The Allstate Corporation (ALL) Q3 2026 Earnings Call Transcript

Operator: Good day, and thank you for standing by. Welcome to Allstate's third quarter earnings investor call. At this time, all participants are in listen-only mode. After the prepared remarks, there will be a question-and-answer session. To ask a question during this session, you will need to press star one one on your telephone. If your question has been answered and you would like to remove yourself from the queue, simply press star one one again. Please limit your inquiry to one question and one follow-up. As a reminder, please be aware this call is being recorded. Now I would like to introduce your host for today's program, Mr. Allister Gobin, Head of Investor Relations. Please go ahead, sir.

Allister Gobin: Good morning, everyone. Welcome to Allstate's third quarter 2026 earnings call. Yesterday, following close of the market, we issued our news release and investor supplement and posted related materials on our website at allstateinvestors.com. Today, our management team will discuss how Allstate is creating shareholder value. Then we will open up the line for your questions. As noted in the first slide of the presentation, our discussion will include non-GAAP measures for which reconciliations are provided in the news release and investor supplement. We will also make forward-looking statements about Allstate's operations. Actual results may differ materially from those statements, so please refer to our 2025 10-K and other public filings for more information on potential risks. Now, I will turn it over to Tom.

Tom Wilson: Good morning. Thank you for investing time in Allstate. Let us begin on slide two. Allstate's strategic objectives on the left there are to increase market share in personal Property-Liability and expand protection services by leveraging the Allstate brand, customer base, and capabilities. This quarter, operating financial results were excellent, as you can see on the right. Property-Liability policy growth continued to increase and was 3.3% above the prior year. At the same time, the adjusted underlying combined ratios were attractive at 91.4 for auto insurance and 61.9 for homeowners insurance. Higher investment income and good protection services results led to adjusted net income of $2.5 billion, or $9.76 per common share. We will go through these results in detail after updating you on the longer-term value creation through Transformative Growth in ALLIE. We will turn to slide three, so we can discuss the progress in executing the Transformative Growth plan. As you know, Transformative Growth was initiated in 2019, and it has five components, which are shown on the left. It begins with improving customer value through more competitive prices and new products. Our progress is shown on the right. Reducing expenses enables us to offer more competitive prices while maintaining attractive margins. New Property-Liability products that are affordable, simple, and connected, or in our world of acronyms, ASC, are being launched. The Allstate-branded ASC auto insurance products are available in 46 states, and homeowners insurance products are in 42 states. Similarly, in the independent agent channel, the Custom360 products are available in over 40 states. Expanding customer access by broadening distribution and increasing the sophistication and investment in marketing has also been successful. As you can see on the right, new business is up 125% to 12.4 million items, is now equally split between the three primary distribution channels for personal lines: exclusive agents, independent agents, and directly from call centers or on the web. That is supported by over $2.2 billion in marketing with sophisticated analytics and an improving return on allowable acquisition costs. All of this required a new technology ecosystem, which is being deployed, including an orchestration layer, which enables us to rapidly leverage artificial intelligence capabilities. Several years ago, we designed and began building Allstate's Large Language Intelligent Ecosystem, or ALLIE, and its rollout is accelerating. A few of the larger impacts are shown on slide four. Like most companies, we are using AI to do an increasing amount of coding, which as you can see was over 12 million lines of code in the third quarter, which is 3.6x higher than the first quarter. In July, we successfully launched Ask Allstate, which handles a variety of service-related inquiries for auto insurance customers with excellent outcomes. 15% of the ASC chat volume has been handled by Ask Allstate, and customer satisfaction is high. Smart Concierge at Allstate Protection Plans has handled over 1 million calls. These capabilities are being extended to more products and services, such as claims, as you can see on the right. There is also a much broader adoption of AI capabilities really throughout the whole company, just beyond these tools, with 97% of employees are now certified as AI-enabled, and adoption is happening in all areas of the company. After the launch of Muse this summer, many of you asked how this impacts our go-to-market approach, which is laid out on slide five. The summary is this creates opportunity for Allstate, and we will be launching sales-related AI capabilities in early 2027. To put this in perspective, though, let us start with customers or consumers, what they want in buying, and how they like to use protection. The top of the slide shows how consumers choose to purchase protection. Starting at the left, some consumers are brand-sensitive and purchase through exclusive agents or directly from companies. In the exclusive agent segment, State Farm and Allstate over-index in market share. For those customers that are brand-sensitive and prefer to purchase directly, GEICO and Progressive over-index in market share there. Allstate has been increasing share in this segment since we replaced Esurance with the Allstate brand and began executing Transformative Growth. Moving to the middle, there are customers who prefer a choice of carriers and want someone to provide advice and handle the transaction. They typically go through independent agents. Allstate serves this segment with the National General brand and through Answer Financial, which is a direct independent agency that writes over $500 million of premium annually, representing over 30 insurance companies. The advent of personal agentic assistants, or PAAs, gives us the opportunity to serve consumers who prefer to have a machine to handle their transactions, such as Muse, Spark, Dot, you go down the list. The introduction of PAAs creates opportunity for Allstate, given our capabilities, and an upcoming launch of our own personal agentic assistant that is focused on protection. So let's start with capabilities on the bottom left, and then we'll discuss the PAA's value proposition. To thrive in this world of highly sophisticated agents, you need sophisticated pricing, since they're likely going to do more price comparisons. Secondly, some consumers may still have a brand preference or want a local agent, so we have a broad set of opportunities for them. When ALLIE is fully deployed and the PAAs have expanded their capabilities, this will give customers straight-through processing on servicing claims. Our brand and marketing expertise are being utilized to engage with LLMs. Essentially, we're now marketing to machines and people, not just marketing to people. These capabilities also give us the opportunity to be offensive in leveraging capabilities by launching a PAA that's an expert in protection, that still meets consumers' needs and their desires for fewer clicks. Its value proposition includes more accurate pricing, given our insurance expertise, than the general-purpose PAAs will have. It will also provide consumers with context on the type of coverage that's right for them and straight-through processing. In addition to serving customers, our PAA will serve other PAAs, as shown at the top. In addition to the benefits of accuracy, context, and ease, it may also require less compute time for them. For example, today, a lot of the PAAs go to a variety of websites to get quotes. By working with Answer Financial, we can review multiple companies at once, reducing compute requirements. Insurers also benefit from Allstate Financial's capabilities because they can target more customers and optimize their risk-adjusted returns. While it's unclear how quickly this market will develop, this will enable Allstate to provide increased value for customers. We've been building this platform for over a year. The foundational technology is in place, the PAA is in development, the algorithms are being tested, and our relationships with publishers and insurers continue to expand and strengthen. Look for this to come to market in early 2027. Chris is going to cover third-quarter results, but let me begin by welcoming Chris. We're thrilled he chose to join us in August. As you know, he was the Chief Financial Officer at CoStar Group before he joined us. He brings tremendous capital markets expertise from his time at Freddie Mac and Morgan Stanley, and his leadership is already making a difference for us. Chris?

Chris Lown: Thanks, Tom. It's a pleasure to participate in my first Allstate earnings call. We had a strong quarter, and we will cover results in a moment. Before we do that, I want to spend a minute on how we communicate and evaluate our performance. Since joining Allstate, our financial reporting cadence has been one area we have looked at closely. The SEC has proposed giving companies the option to report semi-annually rather than quarterly. At Allstate, we are not going to report less frequently. We believe quarterly reporting provides important transparency, especially as an investor managing a nearly $90 billion investment portfolio. However, Allstate does not manage the business to deliver short-term quarterly results, so the concept of long-term value for shareholders aligns with our approach. That is why we are adding more emphasis in our presentations on longer-term trends. We intend to continue reporting quarterly, but we also want investors to focus on longer-term trends that are more representative of how we manage the business, evaluate performance, and deliver value to shareholders. With that, let's review results on slide six. Allstate is delivering increased growth and attractive returns. Starting with the quarter in the two middle columns, revenues were $17.3 billion. Property-Liability net premiums written increased 2.5% to $16 billion. Total policies in force increased 3.9%, with Property-Liability policies up 3.3% and Protection Services up 4%. Investment income increased by 24%, reflecting higher yields and growth in premiums. The Property-Liability combined ratio is 86.6, 6.5 points above the prior year, primarily due to 7.1 points of higher catastrophe losses, partially offset by 2.4 points of more favorable non-catastrophe prior-year reserve releases. Adjusted net income was nearly $2.5 billion, or $9.76 per share. This generated a 39.7% adjusted return on equity. Let's move to the right-hand column to put this in perspective to the last 12 months. There are two important trends related to premium growth: average premiums and policies in force. There has been a 2.7% decrease in average auto insurance premiums over the last year as the percentage of new ASC products increases. On the other hand, homeowners insurance average premiums continue to increase, reflecting increased loss costs and higher home values. As it relates to policies in force, the growth rate has increased in each of the last four quarters, and at 3.3%, we now believe it is higher than the underlying growth in the number of cars and homes. Investment income has been steadily increasing with a larger portfolio and reinvestment at higher rates. We extended duration, locking in higher yields for longer. Underlying margins are approaching target levels, as Jess will discuss. While catastrophe losses were higher this quarter, on a trailing 12-month basis, they were broadly consistent with the prior year. Prior year net income benefited from the sales of the employer voluntary benefits and group health businesses in Q2 and Q3 of last year. Now I'll hand it over to Jess to discuss the Property-Liability business.

Jess Merten: Thank you, Chris. Let's look at the Property-Liability results on slide seven. Property-Liability generated strong results in the quarter, with continued policy growth, premium expansion, and excellent underwriting performance. Policies in force increased 3.3% from the prior year, with 3.4% growth in auto insurance and 3.3% growth in homeowners insurance. Premiums earned increased 3.8% to $15.1 billion in the third quarter compared to the prior year quarter, reflecting increased policies in force and higher average homeowners insurance premiums. The Property-Liability combined ratio was 86.6, as both auto and homeowners insurance recorded profitability was better than our targeted levels. The auto insurance combined ratio was 83.2, which benefited from 9.1 points of prior year reserve re-estimates. The underlying combined ratio was 89.8 and benefited from reserve adjustments related to claims reported in the prior two quarters of this year, which I'll provide a little more detail on in a later slide. Homeowners insurance margins were strong with an 89.5 combined ratio, despite higher catastrophe losses because of an underlying combined ratio of 60.1. Property-Liability underwriting income was $2 billion for the quarter. Turning to slide eight, let's dive deeper into the continued momentum in policies in force growth. Starting on the left, improvements in customer acquisition capabilities have enabled increased advertising while maintaining attractive returns on investment. Advertising spend has increased from $638 million in 2023 to over $2 billion for the latest 12 months. The center of this slide shows how advertising investment, paired with broad distribution, is translating into new business growth. Auto insurance new business is shown on the top chart. New business issued applications increased 59% from 5.9 million in the third quarter of 2023 to 9.4 million in the most recent 12 months, with increases across all three distribution channels. There's also momentum in homeowners insurance. New business increased 39% from 1.1 million policies to more than 1.5 million over the same three year period. Allstate agents continue to bundle at record levels, and we're having success in the direct channel with affordable, simple, connected homeowners insurance products. The result is shown on the right of the slide. The growth rate of auto insurance policies in force increased from 1.3% last year to 3.4% over the latest 12 months. The longer-term trend reflects a decision made to reduce growth given post-pandemic margin pressure. Homeowners insurance policies in force have been increasing over this period and increased 3.3% in the latest 12 months. Expanded customer access, increased marketing investment, and new products are generating market share growth. Slide nine provides a deeper look into auto insurance margins. The chart shows auto insurance premium per policy compared to adjusted underlying loss and expense per policy. The dark blue line represents annualized average premium. The red line represents underlying loss and expense adjusted for prior year and prior quarter reserve re-estimates. As you can see, Allstate has improved affordability while maintaining strong margins, with the blue line of average premium being above the red cost line and the blue line being relatively flat. The spread between the two demonstrates the strong margins we're generating in auto insurance. Annualized average premium per policy was $1,474 in the third quarter, as actions to improve affordability for customers have slightly reduced average premiums. Adjusted underlying loss and expense per policy was $1,347. Shown at the bottom of the slide, the recorded combined ratio for the third quarter was 83.2. The underlying combined ratio was 89.8, but as I mentioned, this included favorable development on claims reported in the first half of 2026. Adjusting for this, the combined ratio was 91.4, better than our mid-90s target. Turning to slide 10, to keep costs down, we're focused on improving productivity while investing in technology. We've lowered the cost structure by eliminating, outsourcing, and digitizing work, reducing our real estate footprint, aligning distribution expenses with customer value, and improving operating efficiency. We're committed to reducing costs and have more to do, but good progress has been made. At the same time, more technology investments are required. This includes investments in ALLIE and cybersecurity, as well as managing higher software and computing costs. These investments will enable us to retire legacy technologies and focus more resources on the new technology platform. With that, I'll pass it to Mario.

Mario Rizzo: Thanks, Jess. Turning to slide 11, the Protection Services business continues to profitably grow while expanding Allstate's customer value proposition. Protection Services has 178 million policies in force, an increase of 4% from the prior year. Over the last 12 months, premiums earned increased 7.5% to $3.5 billion, while adjusted net income increased 4.7% to $221 million. The five Protection Services businesses support customers across a broad range of needs. Protection Plans offers coverage both domestically and internationally for products such as mobile phones, electronics, major appliances, and furniture, and is distributed through over 30 major retailers like Walmart, Costco, and Home Depot. Roadside Services provides towing, jumpstart, lockout, fuel delivery, and tire change services through nearly 1.8 million customer rescues per year. Roadside Assistance is offered as a bundled option with affordable, simple, and connected auto insurance products, and through third-party partner relationships. Dealer Services provides vehicle service contracts and other protection products through more than 1,100 automotive dealers. These products protect vehicle value and reduce the cost of unexpected repairs. Identity Protection extends our promise to customers into an area of growing need, namely their digital protection. This business helps customers protect and restore, excuse me, their identities, and offers consumer cybersecurity and digital safety services. Arity is a telematics-enabled mobility insights business. Its capabilities are being expanded through the acquisitions of INRIX and Replica. With over 3 trillion miles of driving data, mobility intelligence capabilities provide a better understanding of how people move, improve risk insights, and generate third-party revenue. The chart on the right shows the growth of the segment. Premiums earned and other revenue increased from $2.1 billion in 2021 to $3.5 billion over the trailing 12 months, reflecting annual average growth of 13%. Policies in force increased from 148 million to 178 million over that same period. Adjusted net income was $221 million over the last 12 months, compared to $218 million in 2025. The segment is continuing to expand while maintaining profitability. Protection Services expands our addressable market, broadens our distribution footprint, deepens customer relationships, and creates additional opportunities for growth. Now I'll turn it over to John to discuss the investment portfolio.

John Dugenske: Thank you, Mario. Let's move to slide 12, where we see strong investment performance is delivering earnings growth. Starting on the left, the portfolio is anchored by interest-bearing assets, which represent 80% of the $87 billion portfolio and provide a durable base of recurring income. Equity securities and performance-based investments complement that foundation with growth-oriented exposure. Together, this balanced mix is designed to generate consistent income while delivering attractive risk-adjusted returns on capital. On the right, since 2022, net investment income has increased 67% to $4 billion on a 12 month trailing basis, while the contribution to adjusted net income per diluted share increased from $7 to $12.2. Quarterly returns obviously vary across market cycles. Higher interest rates this quarter reduced fixed income unrealized valuations, resulting in a negative 2.1% return. Looking across cycles, the portfolio has generated 4.9% annualized return over three years and 3.9% over 10 years. We view our investment expertise as a competitive advantage, which is integrated into enterprise risk and return decisions. We actively adjust risk and position the portfolio to create value for Allstate and its shareholders. Now let's turn to slide 13, which shows how our broad investment capabilities translate into attractive performance across market cycles. Our integrated investment platform combines enterprise-level risk and return decision-making with active portfolio management. This means opportunities are pursued in the context of the entire company, considering market conditions, relative value, capital, and risk capacity. Internal teams provide fundamental research and analytical depth across a broad range of asset classes. We also use external managers to access differentiated expertise and scale. Together, this has resulted in first and second quartile performance across key asset classes, as you can see on the right. As an example of active management was a reduction in investment risk in 2022 as negative underwriting income and prospects for additional inflation increased overall enterprise risk. In 2025, as underwriting profitability improved and economic uncertainty was reduced, investment risk was increased by adding duration and $4 billion of public equity, as you can see from the charts on the bottom. These actions demonstrate our ability and willingness to adjust risk as market conditions and the company's risk profile changes. Now I'll turn it back over to Chris.

Chris Lown: Let's move to slide 14 to discuss how our capital allocation framework consistently creates value and generates attractive shareholder returns. Allstate deploys capital in multiple ways to create shareholder value: investing in existing capabilities, acquisitions and divestitures, and returning cash to shareholders. Investing capital in existing capabilities leverages Allstate's market presence with well-understood and attractive risk and return opportunities. We also use acquisitions and divestitures to optimize shareholder returns. We have a successful track record of acquiring companies with complementary capabilities, generating returns well in excess of our cost of capital. For example, the acquisition of SquareTrade expanded protection offerings and brought strong retail distribution partnerships. National General strengthened our independent agent presence. The recently announced acquisitions of INRIX and Replica will expand mobility intelligence solutions at Arity, providing another source of long-term growth. We also divest businesses to redeploy capital. Finally, the third component of our capital allocation framework is returning cash to shareholders. In the third quarter, we returned $1.5 billion to shareholders. Over the last 12 months, we returned $4.4 billion to shareholders and repurchased 5.7% of outstanding shares. Wrapping up with slide 15, we are delivering excellent operating results and increasing shareholder value while successfully executing Transformative Growth and deploying ALLIE. And with that I'd now like to open the line for Q&A.

Operator: Certainly. As a reminder, ladies and gentlemen, we would like to ask that you please limit yourself to one question and one follow-up. Our first question for today comes from the line of Josh Shanker from Bank of America. Your question, please.

Josh Shanker: Yeah, thank you for taking my question. Some questions on margins. If we look at 3Q 2025 and 3Q 2026, the auto underlying loss ratio deteriorated by about 600 basis points. It looks like earned pricing is down about 3% year-over-year. Can we talk about some of the severity and frequency trends? It just seems like a pretty high degree of deterioration for such a short period of time.

Tom Wilson: Josh, let me start, and then Jess will fill in. First, I would point out 91.6 is still well below our target of mid-90s, so we are in a really terrific return. When we decided to move that combined ratio up more towards the target, it was to drive growth, and you can see we did that successfully. That does not mean that that is the greatest trajectory is going to be. You can extrapolate from this a year ago to next year and get the same kind of deterioration. This is in our hands. We choose what we choose to do, and we are happy with where we are at. Jess can talk about frequency and severity.

Jess Merten: Yeah, thanks, Tom. I mean, first, I would start where Tom started, which is 91.4 on an adjusted basis is still better than our 95 target, and there's a few things that go into that. It's more than just the underlying loss trend. It's also what we've done from a top-line perspective to help affordability for our customers, right. Last year, we talked a lot about the SAVE program that included shifting customers to our most contemporary, affordable, simple, and connected products, which is better experience, more sophisticated rating plans, and updated features. We've adjusted deductibles. We adjusted limits. We did a lot to save our customers money, and you're seeing that in the average earned premium trend, which sort of has flattened out and was down a bit this quarter. The underlying loss trend, we look at that at a very granular basis. A couple of quarters ago, I talked about a multifaceted system that we use to manage the business, which includes operational levers and an intelligence layer, which is advanced analytics that helps us decide what to do on a state and sub-state basis, as well as an organizational structure. We have a go-to-market structure in place that's watching those underlying trends, Josh, on a very granular basis. We disclose pure premium in the investor supplement. In the deck, you can see what the trend looks like. I think we all observe that there is inflation in the environment, whether it's physical damage, bodily injury trends. There's clearly inflation out there, and we're watching that to make sure that we approach those targeted margins, again, mid-90s. The other thing I would highlight is we did take rate. It's disclosed in the investor supplement. In 39 locations, we made rate adjustments, and it had a 0.4 of a point impact overall on auto. We're addressing margin in a very granular way and watching what those trends look like, not just on a national basis, but on a state-by-state, market-by-market basis.

Josh Shanker: Look, what you do is very difficult. I get it. I'm being a little picky here, but I'm just trying to understand how reserving works. The first half of the year has proven intra-year favorable by 160 basis points, or at least 160 basis points to the third quarter. I noticed that, it's just one quarter, but 4Q 2025, slightly unfavorable as it's developed. These things are hard to get exactly right, but maybe you can talk a little bit about how the loss picks are initially set and when information comes in and when you react to it.

Tom Wilson: Let me start, and then Chris can jump in. First, of course, we always try to make sure it's as good a number as we think we can give you at the end of a quarter. We try to estimate, okay, what do we think we owe for the stuff that's out there? There's, of course, a whole bunch of different kinds of claims, Josh. A physical damage claim is settled within 90 days, so you kind of get that one right away. Property damage, a little bit longer, but still relatively short. The biggest swing, of course, you saw in the reserves is really bodily injury, where it takes about four years to get about 80% paid out on one cohort of claims. As that develops, you have that going on. At the same time, we're looking at the historical math and what are the trends, we're looking at the external trends, what's going on there. Then, we have claim practices that we're constantly getting better at. We feel like we've done a really terrific job over the last couple of years under some new leadership in claims, managing our costs, and you see that coming through. Chris, do you want to talk about the reserving process some and how you come up with a number?

Chris Lown: Yeah. Thanks, Tom. I think a few things. Obviously, it's a highly governed process. Obviously, we've been doing it a very long time. I think that we've been making a lot of meaningful changes over the last couple years. We brought machine learning in, so we're getting better, we're getting faster. Tom mentioned we have some new leadership, and I'd also point out that our trends aren't that dissimilar than the rest of the industry. There's a lot going on, but I think we feel very good about the reserve when we put up. It is our best estimate by the time we make it, and obviously the trend has been beneficial to us.

Josh Shanker: Thank you for all the details.

Operator: Thank you. Our next question comes from the line of Gregory Peters from Raymond James. Your question, please.

Greg Peters: Well, good morning, everyone. I am going to focus principally on slides four and five, maybe a little bit on 10. On slide five, you spent some time in your opening comments talking about the launch of the agentic assistance. I am sure there was a considerable amount of effort put forth to get you to this place. Can you provide some perspective on how you are going to be measuring the success of the launch? It is a loaded question, talk about some of the regulatory considerations, the cybersecurity issues, and maybe customer experiences that are rolled up into this launch.

Tom Wilson: Wow. They did not tell us that it was one question, but less than an hour answer.

Greg Peters: Sorry about that, Tom.

Tom Wilson: No, that is all right, Greg. I appreciate the comprehensive nature of it. Let me take a shot at some of it, trying to get to where you are headed with the underlying stuff. First, we think this will be really good for customers. I think consumers are going to like it. None of us like to do more clicks than we have to. Insurance is not a high engagement product, so people are not waiting to get in and scroll our website. Trying to find a way to increase our connectivity with them and make it easier and better and effective for them is really something we have been working on for a decade. This is just another leg. For example, if we get to mobility intelligence, we do crash detection. We want to give customers more value every day in as many ways as we can, including leveraging technology, and having a PAA will enable more people to come to us. If they do not feel like calling an Allstate agent or feel like getting on our website and they just want to have a PAA do that, we are fine with that. We think it is going to be good. I do not know how rapidly it will grow, but my sense is that the consumer demand is going to be strong enough. It will grow. We will figure out stuff like, what do we do with the regulators, what do we do with privacy, what do we do with cybersecurity, as we go. We are actively working with regulators now on AI, so we have been out proactively, not just on PAAs, but on claims and everything else, saying, "Let us tell you what it can do. Let us tell you how it can make our customers happier, how it can reduce their costs." We are working with There is a bunch of things that AI can do inside existing regulations. We do not need new stuff. There are some things that we will need to get additional regulatory approval to do, and there is some things we are going to have to do to make sure the regulators feel like they can look back at our processes and make sure we are actually doing what we said we did, because it is being done at light speed today. There is work to be done, but the benefit to customers and the economic rationale is going to move this train down the path. This is going to happen. We decided get in on it, accelerate it, go as fast as we can, which is why we started leaning into the PAA. In terms of the economics of a PAA for us, we think it is going to be fine. We are making money at Answer Financial, we make money in our other business. People are willing to pay you when you add good value. We do not know exactly how it is going to shake out, who gets what part, all that kind of stuff, but we are comfortable that we have a value proposition that works for our customers. Second to that, we believe that this will actually help the existing PAAs out there. If you are running one of these, like News, for example, and somebody asks you an insurance question and you have to go to eight different websites to get quotes, and you get blocked at some, you do not get blocked at others, we are like, "Let us make it easy for them." We think we can reduce their compute costs. We could make them do a better job for their customers. We feel very positive about where this might head, but do not know what the pace will be. We know our intention is not to be behind.

Greg Peters: Yeah. Related to this, on slide 10, you talk about increasing your investment in technology. You call it ALLIE, you call out higher software and computing costs. How do we reconcile this rollout and what you are doing with agentic AI and the overall cost structure of Allstate?

Tom Wilson: On the technology that we need to run our existing business, our commitment is to continue to reduce overall expenses while funding that. I might go back to Josh's thing. It is like reserving. It is not easy. But we get paid to do this. We have made a lot of changes in the way we deploy technology. Today, we have 25% less people in technology than we did four years ago, and that is not just using AI. Most of that happened long before we started using AI to code. We are constantly trying to come up with new ways to do it. As it relates to what it costs to launch it, and the PAA, and what we do, maybe I will ask Elizabeth to talk a little bit about where we are going with branding, and then I will share some conversations Chris and I have been having about how to give you transparency.

Elizabeth Brady: Sure. We are still evaluating names to determine what we should call our PAA, but we feel really confident that we have depth in this area, given our track record of building brands that are both broad-reaching and recognized, and also trusted, which is really important in this space. What really matters most is that customers know that our personal assistant exists, they know they can trust it, and they see the value that they get from it. Our PAA will sit directly in the flow of commerce, which gives consumers a new way to shop for insurance and bind, and it has the benefit, as Tom had mentioned, of end-to-end processing, which means you do not need to leave the experience in order to get your insurance completed. We are also interacting with other personal assistants. As you all know, there are many people who are starting to think about how they incorporate personal assistants into their day-to-day lives. We will be in that flow and connected to consumers' personal assistants as well. When we launch, we think that we have a strong acquisition tool and we have a strong customer connectivity tool that is trusted for shopping experiences and insurance. It should be a net positive for us.

Tom Wilson: As it relates to the cost to do this, Chris and I are talking about how can we show that to you. It might not show up in the existing expense ratio. Maybe we will break it out separately. Maybe we will footnote it. I do not know. But we will make sure, Greg, that you can understand what we are doing in the left-hand side of slide five, and what we are doing on the right-hand side.

Chris Lown: Yeah, the only thing I would add is that all this investment will allow us to decommission a number of systems over the next few years. Again, as Tom mentioned, we are highly committed to reducing costs. I see a lot of significant unlocks over the next couple years, so there will be benefits that will come through away from this as well, which we are excited about.

Greg Peters: Thank you for the additional information.

Operator: Thank you. Our next question comes from the line of Michael Zaremski from BMO. Your question, please.

Michael Zaremski: Hey, thanks. Good morning. On the capital discussion, just wanted to clarify the $9 billion of parent company, holdco capital. Does any of that need to be, or potentially need to be put down into any of the subs, or is that mostly free and clear? Obviously, you are accelerating some of your policy growth, so just curious.

Chris Lown: Yeah, it broadly equates to deployable capital. There is not a requirement to push any capital into the subs.

Michael Zaremski: Okay, perfect. My follow-up is about the accelerating auto insurance policy growth. I think in some of your we can see where it's coming from, exclusive agent versus independent versus direct. It seems like it's coming from all of the above. Maybe you could provide some texture. Is the ASC product a big driver? Is advertising spend maybe getting a higher ROI? Any additional, maybe new states, right? New York might be a better state today versus a year ago. Just any texture would be helpful. Thanks.

Tom Wilson: Let me start, and then Jess can jump in. We did show it comes from all three areas. That said, as we look forward, the direct, we believe, will continue to grow, and we made real progress in that segment, as I mentioned, and we'll talk about slide five. We think there's more room to grow there. Independent agents, we feel good about that. Jess should talk a little bit about the transition of Allstate agents to new business models. Then he can also talk about New York, or maybe I could just answer it. New York, we're still hoping to get our ASC product approved in New York so that we can go back into a growth mode, because as you know, we have a really significant share there, particularly downstate, and we can turn on the growth engines there when we get our new product, not until then.

Jess Merten: To build on that, maybe a couple things. One, I would start with, kind of as evidenced by the nature of your question and all the things you were wondering if there are unique drivers, there isn't one single answer to the growth equation. There's a bunch of different answers based on line of business, distribution channel, and really the go-to-market system that we have in place. A lot of this is organization that we have in place to make sure that we're optimizing across channels, products, and within the system. I just think it's important. That was built, we put it fully in place in this version a year or so ago. I think some of what you're seeing through this growth and adaptation is a result of that go-to-market system that we have in place. If you go a step below that, Tom mentioned as an example what we're doing in the exclusive agent channel. When you look at the numbers year-over-year, one of the things you have to sort of take note of is that we have about 5% fewer exclusive agents in the system this year versus last year. They continue to be more productive. That's important. We're doing this through a period where really, we're investing in the exclusive agent system to transition them to the future. We're intentionally shifting the agency capacity away from transactional work and service work towards higher value activities that we know customers value, whether that's new customer acquisition, protection reviews, claims advocacy, basically building deeper relationships, all supported by modern digital and AI-enabled tools, right? We're taking the agents who have been a critical component of our success, and we're bringing them into the future. We recently announced three new operating models that they're going to have an opportunity to choose between, effective January 1st of 2028. Middle of 2027, our agents are going to be able to pick from three different differentiated operating models, and we think all of them will set them up for the future. One is a sort of one of the models that we're calling the Protection Agent, as an example, is a simple, stable, virtual first model where an agent does a lot of the work by themselves. We have another model that's going to be more geared towards an agency that has a small team that's grounded in a local community. We're calling that the managing agent model. Then we have a separate model for agents that want to meet a broader range of customer needs, maybe have some dedicated specialized resources, maybe locations across several markets, and we're calling that Market Managing Agent. A key component, though, is the fact that we're providing choice. We're providing a system of support behind the agencies so that they have a very clear vision for the future. What I think that's going to allow us to do is to begin growing the agency footprint in late 2027 and into 2028, as we have clarity on models, we have a support structure for the future in our agency, not just existing agents, but potentially new agents see the fantastic opportunity that there is for them to create customer value and meet more customer needs. That's a sort of an answer about what's going on in EA within the overall growth landscape. But again, to your question, outside of Tom's response on New York, I wouldn't pinpoint any one thing. I would pinpoint the way that the system's working to generate the trends that you see.

Tom Wilson: Let me give you an example of maybe just from July, a conversation Jess and I were having. There are 11 states, about 13% of the population, which are growing in the mid-teens. Then there is another 23 states that maybe make up another 40% that are a decent amount above our average. Then there is obviously ones that are below that. Jess is looking at the ones below, he is looking at the ones above. He looks at the ones above and he says, "Hey, what is our combined ratio look like?" Yesterday we were talking about a couple of states who have some higher risk stuff, and they are growing, and the combined ratio is currently projected to be something above the target. He goes in, and he changes prices in those states. It is just to give you something tactical. He is just managing this all the time to try to get to what we commit to you, which is at a 95 combined ratio, this business generates a return on economic capital that is really attractive and above what the industry generates. That is the machine we are running.

Michael Zaremski: Thank you.

Operator: Thank you. Our next question comes from the line of Pablo Singzon from JPMorgan. Your question, please.

Pablo Singzon: Hi, good morning. There is more regulatory scrutiny these days in two areas. One is the cost of insurance that consumers pay and claims practices. Can you talk about Allstate's approach to these two topics and your confidence in your ability to respond to increased regulatory attention?

Tom Wilson: Yeah, Pablo, let me maybe make it even more contemporary and just talk about what just happened this week, because that may be a follow-on from somebody. The answer is we think we're in good shape to manage all of it. We don't think we have any big problems. But the impact of the blue wave is most important at the state level where we're regulated, so changes in governors, legislators, attorney generals, all that makes a difference to us, rather than a flip in D.C. So, five governors changed, seven legislators flipped. That said, when you're regulated at the state level, you don't have any favorites. We try to up our political activity. We're pretty balanced between these and ours. We just want people that work with us on important issues. Let me give you an example. Same topic, different parties, different states. In Florida, the Republicans pushed through tort reform, and we were able to lower our prices for customers by 6%. And we feel really good about that. They feel good about it. It goes to the issue you're talking about, affordability. In New York, Governor Hochul just got reelected. She pushed through some stuff on tort reform. That's not yet manifested itself in day-to-day claims costs, so we got to work with her administration and the regulators in New York to make that a reality. So, we kind of work both of them. The broader issue really is you got all new people, all with different agendas, different coalitions, and so that kind of means we got to just lean in. Now, we have people work on every state with the state legislators. We have, under Chris DeBiase, we have a government relations group, and they work in all these states. And the issues are typically availability of homeowners insurance, affordability, both auto and home, and some focuses on specific activities like claims and privacy. And our goal is just make sure we have affordable good products for customers. We work with people the way they want to work with them. And our big priority right now is trying to take on billboard lawyers who have decided that insurance is a stream of money they'd like to extract some wealth from, and so we're working aggressively to try to do that. But we're just trying to give the right product for our customers. At the federal level, there's been some conversation about the issues you're talking about, but the solutions are typically at the state level. We respond to those, we work with whatever senator or congressperson has an issue, but the solution really comes by getting something done at the state level where we're regulated. We are hopeful that with this change in Washington, there's a couple of issues we can go back to. One is FEMA. We need to ensure it has the capacity to deal with events that are bigger than can be handled by any local or state government. I'm not saying what was done was wrong, I'm just saying we just need to make sure looking forward it's got what it needs. NFIP reauthorization is still on the table, and we're a big servicer for NFIP, a bunch of our customers buy it, so obviously we want it to be reauthorized. We'd also like to see there's some potential to get some longer term solutions here. One of the problems with flood is not enough people buy it, and so if we can come up with a way to work with whatever party wants to take on those issues at the federal level, we'll do that.

Pablo Singzon: Thanks, Tom. That was helpful detail. My second question, many other large insurers with captive channels are changing how they compensate their agents and what sort of work is expected of them today, right? The question is, are you making similar changes currently, or is most of that work and restructuring behind you at this point? Thank you.

Tom Wilson: Well, if you're referring to what our big competitor to the south did, we made many of those changes before, so we would say we're past those. I think Jess started to talk about what we're doing with the exclusive agents. Maybe you want to give a little color on that.

Jess Merten: Yeah, I mean, I would agree with Tom that a lot of the changes that have been made across the industry are things that we did a while back. We continue to work with agents, and we will modify the compensation program to adjust for the work that they do. I mentioned that we're taking a lot of the service-related work out of the agencies because that isn't what generates customer value. We want to free them up to do the value-creating activity for customers. So we'll adapt our compensation programs to reflect the work that they're doing and to incentivize them to drive and deliver the value that customers are willing to pay for. I think rather than broad-based change like you've seen from some of the competitors, we're going to continue to refine our compensation program to align with what customers value, because we think that's what's important. It affects affordability, and it's really a customer-focused approach to looking at compensation, where agents can still make a great return on their investment in their agency, but we can better align what our customers are paying for distribution costs with the value that's being delivered.

Pablo Singzon: Thank you.

Operator: Thank you. Our next question comes from the line of Andrew Kligerman from TD Cowen. Your question, please.

Andrew Kligerman: Hey, thanks for taking my questions. Looks like you've got a lot of outstanding initiatives at Allstate and lots to follow up on. As I look at this quarter specifically, a couple of things I wanted to follow up on that Josh Shanker was touching on earlier. With regards to the prior year reserve releases, 9.1 points this quarter plus another 1.6 Q-to-date, and then if I look at the second quarter, it was 6.6 points, first quarter, 8.8, fourth quarter, 7.5. The gist of it is, I'd love to know more specifically what exactly is driving these releases, and maybe should I expect to see more of this in the coming quarters? It looks great. It's a great situation, so I'm wondering if I should model for it. I'd like to know what actually is behind these releases specifically.

Tom Wilson: Let me take a shot at it, and Rich can jump in. Of course, I wouldn't model anything. We think the number's exactly what it should be, and if we've got it right, then you would have zero next quarter. Our intention is to put up what we think it will cost. What's happened, and you quoted some numbers there, which are a lot relative to the combined ratio. You can also look at it relative to the reserve balances, which are slightly higher than that, or a lot higher than that. Nevertheless, it's a lot of money. The good news is that it's positive, and we're settling claims for less. The net of all that is, we just think we're going to pay out less money on these claims than we did before. There's a bunch of work we do in claims to try to make sure we do that. For example, we try to get to customers quickly after they have an accident, because most of these reserves and the releases come from bodily injury claims, not from physical damage. Remember, I was talking earlier, physical damage kind of happens right away, and you can adjust those relatively quickly, which I think a little bit of this quarter versus the prior two quarters of this year was a little bit of physical damage, because we're doing better at controlling how much it costs to fix a car, and making sure there's not extra cost loaded onto it. That said, on the bigger trend, it's all bodily. It's not all, but it's mostly bodily injury, which is related to uninsured motorists and UIM as well. That comes from trying to get there early, for example. What we find is the speed with which trial attorneys get to somebody who has an accident these days, using their data and analytics capabilities, is unbelievable. I'm not even sure how they get it done, because I'm not sure how they even get the data, because I'm not sure it's all legal, but I'm being defensive about it. The big law firms have data and analytics people, and they're out there crawling around. They get to a customer right away and they say, "Hey, buddy up with us. We'll get you more money." That stretches it out. So what we do is we try to get to them ahead of time, or we try to settle it faster, because we say, "Let's just not have this one delay, linger, and wait, because some lawyer wants to keep running up fees. Let's settle this thing faster." There's a variety of things we're doing to try to control costs, pay customers what they're owed. And remember, most of these bodily injury claims are people that our customers got into an accident with and hurt. So, in everything that we have to incur for them, our customers have to pay for in the end. So we're trying to make sure we pay what our customers really owe, and no more, no less. So, it's really no more complicated than trying to manage it that way. Chris talked about there's lots of math. I don't know if you have questions about how the math is done, but Chris has also got We've got Deloitte & Touche, KPMG. We've got enough eyes on this thing where we're always looking at to make sure that it is as accurate as it can be. But I would not forecast in your earnings estimate reserve releases in the future. It's not the way we run the place.

Andrew Kligerman: That was very helpful. Just help me square frequency and severity. Could you give any detail on what you saw in the quarter with respect to frequency and severity? Then I want to square that with written premium impact. So you had 0.4% impact of rate on written premium this quarter. Last two quarters, it was flat. So I'd like to just kind of square frequency and severity versus what you're doing in rate.

Jess Merten: Well, it all works together, right? I think that you're hitting on the right issues. Tom mentioned earlier his example of looking at all the different states. So what you see in the 0.4% is 39 locations, which means we're taking rate across a number of states to address the underlying loss trend. We don't break out frequency and severity, as you know. But you do see, whether it was in the slide or in the supplement, what our average loss and expense ratio is doing. It has been ticking up. There's a frequency and severity component to that. I mean, you know overall miles driven continue to trend down. We're seeing a little bit of frequency change in our book due to mix shift. Again, I will not put a number on that, but whether it is mix shift towards more new business, whether it is mix shift towards slightly higher risk drivers, that all is unique to the book that we are running. Again, not to be a broken record here, but that is state by state, market by market, where we are looking at what those trends are, how they develop alongside the severity trends, and we make adjustments. Those adjustments are what you are seeing in the rate taken in the four-tenths this quarter, which is we have some markets where we think we need to start to take some rate to offset the loss cost trend. Again, it is very granular. I know it probably does not put a number on exactly what you are looking for, but that is the way the system works.

Andrew Kligerman: But you feel like you are pretty close to frequency and severity in terms of not needing to take more rate?

Jess Merten: The forward trend? I mean, we are taking rate we need by market to address frequency and severity trends, knowing that we have a mid-90s combined ratio target that we want to get to. We have a rate trend that you have heard me say a million times, rates earn in over 12-18 months. As we have made rate adjustments a year ago, some of that rate trend is earning in now. So we are looking at the future trend of earned rate along with an increasing underlying loss and expense trend, and we are taking rate by state, by market as needed to anticipate what those trends are going to do together, Andrew.

Andrew Kligerman: Thanks. I appreciate it.

Tom Wilson: Why don't we do one more question?

Operator: Certainly. Our final question for today comes from the line of Tracy Benguigui from Wolfe Research. Your question, please.

Tracy Benguigui: Hey, Tom. You pointed to slide five, the upcoming Allstate Personal Agentic Assistant launch as your answer to the questions you're getting on Muse, and you sound positive about the opportunity. I'd assume, though, that customers may turn to Muse for price comparisons. I see on this chart you are showing other carriers on the bottom right. It is not clear to me. Are you saying that your proprietary agentic assistant will quote both Allstate and your competitors?

Tom Wilson: Yes. Yeah, it does today. Answer Financial does today.

Tracy Benguigui: Okay.

Tom Wilson: We represent over 30 companies. I can't tell you how much they sell for other companies because to respect their competitive position, I don't know. But yeah, we write $500 million of business, and not that much of it's Allstate. And they like it. I mean, it's good. It's another outlet for them to get business. And they can target what kind of customers they want. It helps them manage their portfolio. And most of them are excited about what we're In fact, all of them have been excited, and we're building straight-through processing with some of them.

Tracy Benguigui: Okay. So for my follow-up then is, even though this tool exists, for those customers that want to shop on Muse, how would ALLIE interact with these third-party agents? You talked about some folks blocking websites. Will you be simply redirecting Muse inquiries into ALLIE? Is that also a way to kind of meet your underwriting control requirements?

Tom Wilson: Well, first, we will be available to all personal agentic agents. We'd like to think we can have stronger relationships with some of them other just than MCP connections and linkage. So, we think that But we think the answer is we can help them do a better job for their customers. So, if you got to worry about credit cards and insurance and grocery shopping and everything else, you need some expertise. We have some really good expertise in protection, and so we can help them do a better job for their customers. It could be cheaper for them, and we can help do straight-through processing. So, we think it's a win-win for everybody, for the customer, for insurers, for us. It just, what it will do is, you're right, it will have increased price comparisons, but people shop three to four companies today anyway. This just makes it a little easier. And our pricing sophistication, you can see from our growth, is enough to enable us to compete aggressively in the market today, and we feel good about where we're going in the future.

Tracy Benguigui: Are you operating under the premise that Muse won't disintermediate your captive agency channel and that shopping demographics won't shift?

Tom Wilson: I'm operating under the assumption that Muse will create additional value for customers, that some of those customers may still want a local agent, even though they have Muse, because they had a problem with a claim once and an agent helped them, or they get access, or they may want to just do They want a branded product. Muse might show them XYZ company, it might show them Allstate, and they may say, "I'm more comfortable with Allstate." I think it's And then I would also say, I think people aren't really thinking as much about how much AI can help those local agents do a better job. It's not one or another. This happened kind of with the internet. Pretty soon, all retailers were going to go away, and they didn't. They adapted. Customers adapted and adjusted. We're just in the middle of it, and we're going to do whatever customers want, leverage as much technology in our brand. We think it's good for us, and it'll be good for our competitors. We're happy to be serving more customers. In some ways, PAA will be indifferent as to who the business goes to. It's going to go to who the customer wants. This is not like trying to steer stuff to us. The PAA will serve those insurers as an independent PAA that helps customers do what they want to do. Then if people want to call an Allstate agent, they'll call an Allstate agent, or they'll go by their office, or they'll do whatever.

Tracy Benguigui: Thank you.

Tom Wilson: Okay. Thank you all for leading in on Allstate. We will talk to you next quarter.

Operator: Thank you, ladies and gentlemen, for your participation in today's conference. This does conclude the program. You may now disconnect. Good day.