State Auto Financial Corporation (STFC) 2020 Q3 法說會逐字稿

Welcome, and thank you for standing by. (Operator Instructions) Today's call is being recorded. (Operator Instructions)

I would now like to turn the call over to State Auto Financial Corporation's Director of Investor Relations, Natalie Schoolcraft.

Thank you, Jumeriah. Good morning, everyone. Welcome to our Third Quarter 2020 Earnings Conference Call. Today, I'm joined by our Chairman, President and CEO, Mike LaRocco; Senior Vice President and CFO, Steve English; Senior Vice President of Personal and Commercial Lines and Managing Director of State Auto Labs, Kim Garland; Senior Vice President of Data and Analytics, Jason Berkey; Chief Actuarial Officer, Matt Mrozek; and Chief Investment Officer, Scott Jones.

After our prepared remarks, we'll open the lines for questions. Our comments today may include forward-looking statements, which by their nature, involve a number of risk factors and uncertainties, which may affect future financial performance. Such risk factors may cause actual results to differ materially from those contained in our projections or forward-looking statements. These types of factors are discussed at the end of our press release as well as in our annual and quarterly filings with the Securities and Exchange Commission.

Financial schedules containing reconciliations of certain non-GAAP measures, along with other supplemental financial information, are included as part of our press release. Additional material titled Monitoring Our Progress has been made available on our website, saleauto.com, and along with the press release can be found under the Investors section.

Now I will turn the call over to STFC's Chairman, President and CEO, Mike LaRocco.

Thanks, Nat, and good morning, everyone. First, I'd like to say, hope you and your families are safe and healthy as we continue to navigate our way through these difficult times. Well, the third quarter continued the story of 2020 for State Auto, cats. We had a third consecutive quarter of cat activity that exceeded our planned number and was significantly higher than the prior year. As usual, our claims team was terrific. Handling these claims with the speed and empathy our customers have come to expect from this outstanding team.

The impact of catastrophes and weather, in general, has obviously been the main story of 2020 regarding our results. However, the most important message of the quarter and the year is the ongoing improvement across all of State Auto. It begins with our underlying non-cat loss ratios, the key indicator when it comes to profit. Across nearly all our lines of business, we continued to make progress in this area. I'm very pleased with the pricing and position in the market of our products. Our product management teams have continued to analyze our results, review the competitive position and then adjust and improve our pricing models.

The level of sophistication in this work demonstrates the benefits of our talent and our digital platform. We've built an organization and technology that can deliver continuous improvement. Kim will cover the details of our non-cat loss ratios by product.

Another area where we see the value of our platform is our ability to efficiently add scale, which, of course, leads to improving expense ratios. In the midst of a pandemic, we continue to grow. Our agents have seen that our technology and our team has not missed a beat during these challenging times. This is a clear differentiator in the market, and our agents have responded. They know we can issue and service business regardless of the challenges. I'm very pleased with the growth we've achieved, and we still haven't hit our stride.

Let me speak specifically to auto. I'm thrilled with where we're positioned with our auto product, it's a one product where we had to make a number of adjustments and changes to our pricing model. In addition, as our first product on the digital platform, this might also have to deal with early bugs in the system. I take full accountability for the early missteps. Now we've overcome these issues and are in the market with a product and platform delivering in this critical line. Our current accident year results demonstrate improvement. Our mix of new business from preferred to nonstandard is meeting our expectation. More importantly, early frequency and severity numbers are behaving as planned.

I'd like to comment on COVID-19 and, specifically, business interruption coverage under commercial property policies. I'm pleased and not surprised, in the majority of cases, early rulings nationwide mostly had motions to dismiss, have found in favor of the industry's position that the pandemic does not create a physical loss of or damage to property and that insureds cannot extend their property coverage for pure economic losses.

We have confidence that this prevailing view will continue and will ensure that coverage will not be expanded beyond what was anticipated under the various policies.

For those very few early rulings that we have seen that do not support the industry's position, we believe these will be the exception and, in each case, we believe that we are able to differentiate our situations from those decisions on specific facts related to our approach and position.

Now before I conclude, I'd like to take a moment to share a milestone achievement and reflect just a bit. Last month, we launched our workers' comp product on our digital platform, Connect. This means that in 5 years, we were able to bring every one of our products onto this new platform: personal auto, home, dwelling fire, personal umbrella, commercial auto, BOP, CPP, commercial umbrella, farm and ranch, and now workers' compensation. We took an outdated company with outdated products and technologies and in 5 years have transformed it into a digital organization. Think about that for just a minute. Kim will cover the impact of what we have done in more detail. Now before he gets that, let me first pass this along to Steve. Steve?

Thanks, Mike. Good morning, everyone. For the quarter, STFC reported $0.26 net income per diluted share with an operating loss of $0.10 per diluted share. This compares to $0.25 net income and $0.34 operating income for the third quarter of 2019 on the same per-share basis. For the 9 months, STFC reported on a diluted per share basis net loss of $1.57, operating loss of $0.86, and this compares to $1.25 net income per diluted share and $0.34 operating income per diluted share for the first 9 months of 2019.

Cat losses, net investment gain and loss in COVID all impact these period over period comparisons and not necessarily in the same way, depending on the focus on whether looking at quarter or year-to-date results.

Let's start with cats. Catastrophe losses are a big part of 2020 results, with cat loss ratios increasing 13.3 points for the quarter and 10.5 points year-to-date as compared to 2019. Third quarter 2020 was impacted by 2 significant storms, one from 2017 and one that occurred in the quarter.

We have spoken of the impact on Irma loss estimates from the legal environment in the State of Florida and the involvement of public adjusters. Recent court decisions have caused us to reevaluate, and we made a decision to increase our estimated ultimate losses by recording additional IBNR. We are now fully reserved up to the retention of our property cat reinsurance treaty that was in place at the time of the storm and do not expect any significant further adverse development as a consequence.

The second significant event of the quarter was the Derecho that primarily impacted the State of Iowa. Our loss estimate for that storm impacted our cat loss ratio by 4.6 points. Storm losses, both cat and non-cat, have been elevated every quarter of 2020, including the severe tornado that hit the Nashville area this past March.

Fluctuation of reported net investment gain or loss, which included unrealized gains and losses on equity securities and other invested assets, has also impacted quarterly and year-to-date comparisons. One item of note is we completed the exit of MLPs that began in late first quarter. You might have noticed $42.3 million of net realized losses on sale of equity securities in the quarter, bringing the 9-month total to $50.9 million. This reflects the exit of these securities. However, the impact to 2020 results is less, has included an unrealized gain or loss per equity securities, are the reversals of previously reported net unrealized losses on the same securities, offsetting the net realized losses on sales with reversal of previously reported net unrealized losses.

The reduction to income or loss before federal income taxes for the sale of these MLPs was $2.8 million in the quarter, and for the 9 months, $35.1 million. The reduction in net investment income in the quarter and year-to-date are also impacted by the decision to exit this asset class. While carrying a high dividend yield, the total returns were not acceptable.

COVID is the third item noted. Kim Garland will comment further on this in his remarks. The COVID continues to favorably impact claim frequency in the automobile lines as well as small and middle commercial lines. We made no changes to our loss adjustment expense estimates for expected defense costs related to business interruption cases, and we have not incurred any indemnity losses to date.

Moving to reserve development, I mentioned the unfavorable development in Irma, so I won't repeat that discussion. For non-cat loss development, we continue to see overall favorable development of prior-year loss reserves. Personal auto is an exception, where bodily injury severity has exceeded prior estimates, primarily from the 2018 and 2019 accident years. Commercial lines development continues to be favorable across all product lines and, in particular, small and middle commercial, along with workers' compensation. The GAAP expense ratio was essentially flat with 2019 after 9 months. No significant adjustments were recorded to the bad debt allowance directly as a result of COVID beyond what we booked in the first quarter of 2020.

There were some note payable changes in the quarter. In early September, we retired STFC's 5-year note with the Federal Home Loan Bank, replacing it with a new 10-year note in the same amount of $21.5 million. The new interest rate is fixed at 1.37% versus the expiring rate of 1.73%. In addition, the $60 million Federal Home Loan Bank short-term loan entered into on March 19 matured and was repaid in full on September 22.

And with that I will turn the call over to Kim.

Thanks, Steve, and good morning, everyone. On October 29, workers' comp Connect launched in Texas. This is our last product line to launch on the Connect platform. We are thrilled, and it's probably appropriate to look back at the impact that the Connect program has driven at State Auto since 2015.

From the first day in 2015, the goals of the Connect program were the following: increase scale, reduce expense ratio, reduce our non-cat loss and ALAE ratio, replace and upgrade our technology to give us a competitive advantage, replace and upgrade our products and pricing models to compete with the best prices in the industry, of all of our traditional retail independent agent distribution plan to expand to additional types of growing independent distributors that were emerging, driven by the significant innovation taking place in the world of independent agents and significantly participate in the world of InsurTech innovation at a moment in time when the insurance industry was entering a period of massive innovation and disruption.

Our progress on scale. Direct new business written premium for the State Auto Group in the year 2015 for personal and commercial lines was $226 million. For 2019, it was $557.9 million, which equates to a 25.3% compound annual growth rate. For the 12 months ending September 30, 2020, it was $661.3 million. Connect has helped drive a 3x increase in our new business production since 2015, and this is without a full year of impact from farm and ranch, middle market and workers' compensation Connect. This has driven our total written premium for personal lines and commercial lines to grow at double-digit rates over the last 3 years. This increase in the size of our new business engine has positioned us well for continued growth into the future.

Progress on reducing expense ratio. Our strategy was to take a short-term expense ratio hit through the investment in Connect in exchange for a longer-term expense ratio improvement through a more efficient platform and an increase in scale. Four of our product lines, personal auto, homeowners, commercial auto and small commercial, are far enough along to understand if this strategy is working or not. It has worked in 3 of our 4 product lines.

Commercial auto is the best example of this with a 2 -- 2016 pre-Connect expense ratio of 38.5%. This increased to 46.5% in 2018, as we made our investment in Connect. But the year-to-date 2020 expense ratio is at 34.3%, a 4.2 point improvement than -- over pre-Connect.

Over the next 2 years, we believe the commercial auto expense ratio will get below 30%. This expense ratio reduction strategy has also worked in small commercial whose pattern is very similar to commercial auto and homeowners, where this pattern has been slower and more gradual.

Personal auto is the product line where the strategy has not yet been effective, as the expense ratio has actually increased after the implementation of Connect. As I have told you on prior calls, we made missteps in personal auto with the primary one being our balance between preferred and nonstandard business skewed too heavily towards nonstandard, which has a very low retention. We are putting inefficient business on an efficient platform. This combination still produces overall inefficiency.

The increase in the personal auto expense ratio is not from a bad strategy but from poor execution of a good strategy. I will update you on our progress in personal auto later in this discussion.

With the Connect product builds being completed, you should start to see 2 things: one, the same pattern of expense ratio reduction to show up in the remaining product lines over the next couple of years; and two, since there are no more new Connect builds to invest in, improvements in all the product lines should result in State Auto's total expense ratio starting to decline.

Progress on our non-cat loss and ALAE ratios. The non-cat loss and ALAE ratios have been driven from the 57%, 58% range in 2015 down to the 49% to 51% percent range currently. Non-cat loss and ALAE ratios in the 50 range with declining expense ratios should be sufficient to produce mid '90s combined ratio consistently over time.

Progress on evolving our distribution plans. Insurers categorize agents in a variety of ways. We categorize them in the following way: retail agents, network agents, platform agents and corporate agents. In 2015, we were heavily weighted in the world of retail agents who we love, but we are underweighted in the network agent world, and we did zero business with platform agents, 2 [IA] categories that were rapidly innovating and growing.

Over the last 5 years, we have seen a significant shift in the distribution of our business by agency category. Personal lines is a good example of this phenomenon. In 2015, our personal lines' new business was distributed in the following proportions: retail 63%, network 30%, platform 0%, corporate agent 7%. Through 9 months of 2020, our personal lines' new business is distributed in the following proportions: retail 35%, networks 54%, platforms 8% and corporate agents 3%. This major change in how our business is distributed is a sign that we are now effectively competing in all the methods of distribution in the independent agency channel.

Progress on our technology products and pricing models or what we call the Connect platform. The historical model for insurance technology, products and pricing models, was that a carrier would do a big build and then wait an extended period of time, 2 to 5 years, to do another big build. Our technology product design and pricing models for every single product line have converted from the periodic big build model to a continuous improvement model. For pricing models, you hear us talk about versions: personal auto version 2.1, homeowners version 3.0, commercial auto version 2.0, small commercial version 2.2. In this approach our pricing models are continually improving. Every quarter, enhancements are made, and this has a very different impact on the approach where one's pricing models improve only once every couple of years.

And for our technology, we are converting from the big build teams for Connect to small product teams that are continually enhancing different parts of Connect. In 2015, we started from a place where we have fallen behind in the industry in a number of areas. Our conversion to this operational model of continuous improvement is to help ensure that this never happens again, that we never fall behind and have to dig ourselves out of a hole.

Progress on the participation in the world of InsurTech and being prepared for this era of innovation and disruption in the insurance industry. In 2015, State Auto did not participate in the world of InsurTech. State Auto Labs, State Auto's corporate venture capital fund, was created in 2016, and since that time, 24 InsurTech start-ups have been implemented in the State Auto's insurance operations and 6 investments have been made. Without Connect, we would not have been able to integrate this number of InsurTechs into our insurance operations. It would have been impossible on our legacy system.

In the first quarter of 2020, Coverager, a new source and researcher of the InsurTech world, surveyed the participants of the InsurTech world and released their list of the top 11 Insurance corporate venture capital funds and InsurTech partners in the industry. State Auto Labs was on that list of the top 11. We believe that the insurance industry is still at the beginning of a period of massive innovation and disruption, and it is critical to be connected into the InsurTech world.

As we reflect on the impact of Connect and all of the efforts surrounding Connect, we believe it has accomplished all of the goals that we established back in 2015 when we started and that has positioned us well to be an extremely successful insurer over the next decade. It is impossible for us to adequately thank all of the State Auto associates for what they did to implement the Connect program. We have posted to the Investor section of our website in greater detail the data I have referenced this morning.

Now on the third quarter results. Overall personal lines and commercial lines statutory results are the following: third quarter '20 combined ratio was 102.2 compared to 98.7 for third quarter '19. Third Quarter '20 year-to-date combined ratio was 106.9 compared to the 102.7 for 3Q '19 year-to-date. Written premium growth was 9.2% third quarter '20 versus third quarter '19; 11.4% year-to-date 2020 compared to 2019 year-to-date. For the quarter, commercial lines had a combined ratio of 88.9 and written premium growth of 8.3%. And personal lines had a combined ratio of 111.5 and written premium growth of 9.7%.

To appropriately understand our results, there are quite a few things to unwind.

Our quarterly results. Catastrophes are again the story of the quarter.

The catastrophe impact on our third quarter '20 results were the following: personal lines and commercial lines cat loss ratios were 9.9 points higher than 3Q '19, where personal lines were 11.6 points higher, commercial lines 7.6 points higher. Non-cat loss and ALAE ratios are out of our plans of 3Q '20. The personal lines and commercial lines non-cat loss ratio was 6.3 points lower than 3Q '19, with personal lines being 2.3 points lower and commercial lines being 12.2 points lower.

Non-cat loss and ALAE prior accident year loss ratio impact. The peak personal lines and commercial lines non-cat loss ratio accident year impact was 0.4 points lower than third quarter '19. The personal lines non-cat loss ratio prior year accident year impact was 5 points higher than 3Q '19, while the commercial lines non-cat loss ratio prior accident year impact of -- was 7.9 points lower than 3Q '19. The overall impact of prior accident years was roughly the same as last year, but there was a meaningful shift between personal lines and commercial lines.

Non-cat loss and ALAE current accident year loss ratios. The personal lines and commercial lines non-cat current accident year loss ratios were 5.9 points lower than 3Q '19, with personal lines being 7.3 points lower and commercial lines being 4.3 points lower. This improvement in the current accident year is driven by 2 factors: COVID and the actions that we have taken in personal lines to address the issues that we discussed on the first quarter earnings call.

We feel very good about where we are positioned. Catastrophes happen, and 2020 is a horrible catastrophe year. But when we look back at the total over the past 6 years, our actual cat loss ratio is virtually identical to our planned cat loss ratio. Catastrophes have to be managed over an extended period of time, understanding there will be individual horrible catastrophe years along the way. We feel good about our catastrophe management.

The adverse prior year reserve development in personal auto is part of the price we are paying for the personal auto missteps we made in 2018 and 2019. The improvement in current accident year loss ratios get them to where they need to be to consistently produce mid-'90s combined ratios. And as I explained earlier, being at the end of the Connect program means that expense ratio improvement should start to flow into our total results.

Some updates on our individual product lines. Personal auto. The personal auto combined ratio for third quarter '20 was 104.7. This was driven by 7.8 loss ratio points from prior accident years. This is 9 points higher -- 9-point higher impact from prior accident years than in third quarter '19. And as I said before, this adverse prior year reserve development in personal auto is part of the price we are paying for personal auto missteps we made in 2018 and 2019.

The third quarter '20 current accident year loss and ALAE ratio of 56.4 is 11 points better than a year ago and third quarter '20 year-to-date is 11.3 points better than a year ago.

Our shift to less nonstandard and more preferred is progressing as expected. Connect version 2.1 has been approved in 20 states, and we are seeing the changes in competitiveness across the preferred to nonstandard risk spectrum that we expected.

Connect ultra-preferred and preferred policies in force growth is 42% as of 9/30/20 versus prior year, while Connect nonstandard PIF growth is minus 13% as of 9/30/20 versus the prior year. As of 9/30/20, for the first time in Connect's history, ultra-preferred and preferred Connect PIF is larger than nonstandard PIF. This shift is driving an increase in personal auto retention as personal auto retention is up 1.7 points to 68.4 since December 31, '19.

In another side of this shift, personal umbrella new business sales were up 123% in third quarter '20 versus third quarter '19. We are pleased with our progress through 2020 in personal auto, but there is still work to do in this product line.

Homeowners. Catastrophes are again the story of the quarter for homeowners with a cat loss and ALAE ratio of 30.4 compared to 9.2 in third quarter '19. Growth continues to be strong in homeowners with written premium growth rate of plus 21.7%, policies in force growth rate of plus 16.1%, new business count growth rate of 16.2% and retention level of 76.7, which is up 1.2 points since December 31, 2019. The improved personal auto rate competitiveness from personal auto Connect version 2.1 is helping homeowners retention.

Commercial auto. Our commercial auto results continue to be just terrific. With third quarter '20 results of the combined ratio of 94.3, total written premium growth of 44.4, new business premium growth rate of 46.7 and a premium retention level of 84.1.

Middle market, small commercial and workers' compensation results in third quarter '20 were all impacted by prior accident year loss ratios that were much more favorable than a year ago, while middle market and small commercial results in third quarter '20 were also impacted by a much higher cat loss ratios than a year ago. The workers' comp third quarter '20 current accident year loss ratio of 82.2 was 18 points higher than a year ago, driven by 5 large losses in the quarter.

Middle market growth. We are seeing a slowdown in middle market submissions. Agents are telling us that middle market policyholders are more hesitant to move carriers at this moment in time. Our third quarter '20 middle market new business written premium declined 39.8% from a year ago, and our total middle market new business written premium was flat from a year ago. We rolled out 8 additional states for middle market Connect in September, bringing the total states on middle market Connect to 16.

Workers' compensation growth. Our workers' compensation business has shrunk for several years due to 4 reasons: one, we exited monoline small workers' compensation business sold through wholesalers, which has consistently been unprofitable; two, we have maintained rate discipline while the market has been significantly decreasing rates over the last several years, and this put significant downward pressure on retention; three, we are currently non-renewing our nursing home workers' comp policies; and fourth and the last reason is that our workers' compensation business was sold on a separate system from the rest of State Auto's products, making it significantly harder to package workers' comp with the rest of our commercial products.

This changed on October 29 with the launch of our first workers' comp Connect state, Texas. Workers' compensation will now be on the same platform as the rest of our products, and we anticipate this significantly accelerating our growth of our workers' compensation business reversing several years of premium decline. The last few years of work have now positioned our workers' comp business for significant future success.

Farm and ranch results continue to be terrific. Farm and ranch Connect is now launched in 27 states. Third quarter 20 results versus a year ago are the following: new business written premium growth rate of 202%; total written premium growth rate of 31.6%; a combined ratio of 101.4%, which includes a catastrophe loss ratio of 22.1, which is 18.9 points higher than the third quarter '19, and elevated expense ratio of 45.1%, as farm and ranch is at the height of its Connect investment; and a non-cat loss and ALAE ratio of 30.8%, which is 13.5 points lower than the third quarter '19. Our farm and ranch business is positioned incredibly well for future profitable growth. We are excited about how our product lineup is currently positioned for future success. And with that, we'll open the line for questions.

(Operator Instructions) Your first question will come from Paul Newsome.

Hopefully, you are all safe and well. I wanted to ask about the underlying combined ratio improvement in the year, obviously, been positive. But how should we think or do you have any thoughts about how we should think about the windfall that we've had because of the pandemic versus sort of the "real underlying improvement" to give us a better sense of where you are as we look past, not done with the pandemic?

Kim, you want to start off? I may jump in a little bit.

Sure. So I think how I would think about the COVID benefits is, I think second quarter '20 results were probably the vast majority of the benefit was from COVID. And we had just started the process making improvements in personal auto. I think the other thing that is worth mentioning from a COVID-type world and after it is, we and others in the industry are in the process of pumping a lot of rate into commercial property. So that's a -- and not a lot of that rate had gotten into the system yet. So I'd say second quarter benefit is mostly COVID.

In third quarter, what we see is, miles driven has gone up, but it's not gone up to pre-COVID levels and the miles driven are different in some ways. You don't have really rush hours like you had pre-COVID. So there is still COVID benefit in the third quarter. But the proportions are now -- that part is shrinking and the larger part is, at least for personal auto, the actions that we have been taking and more of the rate is starting to get into commercial property. It's still early. We still have a lot of rate that needs to earn in for commercial property. But I think in each quarter going forward, you will see as the COVID "benefit" burns off, the sort of -- that percentage of impact on our results will go down. But I think we are filling it in with some of the actions that we started back in the first quarter of 2020.

Yes, Paul. Just to tag on a little bit just a couple of areas, I completely agree with Kim. I think that directionally, we're kind of where we believe we can be. There is no question that we've taken some real benefit, and I'm always hate to talk about benefit from something as tragic as COVID, but the impact from a financial standpoint, there's been benefits. But in addition to the rate Kim talked about some of the other changes from a modeling standpoint, we've made some significant improvements in reducing things like loss leakage in areas in our underwriting organization that I believe have been significant.

Our Claims and Risk Engineering organization has had a terrific impact during this period of time. They've been able to implement some real digital approaches to handling claims. They've been able to reduce some of the pending across the spectrum of the many of the older claims.

And I think when you take into consideration the rate, the modeling changes, the underwriting improvements and then the Claims and Risk Engineering organization, we're emerging from this period of time I think in a pretty good place.

The other thing we all have to be cognizant of is people -- in the early days of COVID, we were talking about, as Kim said, miles driven and shelter-in-place impacts, but obviously the impact of COVID is going to extend minimally through the first quarter of next year. Now it may trend a little bit differently, but I think the reality of the COVID impact on losses will linger for minimally another 6 months. So that's all I think I would add to what Kim said.

That's great. And then my -- just can you give us a sense of the competitive environment, particularly in personal lines. Obviously, lots of talk of rate going at least down for some of the companies that have reported so far and lots of InsurTech stuff, which I don't think is a big deal. But just give us a sense, at least from your markets what do you think is going on from a competitive perspective?

Go ahead, Kim. I'll again follow you.

Yes. Okay. So I think in personal lines, we have not seen sort of large changes in a couple of the competitive dynamics. So for us, one of the things -- there are a number of things to look at. I think one is like your quote volumes and your new business levels. In homeowners, we continue to have all-time highs in new business sales. Rates are starting to increase in more states. So it's becoming broader. So quote volumes, the new business sales and the work we do in trying to understand the competitiveness of our rates, we've not seen big changes in that dynamic. So homeowners is kind of minimal.

On the personal auto side, we've seen drops in quotes and new business, but we -- our belief is that it's primarily from our actions, not the market actions, just because we instituted nonstandard filters. So we very cautiously sort of reduced the quote volume. And when we try and sort of take out that impact, personal auto quotes seem to be sort of not impacted that much. When we look at things and we look at sort of -- if I just look at preferred, we look at preferred quote volumes and new business are up for us. And when we look at our competitiveness in preferred, it's actually improved over the last 6 months. Some people are making moves to reduce rates, but we implemented the Connect 2.1 structure, which sort of tilts more towards being more competitive in preferred. So that's more than offset what others might have done in the marketplace.

So for us, we haven't seen a whole bunch of changes in personal lines and competitiveness.

Yes. I think, Paul, just again to tag on to all that, I think Kim is right, and I think there's a lot of wait-and-see feeling just given the uncertainty around COVID. But the other thing I think you have to think about here and be aware of is that there are a lot of companies that have struggled through this. And so our growth across homeowners and now the trending improvement that we're starting to see in new business and personal lines with regard to more preferred business, and I think directionally, that will pick up some momentum here in the fourth quarter and moving forward. We are in the market, and we, again, have not missed a beat. And this isn't just me blowing smoke. I mean this is just the reality of State Auto in this digital platform. And that means a lot to agents.

The second thing I'll say is that there are certain types of agents that continue to get a disproportionate amount of business. And of course, we are in those spaces. Kim talks about that when he talked about our shift in distribution, the platform agents that we use who really heavily leverage a digital approach to sales, whether it's online or through a call center. They are still getting quite a bit of quote activity, and they will only partner with others who can efficiently handle that business.

So I think the impact of companies that are struggling through this, who aren't as accessible as a market, that's something else when you talk about the marketplace competitiveness and personal insurance. The second thing I would suggest is that there's a -- there's a growing difference in certain agencies who are driving an increasing amount of business. And I think in both of those counts, it directly impacts the competitiveness of certain companies. It's not just a rate issue, but it's an availability and an efficiency issue. And on those 2 counts, I think we're -- is one of the reasons we're continuing to see the really positive results in homeowners, and we believe we'll add auto to that as we rebuild. So I would just want to tag that on.

Your next question will come from Marla Backer.

So since the industry is seeing severe weather this quarter, last quarter, and it seems like, in the fourth quarter potentially as well, can you talk a little bit about what that means, if it means anything for your reinsurance strategy?

Steve, do you want to talk about where we're at right now and your thoughts?

Sure. We've touched on this previously. Our strategy on reinsurance, especially when it comes to a property cat is more from a capital preservation or protection angle versus a current earnings management philosophy. So we realize that cats go through -- seem to go through cycles. We can't really predict the weather, but -- and so our products are priced in the property lines to take that into account, making profit over the cycles knowing in any individual quarter or, in some cases, years you might have poor results.

So we're not planning any changes presently. We just renewed our treaties this past July 1. We did increase our retention on a group level when we do the analysis and look at what the property reinsurers are wanting to charge for layers below our retention relative to the premium and the probabilities of hitting those layers. It just didn't make sense to purchase that. So instead we deployed more money on top, again, from a capital protection perspective.

Yes, I'll just again tag on briefly. I mean it's a terrific question and, look, global warming is for real, and there's been more weather activity, and we all have to be super-cognizant of that and aware of that. I really do believe that beyond those concerns, this has just been a significantly bad year. If you think about the impact on us, this wasn't as you would normally expect, which is just pure wind and convective storms in -- particularly, in certain areas. In the second -- in the first quarter, we had a very unusual tornado set down in Nashville that hit us. In the second quarter, you had a handful of wind and convective storm-type issues. The hurricane activity wasn't particularly bad for us, but it's spun off some other weather. And then if you really look at -- and then of course we had the Derecho in the third quarter.

So it was a very unique set of circumstances. Having said that, again, you make a really important point, and we're aware of that. I think what we've done in terms of risk differentiation by geography in which we continue to do is another way we think about this beyond the good point Steve made with regard to how we purchase our reinsurance and how we use it from a capital standpoint. So I appreciate the question. Thanks.

So then there's so much noise going on right now with COVID and the impact of people being on the roads less often. So can you try to -- or maybe it's not possible, can you try to parse through some of this noise to talk about what you see in terms of like underlying more normalized trends as it relates to the telematics initiative because we've talked -- I know you've talked about that in prior calls about encouraging policyholders in the use of telematics. So can you address that?

Yes. At this time, I'll actually start and then turn it over to Kim because I was going to mention telematics in my answer and I failed to Paul because that's another competitive issue. When you think about the competitive positioning, companies that leverage telematics not just have it but leverage it effectively, which I believe we do -- are going to have a competitive advantage.

And so I will turn the main part of the question over to Kim, but I think again you hit a really important issue because digital in general and no one knows exactly what the new normal is going to be coming out of COVID but it's pretty clear that some leverage of the digital interaction with whoever your supplier of whatever the product is, it's just going to exponentially grow.

And quite frankly, I kind of put telematics into that space because it allows people in a very efficient way in both commercial and personal to pay as they drive. And I think as people have learned through COVID, the reality of changing driving patterns we know coming out of this, more employees are going to work from home than ever before. That's going to impact miles driven and time of day when they actually drive. So telematics will be a part of the impact coming out of COVID. Kim, do you want to elaborate on that a little bit?

Yes, I think the question of like normalizing trends and trying to understand things with all these moving pieces is a question we ask ourselves and think about a lot. I think the first thing for telematics specifically that we keep an eye on is almost from day one on both personal auto, commercial auto, we give discounts for telematics, but even with that discount, we see loss ratios better for our telematics business than our non-telematics business. That pattern continues to exist. So for us, it's just how do we drive more adoption.

Outside of telematics, how we think about it, and I'm sure it's not that differently than others, is we keep track of miles driven and then what our collision frequencies are. And so collision frequencies sort of have gone down more than miles driven have gone down and so that is a way to get some sense of what a COVID impact might be to that and we watch that every month. And claim counts, there's lots of them, so it's a relatively stable measure. So that's kind of a known to us.

And then the other side that we are keeping an eye on is there are some signs that some of the accidents are more -- can be more severe, but severity bounce around and so it's a bit fuzzy. Severity up more than normal, if so how much. So we keep an eye on those sorts of things, but telematics and then looking at frequency, collision frequency specifically for -- versus miles driven to estimate sort of the number of accidents, COVID benefit and then trying to overlay, is there any severity dynamic on top of that is how we try and normalize all this.

Your next question will come from Meyer Shields.

A couple of small-ball questions. First, workers' compensation, I guess, the underlying accident year loss ratio has gone up over the course of 2020. Is there a true-up in the third quarter? Or is it some of the trend, in fact, in the line?

Kim?

Yes. No, I think there were -- as I said, there were like 5 large losses that we had in the quarter. And so -- well, I think there are 2 things going on. So if you look at the current accident year, that is driven by 5 large losses. 4 of the large losses were non-COVID, so that just sort of is one of those quarters. You take that out, it looks pretty consistent, pretty good.

I think Steve mentioned this. I think the prior accident year was quite a negative loss ratio impact on the quarter, and so I think we had favorable reserve development in prior years. So those are the moving pieces, but both of those are sort of -- we do not think that there's a signal that things have inherently changed in workers' comp or workers' comp business.

Okay. I guess related follow-up question. There is a fair amount of dispute, I guess, out there in terms of whether workers' compensation pricing is bottoming, is inflecting. What are you seeing? What are you planning for?

I think it's -- I think there are debates or how we debated internally is, it's gone down for several years and, at some point, you figure things just have to stop going down, but in a non-COVID -- there is less activity in offices and restaurants and whatever, so that puts continued downward pressure in doing that. Also, I think in some of the states, they're talking about their rate making. They're going to exclude COVID losses in sort of the go-forward pricing. And so if they do that, that will continue to put downward pressure on rates. I think for us, we are probably flat, marginally down, so we're not looking to take big rate increases. If anything internal to us, you're going to see us sort of very maybe more by state than we normally did in industry type.

The industry type, those that have greater COVID exposure were probably going to maybe go up a little bit, not down. And then for the states, it all becomes trying to keep an eye on sort of the regulatory system, are they going to keep the rules the same, are they going to change the rules that -- to allow more losses into the system? And so for states that do that, we will react to that and probably increase rates versus decrease them.

The only thing I'd add, Meyer, to what Kim said is, we're going to continue to do what we've always done. One of the reasons we've seen a little bit of a decrease in our workers' comp is because profit comes first, and our discipline around that won't change. So if indeed there is a continued downward pressure, which I think at least in the short term there will be, you won't see us going down that rat hole. The other thing, too, with our announcement that workers' comp is now on the Connect platform, this is going to increase our sales of workers' comp as part of the mid-market or small commercial package, which I think will also be a very positive outcome on the workers' comp side.

Right. Yes, no, I definitely think about that with regard to -- I don't want to overstate this, but some ability and desire to grow in what's been a competitive environment. Obviously, the margins are phenomenal. So that's not a problem. I just wanted to read your mindset. And then I'm going to call this philosophical. That's too grandiose. But question on BI. Maybe it's a broader reserving question. How do you incorporate the risk of absurd court decisions that are still going to be enforced?

Steve, do you want to start? And I'll follow up.

If I understand the question, you're talking about how some of these court rulings that are coming out that have not been favorable to the industry. How are we considering those in our reserving? Is that the question?

Yes, the prospects that you will get 100% of the decisions in a way that makes a whole lot of sense.

Yes. So I mean, our philosophy at the moment is, we've been presented with X number of claims. We've adjusted each one of them individually very thoroughly. As I mentioned in my comments, we've yet to book any kind of indemnity loss associated with this. Until, I guess, it hits us home, in other words, a particular suit that we might be involved in or something that emerges that -- as we evaluate those particular cases that we think, gee, that is -- well, let me respond in a different way. To date, as we've evaluated some of those cases and looked at our particular facts and circumstances, they don't align or line up. So we believe there are differentiating factors. And so there has not anything happened to date in our minds that has triggered any need for us to record any sort of indemnity reserve. So we're watching those very closely, but to date, our position and conclusions have been still no need to book a reserve.

(Operator Instructions) Your next question will come from Ron Bobman.

Hope everyone's well and safe. I'm sorry I missed the early part of the call, and I can circle back and get the details on the Irma matter. But I was wondering -- I assume you haven't disclosed this at least in the early part. How many open Irma claims do you still have?

This is Steve English. We haven't disclosed how many open claims we still have. Just as a reminder, that came out of our -- solely out of our specialty book, where that particular book, quite frankly, was in the business of writing wind-exposed Florida coastal. But in my earlier comments, we've now reserved up to our retention under our property cat treaty, and so we don't really expect any kind of meaningful further adverse development on our books. I will tell you that as we analyze the population of claims, that's still not closed. We don't feel at all that, that would go anywhere near to top of our program. So we feel like the balance sheet is pretty protected now from Irma.

Following a little bit on the recent question from Meyer and your comments about COVID, I'm wondering whether you're seeing any new business opportunities from agents who are moving business or biased to move business from carriers who are maybe a little bit more COVID vulnerable and you're being, in essence, the beneficiary of that as agents look to maybe diversify their books at all.

Kim, do you want to comment before I do?

Sure. I think we've seen probably 2 phenomena. One is, this period of time, where people had to reinvent how they are doing business and so, like, having to have sales people come in the office, having to sell business face to face, agents who sort of that was how they did their business and the only way they wanted to do their business. I think they more fully appreciate at least having carriers or doing business with carriers who can do it in other ways than that. So we probably have seen some agents who were maybe more skeptical about our sort of digital approach. I'll say flip and buy-in and say, "Hey, that was good foresight you had, sort of, count me in." And so I think we are probably the beneficiaries of agents diversifying if they had only non-digital carriers. And then this -- there is this phenomena at least we are seeing in middle market, where middle market policy holders are really kind of hunkering down and staying where they are.

So we have a number of phenomenas where, like, we'll get late into a quote, we're competitive, we think we'll get it, and at the last -- but at the last moment before the policyholder pulls the trigger, they say, "I'm just going to stay here one more year and see how it does." Also in middle market, that is more agents go out and sort of hump the business, and it's more face to face and it's more in depth. So I think some of their pipelines have shrunk through this. And so there is opportunity. I think they're either like when this ends and everybody sort of says, I want to go or I want to start moving my policies again. Or if the whole sort of market figures out a non-in-person weighted to do middle market, those are the opportunities there. But that's what we see.

Yes, I'd just like to add. I mean, while it's hard to obviously quantify it with a specific number, I really like the question because I really -- I talk to a lot of agents, and the feedback I've gotten has been very favorable. And we talk a lot about the home growth with the digital platform. We talk about auto now starting to come back.

We also -- we didn't talk a lot about our BOP new business. And again, the reason I mentioned those 3 products is because the digital platform in those 3 is pretty much is kind of quote and issue business. And in each of those cases, including BOP, the number of quote opportunities in our growing new business, there is a direct reflection there of the fact that agents have to pick somebody, obviously, who still open and somebody who they don't have to pick up the phone necessarily and call that they could do everything online in a digital manner, including reaching out to their customers remotely. So again, I wish I could put a more specific number on it, but I think our new business numbers are a pretty direct reflection of the benefit we're getting with our platform during COVID.

I just -- yes, Mike made a great point. We're -- in both September and October, we had record new business months in commercial auto in BOP. And if you just look at sort of the level of business activity out there, that's kind of crazy. So it would be impossible to say that we're not getting opportunity there that others aren't.

Okay. For some time now, I guess, really years, you've been laser-focused on your systems and rolling out your systems and presumably writing new business on those systems, but I know oftentimes, new business goes on to the systems, but renewal does not. And I'm wondering in State Auto's execution, is that the case? Or whenever you've been sort of rolling out these new systems, you've highlighted the workers' comp one, I think, in Texas, for example, most recently and I think the last, what's the split? Whether it's the entire book on the new systems, all lines, or is it just new business?

Yes. To be clear, when we write a new policy on Connect, it also renews on Connect. So we have the new customers now depending on auto. It's been over 4 years, almost 5 years. In each -- as each of the products roll out the new business, and as it renews, it stays on Connect. We still have what we call our legacy business, which is the business that was written before Connect, and that business has -- generally speaking, has remained on the legacy systems. The split between Connect and legacy, Kim, do you want to talk about personal lines where it is probably most significant?

Yes. And we'll have to, like, double-check my numbers, and if I'm way off, get back to you. But I think for both personal, auto and homeowners, about -- if you look at our total premium, I think it is maybe 60-ish percent on Connect, 40% on legacy. And so those products have been out there the most. If you go to workers' comp, we've been in the market for a week. It's probably 99.8% on legacy and very small on Connect. So there is -- I think commercial auto is probably starting to get up into the 20%, 30% on Connect and then BOP is the next down as a percentage. But you raise a good point.

If you think about the path to a lower expense ratio, we've now invested in all the Connect products. So there's no more like new products that we have to invest in. So we're going to sort of reap the benefit of all these savings over time and without having to sort of reinvest in building a new system. But your point is, at some point, we convert the legacy on to Connect, and that's going to be another tranche of expense ratio improvement. But I think Mike has mentioned this before. It's a different way of doing business, a different agreement with the customer. And so that is why we have moved relatively slow in the personal lines world as customers came to us under doing one way, and we did not want to quickly push them to a different way. At some point, the legacy business gets small enough, you need to convert folks over. But it's another sort of arrow in our quiver to drive the expense ratio lower.

When you said a different agreement, you mean policy form or do you mean agency agreement?

Just informal agreement. Like, when you bought a policy from us, you could write a check and you can pay cash and you got printouts and all that stuff. And so we try to be very sensitive to sort of when a 12-year customer came to us, the part of that they came to us was the agent and then how we did business, and we kind of wanted to honor that and go slow. So there is no formal agreement, but there is sort of how we thought about or how we talk about it internally.

Okay. I got one more question. The company and the group, I think, have an A minus Best rating, and results have been disappointing, candidly. Underwriting results have been -- taken longer to get to the underwriting performance than everyone has desired. And I think it's sort of an acknowledged fact over the years. And having -- but recognizing, you've made -- you have made a lot of progress operationally. It's not fully evident in the printed results, partly because of weather and partly because it has taken longer. And it's been harder than expected.

The current backdrop is, the stock price is really, really cheap on a relative to book value basis. And I'm wondering in the context of your A minus rating, and I'm not sure where sort of the mid-year estimated RBC is, but do you have any other levers to pull from -- with respect to sort of capital allocation, capital return to capitalize upon the cheaper price or because of the capital situation, i.e., the A minus rating where you stand in best eyes, the amount of capital at the mutual, you just don't have a lot of other levers to pull?

This is Steve English. I think that was a well thought-out question about whether or not we're contemplating buying back any treasury stock, and the answer to that would be, no, we would -- for many reasons, but most particularly, we believe that we're going to need that capital to support our business plans and our strategies and believe in the long run, we'll deliver a better return for folks in that way.

Okay. It was broader than that. But do you have a limit -- I mean is it best? Or your RBC, is that a constraining factor on how much business you can write? If you want to explain it that way.

No, no. The -- our RBC ratios, for example, are well above any type of level that you start to get regulatory review or scrutiny. And from an A.M. Best perspective, of course, our rating cycle with them -- we typically talk to them in the spring. They make their annual decision typically in June. And so we just went through that process with them, presented them our plans, which is a growth plan. They affirmed our rating. Obviously what we've seen this year with the cats, I think A.M. Best has been around the industry quite a while. So they are not unfamiliar with the periodic elevated cats that can occur when you write property, but it is presently not a constraint. No.

Is A minus satisfactory from a competitive positioning?

Yes. The A minus is based on where we write, which is personal lines for middle market. The A minus is not a restriction at all. And since you raised the stock price, let me just comment, beyond the insanity of the stock price, it's, I think, more reflective of the lack of understanding of business interruption and what the exposure is and also a lack of understanding of the results that have been achieved.

I think Kim kind of laid out a case, and it is out on our Investor site now as well. We don't worry about short term. We're long term believers in the organization and very comfortable and confident with where it's going. I certainly agree that the stated results have to reflect that. We --- in our kind of a normalized cat world, I don't know that we would be having that same conversation, but our view in answer to your question around capital A.M. Best stock price is, we believe in the long term, and we are very bullish on State Auto.

And at this time, there are no further questions in queue. I would like to turn over to the panel for any closing remarks at this time.

Thanks, everyone, for your questions, for participating in our conference call and for your continued interest in and support of State Auto Financial Corporation. We look forward to speaking with you again on our fourth quarter earnings call, which is currently scheduled for Thursday, February 18, 2021. Thank you, and have a wonderful day.

Thank you. That concludes today's third quarter 2020 earnings conference call. Thank you for participating. You may disconnect at this time.