旅行家集團 (TRV) 2021 Q3 法說會逐字稿

Good morning, ladies and gentlemen. Welcome to the third quarter results teleconference for Travelers. (Operator Instructions) As a reminder, this conference is being recorded on October 19, 2021. At this time, I would like to turn the conference over to Ms. Abbe Goldstein, Senior Vice President of Investor Relations. Ms. Goldstein, you may begin.

Thank you. Good morning, and welcome to Travelers' discussion of our third quarter 2021 results. We released our press release, financial supplement and webcast presentation earlier this morning. All of these materials can be found on our website at travelers.com under the Investors section. Speaking today will be Alan Schnitzer, Chairman and CEO; and Dan Frey, Chief Financial Officer; and our 3 segment Presidents: Greg Toczydlowski, Business Insurance; Jeffrey Klenk of Bond & Specialty Insurance; and Michael Klein of Personal Insurance. They will discuss the financial results of our business and the current market environment. They will refer to the webcast presentation as they go through prepared remarks, and then we will take your questions.

Before I turn the call over to Alan, I would like to draw your attention to the explanatory note included at the end of the webcast presentation. Our presentation today includes forward-looking statements. The company cautions investors that any forward-looking statement involves risks and uncertainties and is not a guarantee of future performance. Actual results may differ materially from those expressed or implied in the forward-looking statements due to a variety of factors. These factors are described under forward-looking statements in our earnings press release and in our most recent 10-Q and 10-K filed with the SEC. We do not undertake any obligation to update forward-looking statements.

Also, in our remarks or responses to questions, we may mention some non-GAAP financial measures. Reconciliations are included in our recent earnings press release, financial supplement and other materials available in the Investors section on our website. And now I'd like to turn the call over to Alan Schnitzer.

Thank you, Abbe. Good morning, everyone, and thank you for joining us today. We're pleased to report strong top and bottom line results for the quarter and the first 9 months of the year, including very strong underwriting profitability and healthy top line growth. Core income year-to-date of $2.2 billion is about $800 million higher year-over-year, generating core return on equity of 11.6%. Core income for the quarter was $655 million or $2.60 per diluted share, generating core return on equity of 10.1% despite a high level of catastrophe losses. Our cat losses were well below our market share, well above the prior year quarter and the 10-year average for the quarter. Underlying underwriting income of $632 million pretax was 6% higher than in the prior year quarter, driven by record net earned premiums of $7.8 billion and a very strong underlying combined ratio of 91.4%. We are particularly pleased with the continued strong underlying fundamentals in our commercial businesses. The underlying combined ratio improved by almost 4 points in business insurance and more than 5 points in Bond & Specialty Insurance.

As you'll hear in a few minutes from Michael, underlying results in Personal Insurance were impacted by auto frequency returning to pre-pandemic levels and elevated severity in both auto and property due to higher costs for labor and materials. Our consolidated results demonstrate the value of having a diversified group of businesses.

Turning to investments. Our high-quality investment portfolio generated net investment income of $645 million after tax, reflecting reliable performance in our fixed income portfolio and very strong returns in our nonfixed income portfolio. These results, together with our strong balance sheet and cash flow enable us to grow adjusted book value per share by 10% over the past year after making important investments for the future and returning significant excess capital to our shareholders. During the quarter, we returned $821 million of excess capital to shareholders, including $601 million of share repurchases.

Turning to the top line. Net written premiums grew 7% to a record $8.3 billion. Each of our 3 segments again contributed meaningfully to the top line growth. In Business Insurance, net written premiums grew by 5% with renewal premium change of 9.9%, up more than 200 basis points year-over-year and near an all-time high. Renewal premium change was driven by continued strong renewal rate change and higher exposure growth. Importantly, at the same time, retention was also higher. Our ability to continue to drive price change at historical highs, while increasing retention, reflects excellent marketplace execution and the stability of the pricing environment.

In Bond & Specialty Insurance, net written premiums increased by 19%, driven by record renewal premium change of 13.6% in our management liability business and continued strong retention. We are also pleased to report strong production in our surety business. In our commercial businesses, written pricing continues to outpace estimated loss trend, which will continue to benefit margins as it earns in. Given social and other inflation, the frequency and severity of weather-related loss activity and the low interest rate environment, we expect the pricing environment to remain strong.

In Personal Insurance, net written premiums increased by 7%. Policies in force in both auto and homeowners were at record levels, driven by continued strong retention and growth in new business. Before I wrap up on results, I'd like to spend a minute discussing how our leading data and analytics and risk expertise contributed to our relatively favorable loss experience with Hurricane IDA. As I shared in our first quarter earnings call, our share of the industry's property cat losses over the past 5 years have been meaningfully lower than our corresponding market share. And while there is always the potential for us to have outsized exposure to an event, it was no accident that we again outperformed in Hurricane Ida. Our leading underwriting expertise supported by cutting-edge data analytics are key to an effective assessment of risk and reward. For us, third-party models are a starting point for our more advanced proprietary cat modeling. At the portfolio level, the insights from our models warned us away from the coast to our item at landfall, given the impediments to achieving an appropriate risk-adjusted return.

In the other states who are on Ida's path, we effectively managed risk selection, pricing and other terms and conditions by putting sophisticated data and analytics at the fingertips of our frontline underwriters. These include robust flood risk scoring, location intelligence down to the parcel level, hill dashboards and output from our risk control engineers.

Within Personal Insurance, we continue to see the benefits from our highly segmented Quantum Home 2.0 product, which is now rolled out in more than 40 states. In the northeast, extreme rainfall from Ida resulted in significant claim activity for the industry, including from water and drainage backup, which is a coverage we provide in our QH 2.0 product. The model underneath the product leverages data and analytics to underwrite and price that coverage on a very granular basis. In addition to underwriting, data and analytics are increasingly informing our claims handling strategies. For example, our AI-assisted claim damage detection model was a key part of our Ida claim response. This model uses AI and high-resolution aerial imagery to detect the extensive damage to individual properties as soon as a day after an event. Within 2 days of impact, we were collecting and analyzing aerial imagery of customer properties along Ida's path as it moved across 20 states. This enabled us to remotely identify which of our customers' properties had sustained exterior damage and effectively organize our claim response.

In some cases, we can use this technology to adjust and pay total losses before the customer has even been able to return to their homes. We also utilized other virtual abilities in our Ida response, such as image share and live video capture on a majority of claims with interior damage. These leading-edge capabilities enhance a claim experience for our customers by cutting significant time out of the claim process, expediting an accurate loss assessment, and in many cases, eliminating the need for physical inspection. Again, with Ida, we've successfully closed 90% of all homeowners claims within 30 days. All of this also results in a more efficient outcome for our shareholders. As strategic as the data and analytics are, maybe even more important is the culture that brings it all together, our collaborative approach to developing a holistic 360-degree view of risk incorporating underwriting claim, actuarial, risk control, legal and regulatory input is an important differentiating factor in effectively managing risk and reward. That culture is decades in the making and very hard to replicate.

Before I turn the call over to Dan, I'd like to welcome Jeff Klenk President of our Bond & Specialty Insurance segment to the call. As I shared last quarter, Jeff is a 22-year veteran of Travelers, most recently as a member of Tom Kunkel's leadership team and Head of our Management Liability business. Jeff succeeded Tom following his retirement last month. We're fortunate to have Jeff in the role, and you'll hear from him in a few minutes.

To sum it up, we're pleased with our results for the quarter and year-to-date. Our significant and hard-to-replicate competitive advantages position us very well to continue to deliver meaningful shareholder value over time.

And with that, I'm pleased to turn the call over to Dan.

Thank you, Alan. Core income for the third quarter was $655 million compared to $798 million in the prior year quarter. For the quarter, core return on equity was 10.1%. And on a year-to-date basis, core ROE was 11.6%. The decline in core income was driven by prior year reserve development and catastrophe losses.

Recall that last year's PYD benefited by approximately $400 million of subrogation recoveries from PG&E. Most unfavorable year-over-year comparisons were partially offset by increased investment income and a higher level of underlying underwriting income. Underlying underwriting income increased 6% to $632 million pretax, reflecting a higher level of earned premium in all 3 segments and a strong underlying combined ratio of 91.4%. Improvements in the underlying combined ratio in both Business Insurance and Bond & Specialty were offset by an increase in the underlying combined ratio in Personal Insurance.

Business Insurance and Bond & Specialty results both reflected the benefit of our pricing efforts, as earned price continues to exceed loss trend. We expected a higher underlying combined ratio in Personal Insurance, given that last year's quarter benefited from unusually low auto losses related to the pandemic. As Alan mentioned, the underlying combined ratio in PI was further impacted by higher severity in both the auto and homeowners products. Greg, Jeff and Michael will provide more detail on each segment's results in a few minutes.

On a consolidated basis, the underlying loss ratio for the quarter improved slightly to 62% compared with last year's 62.2%. The expense ratio of 29.4% was in line with the prior year quarter and in line with our expectations. We've improved the expense ratio by more than 2 points from where we were 5 years ago. Having added roughly $7 billion to our annual net written premiums over that period while maintaining a focus on productivity and efficiency all while adding significantly to the level of strategic investment we're making to ensure our future success. Before turning to catastrophe losses, I wanted to point out that within our underlying combined ratio, non-cat weather was higher than we would have assumed for the quarter, although lower than the unusually high levels we experienced in last year's third quarter.

Our third quarter cat losses were $501 million pretax compared to $397 million a year ago. Remember that in last year's third quarter, we had a full recovery under the Property Aggregate Catastrophe XOL Treaty. In this year's third quarter, we recognized a partial recovery of $95 million from the treaty, $83 million benefiting the cat line with $43 million in Business Insurance and $39 million in Personal Insurance and $12 million benefiting our underlying results with $3 million in Business Insurance and $9 million in Personal Insurance. That leaves us with $255 million of potential recovery in the fourth quarter, depending on the level of qualifying losses we actually experience.

In terms of the level of cat losses relative to our assumptions, third quarter cats were elevated compared to what we would have assumed for a typical third quarter. Although our losses from Hurricane Ida are well below our relative market share, the sheer size of IDA on top of the other cat losses in the quarter resulted in overall cat losses that were higher than our assumption.

Turning to prior year reserve development. In Personal Insurance, $30 million of pretax net favorable PYD resulted from better-than-expected experience from recent years in the property line. In Bond & Specialty Insurance, $22 million of pretax net favorable PYD was driven by favorable loss experience in the surety product line for recent accident years. In Business Insurance, net unfavorable PYD was $108 million pretax. Our annual asbestos review resulted in a charge of $225 million, as the level of claim activity did not decline as much as we had assumed in our previous estimate.

Excluding the asbestos charge, Business Insurance had net favorable prior year reserve development of $117 million, driven primarily by better-than-expected loss experience in workers' comp across multiple accident years. Net investment income improved to $645 million after tax this quarter. Our non-fixed income portfolio delivered another strong result, contributing $224 million after tax. As you can see from the detail provided on Page 6 of the webcast presentation, our recent results in the non-fixed income portfolio have been unusually strong, and I would caution you that this level of returns is not likely to continue.

Consistent with our expectations, fixed income returns were down slightly from the prior year quarter as the benefit from higher levels of invested assets was more than offset by a decline in yields. For the fourth quarter of 2021, we expect NII from the fixed income portfolio, including earnings from short-term securities of between $425 million and $435 million after tax. For 2022, we expect that, that figure to be between $420 million and $430 million per quarter.

Turning to capital management. Operating cash flow for the quarter of $2.5 billion was an all-time record. All our capital ratios were at or better than target levels, and we ended the quarter with holding company liquidity of approximately $2 billion. The market value of the bonds in our fixed income portfolio declined as U.S. treasury yields increased and credit spreads widened during the quarter. Accordingly, our after-tax net unrealized investment gain decreased from $3.2 billion as of June 30 to $2.7 billion at September 30.

Adjusted book value per share, which excludes net unrealized investment gains and losses, was $104.77 at quarter end, up 5% since year-end and up 10% since September 30 of last year. We returned $821 million of capital to our shareholders during the third quarter with $220 million of dividends and $601 million in share repurchases. Overall, a very good quarter for us with healthy top line growth, strong and improved underlying margins in our commercial businesses, excellent cash flows and a terrifically strong balance sheet.

And with that, I'm pleased to turn the call over to Greg for a discussion of Business Insurance.

Thanks, Dan. Business Insurance had another great quarter with strong financial results and terrific execution in the marketplace. Segment income was $558 million for the quarter, up more than 50% from the prior year quarter. The improved year-over-year result was driven by higher underlying underwriting income, prior year reserve development and higher net investment income, partially offset by higher catastrophe losses.

We're once again particularly pleased with the underlying combined ratio of 90.2%, which improved by 3.8 points from the third quarter of 2020, primarily attributable to 3 things. First, about 2 points of the improvement resulted from earned pricing that continued to exceed loss cost trends. The other nearly 2 points resulted from a combination of a favorable impact associated with the pandemic and a lower level of property losses.

Turning to the top line. Net written premiums were up 5%, benefiting from strong renewal rate and exposure levels as well as high retentions. As for domestic production, renewal premium change was once again historically high at 9.9%, while retention increased to an exceptional 85%. The 9.9% renewal premium change was up more than 2 points from the third quarter of last year, with strong renewal rate change of 6.3% and continued improvement in our customers' exposure growth.

In addition to our granular price execution, we've also focused on careful management of deductibles, attachment points, limits, sub-limits and exclusions, which can also contribute to an increase in the price per unit of risk. The business was down from the prior year quarter as we continue to be thoughtful about balancing risk and reward for new business in the marketplace.

As for the individual businesses, in Select, renewal premium change was a strong 9.7% while retention improved from recent periods to 82%. Underneath RPC renewal rate change was 4.1%, up well over 1 point from the third quarter of 2020. We're also encouraged with the improving exposure, which was up about 3 points as the economy continues to reopen. New business was up a bit from the prior year, driven by the continued success of our new BOP 2.0 product, which is now live in 39 states. In middle market, renewal premium change of 9.5% and retention of 88% were both historically high. Renewal rate change of 6.2% remains strong. As always, we remain disciplined around risk selection and underwriting.

To sum up, Business Insurance had another terrific quarter. We're pleased with our execution in further improving the underlying margins in the book and we continue to invest in the business for long-term profitable growth.

With that, I'll turn the call over to our new partner in the room, Jeff Klenk.

Thanks, Greg, and thank you, Alan, for that welcome. I'm very pleased to be here. We've built an industry-leading surety and management liability franchise, and we look forward to continuing to perform, innovate and transform to profitably grow these businesses into the future.

Turning to the results. Bond & Specialty delivered excellent returns and growth in the quarter. Segment income was $174 million, up about 50% compared to the prior year quarter, driven by the impact of higher net earned premium, a significantly improved underlying underwriting margin and favorable prior year reserve development. The underlying combined ratio of 83.4% improved by 5.5 points from the prior year quarter, reflecting lower pandemic-related loss activity, earned pricing that exceeded loss cost trends and a lower expense ratio.

Turning to the top line. Net written premiums grew an exceptional 19% in the quarter, with strong contributions from all our businesses. In domestic management liability, renewal premium change was a record 13.6%, driven by record renewal rate change. Retention remained strong at 86%. Consistent with recent quarters but down a few points year-over-year, as we continue to non-renew cyber policies for accounts that don't meet our updated minimum requirements for cyber hygiene. Notably, research indicates that implementing affordable cyber risk mitigation controls, such as multifactor authentication, could prevent the vast majority of ransom attacks.

Domestic surety also posted strong growth relative to the pandemic impact of prior year quarter. In addition, our international businesses again posted excellent growth, including strong management liability retention and rate. So both top and bottom line results for Bond & Specialty were terrific this quarter, demonstrating our thoughtful approach and strong execution across our businesses.

And now I'll turn the call over to Michael.

Thanks, Jeff, and welcome. Good morning, everyone. In Personal Insurance, bottom line results were impacted by weather or returned to pre-pandemic claim frequency in auto and higher loss severity impacting both auto and home results. Segment income declined by $394 million from the prior year quarter. $262 million of that decline is attributable to lower favorable prior year reserve development as the prior year quarter benefited from the PG&E settlement Dan referenced. The remaining unfavorable variance was primarily driven by lower underlying underwriting results. Our underlying combined ratio increased by 6.5 points to 95.2%. We were pleased to see our top line momentum continue in the quarter with net written premiums up 7%.

Automobile underwriting results reflected higher loss levels for the quarter. The combined ratio was 100% and included 3.4 points of catastrophe losses, mostly from Hurricane Ida. The underlying combined ratio was 97%, up approximately 15 points from a prior year quarter, which reflected unusually low loss activity due to the pandemic.

The underlying combined ratio increased mainly due to claims frequency effectively returning to pre-pandemic levels in line with the trend we referenced in our prior quarter call. To a lesser degree, higher loss severity impacted the combined ratio as well, as vehicle replacement and repair costs remained at elevated levels. We believe these profitability challenges are environmental. And in response, we are executing on our plan to file rate increases in about 40 states over the next 3 quarters. As we indicated last quarter, it will take time for future rate actions to earn into results, but we do expect to have higher rates in market in several states by year-end.

In Homeowners and Other, the third quarter combined ratio increased by 16 points from the prior year quarter to 109.3%, driven by a 24-point reduction in net favorable prior year reserve development, primarily related to the PG&E recovery from last year. The combined ratio included 17.6 points of catastrophe losses, mostly from Hurricane Ida. Homeowners catastrophe losses were 4.7 points below a very active prior year quarter.

The underlying combined ratio was 93.3%, an improvement of 3.5 points over the prior year quarter, which experienced a very high level of loss activity. That said, the underlying combined ratio was above our expectations, reflecting elevated non-catastrophe weather and non-weather loss activity, both of which included higher severity related to a combination of labor and materials price increases. Again, we believe these trends are environmental, and we continue to see price increases in response.

Before I shift to discussing production, I'll remind you that looking ahead to the fourth quarter, there tends to be a good amount of seasonality in our combined ratio results by line of business, with the fourth quarter auto losses typically higher and fourth quarter Homeowners losses typically lower than their annual average levels.

Turning to production. We were very pleased to deliver another strong quarter in both auto and home. Automobile policies in force grew 5% to a record level, driven by strong retention at 85% and continued growth in new business, which increased by 8%. Renewal premium change was essentially flat, reflecting the continued impact of the rate increases we filed in response to the pandemic. Domestic Homeowners and Other delivered another excellent quarter with policies in force up 7%, also to a record level, driven by retention of 85% and new business growth of 5%. Renewal premium change increased to 8.8%.

In the quarter, we continued to deliver solutions that meet customers where they are, give them what they need and serve them how they want. A couple of highlights from the quarter include: our new digital self-inspection process for property customers, which improves their onboarding experience and provides valuable information to our underwriters. And IntelliDrive, our proprietary auto telematics offering, which now has distracted driving as a rating variable in 40 U.S. markets and Ontario, Canada, providing valuable feedbacks to drivers and continuing to advance our sophisticated pricing segmentation in automobile. We will continue to invest in new capabilities like these for our customers and distribution partners.

Despite a challenging third quarter on the bottom line, we remain pleased with our overall performance, and we are well positioned to profitably grow our business over time.

Now I'll turn the call back over to Abbe.

Thanks, Michael. We're happy to turn to your questions.

(Operator Instructions) Your first question comes from the line of Michael Phillips with Morgan Stanley.

First question, Alan, I think might be to your opening comments on BI and expecting -- you listed a number of reasons why pricing environment should remain strong. I guess, maybe help us define or think about what you mean by strong and how we can think about maybe margins for BI into next year or I guess the back half of next year? We have pricing now kind of close to where loss trends are. So just kind of help think about when -- with those reasons why prices remain strong? Should we expect this pricing to -- it seems like start to converge with where loss trends are, which obviously could have some impact on margins into 2022? So just comments on what you meant by strong there.

Thanks for the question. You're very narrowly, I think, focusing on rate versus loss trend, and by the way, there's still a margin there. And even if you look at pure rate over some long historical period, the rate we continue to get is actually a pretty good number. But if you look at the overall pricing number, I mean, at 9.9%, that's practically an all-time high, and there's a lot of margin in that exposure. So we continue to see a nice spread between written price and loss trend. I do think that's the right way to think about it.

Nonetheless, as we -- we know what our product returns are. We know what the headwinds are. We know what our objectives are in terms of our returns, and we know what we need to do. So I think that from here, pricing will continue to be strong. Now we've also been getting pretty good pricing for some time now, and it's been compounding for a while. And so you've got different lines in different places. Some need more help than others.

And so from here, the rate of increase is going to be impacted by that. So some lines are -- the rate of increase will be higher, some will go sideways, some still be an increase -- the rate of increase may go down. But overall, we continue to expect pricing to remain strong. I'm not sure we're going to put much more dimension around it other than that but hopefully that gives you some color to the top.

It does. Yes. I appreciate that. Second question for you or Mike maybe, on personal auto. And Mike, I appreciate your comments on -- it was more of a frequency back to pre-pandemic levels, to a lesser degree, the environmental issues with low severity. So I guess, do you think the loss severity side there on auto, has it peaked? Will it get worse from here? Do you think? And how long lasting will that be? I ask that because I wonder if it's to a lesser degree that. The 97% core, it's not that much higher than where you were pre-pandemic, 96%, 95% in that kind of range. So I asked those questions because I wonder how much rate you think you need in personal auto if those things have already peaked? I'm not sure if they have or how long they might last and your comment was there to a lesser degree anyway. So any comments there would be appreciated.

Yes. Sure, Michael. It's Michael here. I would say a couple of things about that. So to dimensionalize that a little bit more for you, the auto underlying combined was up 15 points in the quarter. We'd say 3/4 of that was loss experience returning to pre-pandemic levels and about 1/4 of -- about a remaining 1/4 of that is the trends in increased cost to repair and replace that we talked about last quarter that essentially continued into the quarter.

In terms of how long it's going to last, I can point you to a little bit of external data that might help there. There's certainly pressure on the system, right? A lot of the costs relates to total losses and replacement cost values for total vehicles, which tracks pretty closely with used car prices. And inside the quarter, we were actually encouraged by the July and August data from places like the Moody's used car price index that looked like the peak had happened in May and June and was starting to decline in the first part of the quarter. Now that data ticked back up again in September, and I think its kick back up reflects the pressure on the system.

You had IDA occur in the quarter. Estimates are that over 200,000 vehicles were destroyed in that event at the same time that you have vehicle inventories in the U.S. at historic lows. And so I think we're going to see continued pressure on the cost of repair and replace related to supply chain, related to labor and related to used car parts and prices. But again, that was about 1/4 of the 15 points, and we will factor that into our underwriting pricing decisions as we move forward.

Yes. And I would just add to that. The good news is this is impacting short-tail lines where we can see it very quickly and react to it. And the reserves are obviously significantly way less leveraged, given the duration of those liabilities.

Your next question comes from the line of Mike Zaremski with Wolfe Research.

Maybe moving to focusing on the business segment, in Greg's commentary, parsed apart the year-over-year improvement in the underlying loss ratio, it was noted that 2 points came from a favorable impact from the pandemic. Maybe you could give us more details on that. And I believe that's a more favorable impact versus what you've laid out in previous quarters.

Mike, it's Dan Frey. So just to clarify Greg's comments about a 4-point improvement, you had about 2 points come from earned price versus trend and then 2 points come from the combination of favorable impact of the pandemic and some favorability in non-cat property losses. You're right in terms of the impact on Business Insurance from the pandemic.

Last year, in the third quarter, we gave you that the favorable impact was about 0.5 point and that was the net of direct charges that we took related to COVID, things like workers' comp charges for COVID-19, somebody contracts the disease and that was more than offset by indirect impacts, things like lower frequency on commercial auto. This year, you've got a larger net impact, but it's really the impact of the absence of those direct charges. We're not seeing that anywhere near the same level we saw a year ago. But the 2 points you're talking about was not just attributable to COVID, it was attributable to the combination of COVID and non-cat property experience.

Got it. And has any -- if we parse through all the lost cost commentary between commercial lines and personal lines, I know there's a lot of moving parts. But would you say on a 100,000-foot view that Travelers has a kind of an updated view of loss trend?

Yes. So a little different between the 2. In Business Insurance, you got a lot of moving parts, including mix of business, but we haven't seen a meaningful change in our view of long-term normalized trend. In Personal Insurance, given Michael's comments, we've taken a slightly increased view of what we say is our view of long-term normalized trend.

Your next question is from Elyse Greenspan with Wells Fargo.

My first question is on personal auto as well. So as you're talking about trying to take some price, I believe you said over the next few quarters. Are you concerned as you look for greater price that there could be some pushback from regulators, right? We've -- obviously, loss trend has spiked of late, but it seems like you're also looking at 2020, which was an extremely profitable year for the industry. So how do we think about regulators and just the level of price you're looking to take within your personal/auto book?

Sure, Elyse. It's Michael. Thanks for the question. And before I address the question, I think I may have misspoken. In the script I think I said that auto RPC was impacted by the rate increases we filed in response to the pandemic. Obviously, we didn't file rate increases in response to the pandemic. Those were rate decreases, which, by the way, I think, is relevant to your question, right?

So as we are in the conversations with regulators, which are active and ongoing, as we speak, we're in conversations with them about that loss history. We're also in conversations with them about the actions we took in response, which included rate decreases in a variety of jurisdictions, which included premium refunds to customers in a variety of jurisdictions. And so we're having conversations with them around how we align pricing with loss costs over time and engage in discussions with them about how to view the 2020 loss experience in that context.

If I take a step back and think about the process a little bit more broadly, what we would say about the process of getting rate is that it is an ongoing dialogue between us and the regulators. It's about an exchange of information to justify the pricing that you think you need on a go-forward basis. And generally speaking, what we have found is, we've been able to get to a point where we can align price with loss cost over time.

You'll have some challenging conversations in some jurisdictions and maybe less challenging conversations in others. To put a little bit of a point on the 40 jurisdictions, we do expect some of those increases to become effective in Q4 and actually a couple of them literally this month. So we're on the way to getting those increases filed and approved. And then as we've talked about, it will take time for them to get written and then earned into the book of business but we're making progress.

And then my second question just goes back, I guess, to the prior question as well. You guys have been within BI at around that 5% overall loss trend for, I think, like 6 quarters now. Obviously, it sounds like things have been slow to fully get back to normal with the pandemic. But as course have started to be open, has there been anything you guys are paying attention to just on the inflation side as we think about loss trends as well as expectations as we go into 2022?

Yes. Elyse, as you'd imagine, we're looking at all the data that comes in every quarter. And there are some things that are moving up and there are some other things that are moving down. It varies by line. We've got still some frequency benefit in the commercial auto space. There's some severity in certain elements, but those are things that were more levered to in Michael's business than we are in Greg's business in terms of the relative mix of loss cost. So we've got our finger on the pulse of all those things. Our aggregate conclusion is, there's not much of a change there.

And, Elyse, I'd add to that. There was nothing unusual about this quarter relative to any other quarter. We're always looking at frequency and severity in every line. And occasionally, there's something that will move here and there. But again, this is a view of long-term trend and so nothing really unusual in the quarter.

Your next question comes from the line of Ryan Tunis with Autonomous Research.

I just had a couple for Michael. The first one is thinking about your auto book as being largely preferred. I guess, in your mind, what are some of the advantages and disadvantages of having a preferred book versus a less preferred book when we think about the loss trend environment post pandemic?

Yes. Good question, Ryan. I think there's a couple of things there. We, generally, I would say, are going to be impacted by commuting patterns that are more related to office jobs when you think about the general profile of our preferred book. I think the other dynamic you see in a preferred book though is newer vehicles with more technology. And so when you think about some of the cost pressures on repair and replacement costs, that's going to put upward pressure there.

And I would say that while we saw, I think, some differences over the past 12 to 18 months that you could sort of point to in different parts of the country, different books of business in terms of how the pandemic impacted different segments differently. As we sit here today, the economy is broadly recovering. The country is broadly recovering. So you don't see quite as many differences, I don't think, geographically or by segments that are driving differences in loss experience across the segments of the industry. The other thing that's true of our book of business is it's heavily rounded. And so it's -- while we talk a lot about the individual lines of business, we certainly look at customers in the total portfolio as we look to manage it.

Got it. And my follow-up is hearing you say that frequency is back at pre-pandemic levels, I guess I'd always thought about pre-pandemic as kind of being a ceiling for frequency, where it would return to. Is that how you're thinking about it? Or based on what you're seeing, do you actually think frequency could run about pre-pandemic levels?

Yes. It's a great question, Ryan. I would say, underneath frequency, we look at driving activity, we look at trips, we look at mix of activity. And a lot of the patterns that we've talked about in the past, we continue to see. So on the one hand, we continue to see whether we look at our own IntelliDrive data, whether we look at Google mobility data, again, fewer commute trips. On the other hand, we actually see elevated shopping trips and recreational trips. So the mix of miles has changed.

But mileage broadly, when you look at Department of Transportation data, again, our IntelliDrive data, MS2 sensor data, is back to pre-pandemic levels as is frequency. So again, the mix looks a little bit different, but miles driven trips broadly are back sort of close to where they had been. And so as we talked about a little bit earlier, we see -- the frequency return sort of in line with the trends we outlined last quarter coming into this quarter.

Your next question comes from the line of Jimmy Bhullar with JPMorgan.

I had a couple of questions both on the commercial side. First, can you talk about what you're seeing in terms of pricing in some of your main lines, specifically commercial auto and then workers' comp? Have you seen a bottoming there or is it still soft? And then I had another one.

Yes, we continue to see workers' comp. This is Greg, by the way, workers' comp bump below the water level. And so we had a slight negative with workers' comp. Other than that, we continue to be led with excess casualty and umbrella automobile and property that gives you just a little bit of color in terms of the range of rate activity from the most positive to lease.

Okay. And then how do you think about inflation affecting loss costs in commercial lines? I think Alan had mentioned like loss -- you're assuming loss costs are fairly stable, but have you seen inflation begin to have an impact? Or do you not think that's a factor on the commercial side?

Yes. Jimmy, we certainly see it, but you got to start with what are we seeing relative to our expectations because we certainly expected some level of inflation. Also on the commercial side, say, relative to Personal Insurance, you got to think about the impact on a mix-adjusted basis. So you've got the casualty and the property in there. You don't see this cost of goods sold type of inflation so much on the casualty side.

And while we did see a little of it on the commercial side, there was an offsetting frequency benefit as you've heard both Greg and Dan talk about. So it's not that we don't see it. It's not that we're not aware of it, but we expected some and it just doesn't have the same overall impact on that segment.

Okay. And if I could just ask one more on personal auto, I think you were talking about sort of thinking about raising prices on the call last quarter as well. And it doesn't seem like any of that was in your results this quarter. Have you already started implementing price hikes? Or is it more -- is more of the benefit going to flow through in 2022, just given the timing of getting -- often getting permission and then implementing those?

Sure, Jimmy. It's Michael. I think, first of all, we would not have expected any rate increases to take effect this quarter. As we talked about it last quarter, we were beginning to -- the process of seeking rate, which means we've got to put together the filing package, we've got to file that with the state, we've got to negotiate with the state, we've got to get it approved, we've got an agree upon an effective date, which is always in the future from the date you get the approval.

So there's a fairly long tail on the front end of any rate filing process up or down. And so we really wouldn't have expected any of those increases to take effect in the third quarter. As I mentioned, we've got 40 planned over the course of the next 3 quarters. And a couple -- 2 or 3 will likely take effect this month on several, call it, another handful of take effect probably before year-end, again, all subject to getting approvals from the states that we are working towards those effective dates with and then the remainder will be in the first or second quarter of next year.

The few that we have likely to take effect in the back half of this year will represent 25%, 30% of the written premium volume we have across the country. So it's a relatively small number, but it should be impactful on a written basis. But again, I'll just caution you, it takes time for that to earn into the book, but we've started and we are making progress.

Your next question comes from the line of Paul Newsome with Piper Sandler.

I wanted to ask about social inflation and if you believe or how we should think about the mechanism of sort of regular CPI inflation creeping into the social inflation side of it. Is it something where we should expect some sort of lag there as courts assess the damages over time?

Paul, I don't think it's entirely unrelated or uncorrelated, but I think it's less leveraged. I think of social inflation as being driven more by things like aggressive tactics by the plaintiff bar and advertising and litigation finance and things like that. Now of course, that is -- the whole debate on damages, in any case, starts with the underlying compensable damages. And to the extent there's inflation in that, maybe the starting point goes up a little bit. But so not completely uncorrelated, but also not so leveraged.

Great. And then my second question unrelated to auto -- it's not a question. There seems to be efforts in the industry to effectively replace credit scoring with telematics and not just using together. What do you think of those efforts? And is that something that Travelers has a particular year?

Sure, Paul. It's Michael. Certainly, there are conversations that are linked to 2. I guess I'd start by separating the 2, right? There are certainly conversations with state regulators and a handful, I would say, that are talking about and are actively pursuing the removal of credit scoring as a variable in pricing auto insurance.

The 2 most notable are Washington and Colorado. Colorado has a bill that was approved by the state legislature that's in the rule-making process around removing credit score in that state. In the State of Washington, again, pretty widely followed. But Commissioner Kreidler issued an emergency rule to ban the use of credit scoring in auto insurance. That emergency rule was actually just overturned by a judge in Washington earlier this month. And so we're in the process. And that -- the overturn by the judge is now being challenged by the commissioner. So there's a back and forth in Washington. Those are probably the 2 most widely viewed views on credit scoring.

As respect to credit scoring and telematics, what we would say is, first of all, credit is a powerful variable in pricing auto insurance. It's very predictive of claim experience. And if you remove it then you have subsidization in your rate plan between higher risk drivers and lower risk drivers. And in fact, in the context of a telematics program, the credit score and the telematics data are powerful together.

If you're in a situation where you don't have credit and can't use credit, telematics is very valuable, probably becomes increasingly valuable. And in that kind of a world, we're encouraged by the capability we've got, the tools we have, the success of our IntelliDrive program and the power that it has in segmenting loss experience and helping us segment pricing.

It sounds like you're agnostic to the use, in general.

Paul, I wouldn't say we're necessarily agnostic. I think as Michael pointed out, it is highly predictive. And so we support and promote risk-based pricing, but you end up with some subsidization and so it is what it is. The fact is, we compete without it in places like California just fine, and we'll continue to compete without it. And in that world, having the data and scale to invest in other rating methodologies is important, and we think we have that. So I think either way it would be just fun.

Your next question comes from the line of Josh Shanker with Bank of America.

Guess what, more personal/auto, I apologize. I'm just wondering, it looks like that used car severity peaked in May and then again in September and vehicle traffic levels seem to get back to normal by about May, yet the 2Q loss ratio in personal/auto was materially higher than the -- actually the 3Q was materially higher than the 2Q loss ratio. Is there any intra-year development in that? What's the difference between the 2Q and 3Q loss experience you guys had?

Josh, I would say the difference is sort of precisely what you just described. So you talked about driving levels in May, which is partway through Q2, you talked about severity in May, which is partway through Q2. And so as we talked about in the second quarter, we saw frequency trending toward pre-pandemic levels throughout the quarter, which means it was below at the beginning of the quarter. And similarly, some of the severity impacts we've been describing really started to impact the results in the back half of Q2 and into Q3. So that's really, I think, what's going on there, and again, no intra-period or intra-year that adjustment talking about, right?

Yes, Josh, it's not a catch-up. Our second quarter loss pick looks pretty consistent with what we had coming out of the second quarter. It's really the difference of a full quarter of the higher frequency and severity, Q3 versus Q2.

Great. And your growth looks fairly strong. I think we can all agree that the price is underpriced right now. Can you talk about growing your book at a period of time when the prices are below rate-adequate levels?

Sure, Josh. Michael, again. As we've talked about pretty consistently, right, our goal is to profitably grow the portfolio over time, we remain confident in our ability to do so. And again, as we've talked about, really, the profit pressure at the moment really is timing. And it's a mismatch between our ability to get rate filed and approved and earned relative to the loss experience we're seeing at this point in time, but we expect that's going to align with loss cost over time. And we've seen movies like this before, right, where we've written business, we've grown the business, we've removed the profit on that business we wrote over time. And so that's the path we're headed down. As we do start to put more price in the market, could that impact growth on a go-forward basis? It certainly could. But again, we're comfortable with our ability to profitably grow the auto book over time.

Do you think retention -- yes, go ahead, please.

Just to put a finer point on that. We had this experience a few years ago in Personal Lines, and we ended up with a larger book of very profitable business and we would expect this to play out in exactly the same way. So very pleased with the volume we're putting on.

Your next question is from the line of David Motemaden with Evercore ISI.

I had just a quick question on the Holdco cash balance. It fell versus last quarter, but still well above your target, I think, it was $800 million. Is that still just a function of the debt that you raised? I think it was last quarter. Have you -- you just haven't put that down into the operating companies for growth? Or is that sort of excess capital at the Holdco that we can think about to be used for M&A or share repurchases or other uses?

David, I'm not too terribly worried about cash at the Holdco at any quarters. And we're really looking at overall capital position and still feeling very strong about that. And I think you see that in the capital actions that we took this quarter, the decline from where we were, we also, because of the strength of the position that we started with, didn't need to bring up as much from the underlying companies. We're not signaling anything there with the variability in the holding company liquidity. We're really just trying to make sure that we've got what we need to cover. And if we've got a little more than that at any point in time, that's fine.

Got it. Okay. And then maybe a question just for Greg. I was a little surprised that the workers' comp net premium written is still down year-over-year. I know you said that the rate is still slightly negative. I guess, could you maybe talk about some of the other drivers of the decrease? I would think that there's some sort of wage inflation in there and the exposure base I would think is up versus last year. But maybe just talk about the puts and takes there? And when you think we should start to see some growth in workers' comp net premium written?

Yes. David, let me give you a little bit of color on what's underneath the quarter number. Clearly, as I said earlier, we still have a slight negative in rates. So that's going to have a large weight on the overall net written premium delta. We had strong retention. Overall, if there's an area that brought the number down was on a new business basis. And when we look into the marketplace, particularly larger accounts for this quarter, where we saw the clearing price relative to our price to risk to make sure that we're getting adequate value over the long term, we didn't hit on some larger deals. So really, it's a function of disciplined underwriting. So the combination of all those drove that minus (inaudible).

Your next question is from the line of Brian Meredith with UBS.

Just quickly on workers' comp. I'm just curious, are we getting to a point where maybe we start to see underlying loss ratio start to be stabilized or improve here given rate versus trend? Or are we still a ways away from that happening?

Yes. The loss trend has been pretty benign in workers' comp, but it hasn't been negative. It's been positive. And the rate -- the overall pricing has been sort of closer to 0, so below that. So if you look at a calendar year basis, almost being equal, you'd expect some compression. Of course, we've had a lot of prior year development. So the line has been on a -- it's been -- if you look back in accident years, more probably, we would have predicted the time. And we would have predicted some quarters ago that we would have hit a bottom and made a turn. But again, we've had pretty good results in that line throughout the pandemic. And so I think the outlook is probably going to continue to bounce around, call it, 0-ish for a little bit longer before it makes a turn, but it will. It will bounce around, but it will make a turn and go ahead from there.

Great. And then just quickly, I know you talked a little bit about cyber and all that's going on there with respect to tightening terms and conditions and rate, et cetera, et cetera. I think you said that you're declining in cyber right now. I would have thought this would be a really good time to maybe step on the accelerator a little bit and drive some more growth in cyber. So maybe give us a little bit more perspective on what's going on there?

Jeff, do you want to take that?

Yes. Thanks for the question, Brian. Cyber flow remains strong. It's consistent with broader demand in the marketplace, including first-time buyers. Like I mentioned earlier, in this elevated risk environment, our quoted ratios are understandably down given those dynamics. We've tightened underwriting requirements like multifactor authentication, like I mentioned. We're also aggressively pursuing needed rate on both new and renewal. So that's the dynamic that's still out there in the market in our results.

Got you. So are you just pushing harder than the rest of the market, is that what it is? Or are people just not buying as much?

I think it's just a combination of we're pushing consistent with our strategies and others are executing on their strategies, but it's pretty well known how the cyber dynamic has been in the marketplace.

Brian, I would also add 2 things. One, we've been very, very disciplined about enforcing minimum cyber hygiene. And I don't know that other markets are necessarily as disciplined about that. So when that business leaves us, it's presumably going somewhere. And given the profitability challenges in that line, we're just fine with that.

Your next question comes from the line of Meyer Shields with KBW.

Michael, can you update us on severity trends in personal/auto liability? We've talked a lot about the physical damage side. I just want to get a sense of what's going on the liability side?

Sure, Meyer. I would say, we talk a lot about property damage and liability because that's the place where it's really elevated relative to our expectations. I think on the personal lines side, there certainly is loss trend in bodily injury reliability, but it's much closer to what we would have expected.

Okay. Perfect. And a brief follow-up, if possible. I just was hoping I could tease or parse out that 2 points that Greg had talked about between the weather-related losses and the -- I'm sorry, and the COVID impact?

Meyer, it's Dan. At the risk of getting too granular, we'd rather not parse that out. You could think of them as being fairly equal in terms of their contribution.

And we have time for just one more question, and we'll take that question from Tracy Benguigui with Barclays.

Just really quickly on the commentary of that net favorable impact of COVID on business insurance. I'm looking at your 10-Q, and it looks like Travelers benefited from reduced claim settlement activity, largely due to the disruptions in the judicial system related to COVID-19. So how much of that thought process played into your underlying BI combined ratio improvement this quarter? And could you contextualize how that may change going forward?

Tracy, it's Dan. So really, what you're seeing in the current quarter is the impact on commercial auto and that's a frequency event. What -- so what I think what we expect there over time is, we get to a full post-pandemic recovery. If things go back to the way that we're prepandemic, I don't think we'd expect that frequency benefit to persist. I think the language -- if we could take it off-line, but I think the language you're reading out of the Q is more specific to the cash flows related to operations in the quarter and that would be lower claim payments.

If you look at paid to incurred, on a year-to-date basis, it's almost identical to where paid to incurred was on a year-to-date basis in the first 9 months of last year when we were talking about the fact that ports were closed and it was taking a long time to get settlements of the door.

Okay. That's helpful. Also, Greg mentioned that you focused on your deductibles, attachment point, supplements and exclusion. And that's also really important rather than just looking at a blend rate direction discussion. So could you just let me know your view of loss trend or rate adequacy that's driving those underwriting decisions and which business lines those actions are the most evident?

Tracy, I'm not exactly sure what you're trying to get at with that. I think the point we were trying to make is, we are always very disciplined underwriters starting with risk selection and then evaluating the terms of every single deal that we underwrite, paying attention to all the levers that contribute to price per unit of risk, and so I think that's the point we were trying to make. Those steps are important. It's not overpowering the results in any quarter, but those things compound over time. And to the extent that we are making those types of changes on an individual risk, it will improve the price per unit of risk. That's not apparent in the production statistics. And also, that tends to unwind more slowly than, say, rate unwinds over time, but I don't think we meant to make more of it than that.

Okay. So it sounds like just normal course of business.

And at this time, I'll turn the call back over to Ms. Goldstein for any closing remarks.

Thanks, everyone, for joining us today. We appreciate your time. And as always, if there's any follow-up, please feel free to reach out directly to Investor Relations. Have a good day.

This concludes today's conference call. Thank you for participating. You may now disconnect.