by lowering new business growth, which ultimately should improve the combined ratio.

I guess my question is, and maybe it's the other side of Josh's question earlier, ultimately, can you achieve that simply by lowering the new business growth while achieving kind of a 96 combined ratio for cohorts? Or do you now need -- or are you aiming to achieve a better than 96 for existing cohorts in order to get to the 96 will be as close to the 96 for the year..

Well, we clearly need to get a rate. So that's one part of the formula. We also have really had a lot of restrictions from an underwriting perspective. on new business. And so those are some other things that go into it. We have a line of sight, but again, we'll caveat that by all the comments regarding uncertainty that I've said before..

Okay.

But ultimately, do you believe that beyond the kind of flexing down or up the new business growth, do you need to take any additional actions to lower the combined ratio for the existing business below 96 in order to get to a full year 96 or close to it?.

So clearly, the new renewal mix is working in our favor right now. The other bigger tailwind is the rate that's earning in. So we've taken a lot of rates. We still have about 6 points in Personal Auto for that rate to earn in. Additionally, we're taking 6 more points for the rest of the year.

So the denominator effect is a tailwind for us the rest of the year..

Got it. And then my other question, maybe for Gary, since Gary is on the call. We saw an uptick in bodily injury severity this quarter.

And I am curious as to how that ultimately plays out or how you think about it, how you approach that with regards to prior year reserves? Ultimately, I would think that bodily injury has a longer tail, so more risk there for things going adversely.

So how do you approach that?.

I'll start, and then Gary can weigh in. About half of the BI was actuarial reserves, but it's pretty much in line with the industry. So we're not too concerned about it. A lot of them are soft tissue, attorney rep but not litigated. So they do take longer. But I feel like we're in a good position and understand where they're headed.

And I think in a better position than we were even a year ago.

Gary, do you want to add anything?.

Yes. I think Tricia is spot on there. The only couple of things I'd add from the reserve side is we tend to look at the data off of an accident period basis, right, as opposed to reported calendar information.

The accident period data is more smooth, right, on a calendar basis, you can get some movements up or down, sometimes with changes in mix or closures, et cetera. So on an accident period basis, we're a little bit lower than what that shows. The other thing what Tricia alluded to is there's some large loss mix, et cetera.

If you think back to that presentation, that's one of the reasons we employ that reserve threshold. So when our claims adjusters identify those large losses, once it's above that threshold, they put that up immediately. So it's recognized in the reserves.

And again, there could be a little bit a timing difference between the accident period and the calendar period, which you would see..

We've exhausted our scheduled time, and so that concludes our event. Michelle, I will hand the call back over to you for the closing scripts..

That concludes the Progressive Corporation's Second Quarter Investor Event. Information about a replay of this event will be available on the Investor Relations section of Progressive's website for the next year. You may now disconnect..