Ladies and gentlemen, thank you for standing by, and welcome to the Allstate Fourth Quarter 2019 Earnings Conference Call. At this time, all participants are in listen-only mode. After the speakers' presentation, there will be a question and answer session. To ask a question during the session, you will need to press star one on your telephone. We ask that you please limit yourself to one question and one follow-up. As a reminder, today's program is being recorded. Now I would like to introduce your host for today's program, Mr. Mark Nogal, Head of Investor Relations. Please go ahead, sir.
Thank you, Jonathan. Good morning, and welcome everyone to Allstate's Fourth Quarter 2019 Earnings Conference Call. After prepared remarks, we will have a question and answer session. Yesterday, following the close of the market, we issued our news release and investor supplement and posted today's presentation on our website at allstateinvestors.com. Our management team is here to provide perspective on these results and further context on our recently announced transformative growth plan. As noted on the first slide of the presentation, our discussion will contain non-GAAP measures for which there are reconciliations in the news release and investor supplement and forward-looking statements about Allstate's operations. Allstate's results may differ materially from these statements, so please refer to our 10-K for 2018 and other public documents for information on potential risks. Now, I will turn it over to Tom.
Well, good morning. Thank you for joining us to stay current on Allstate. Let's begin on slide two with Allstate's strategy. As you know, our strategy has two components, increase personal property and liability market share and expand into other protection businesses. Starting with the upper oval, we have been a leader in creating differentiated insurance products with features such as declining deductibles and new car replacement. We use sophisticated pricing, have strong claims expertise, and are building an integrated digital enterprise to lower costs. We are also diversifying our businesses by expanding our protection offerings, which are highlighted in the bottom oval. We leverage the Allstate brand, customer base, and capabilities to drive growth in these businesses. We offer customers a circle of protection that includes Allstate Life, Allstate Benefits, commercial insurance, roadside services, car warranties, Allstate Protection Plans, and Allstate Identity Protection.
These growth platforms have extremely broad distribution that includes major retailers, insurance brokers at the worksite, auto dealers, manufacturers, telcos, and directly to consumers. On the right hand, you can see that this strategy creates shareholder value through customer satisfaction, unit growth, and attractive returns on capital. It also ensures we have sustainable profitability and a diversified business platform. If we move to slide three, Allstate's strategy continued to deliver excellent results in 2019. Revenues were nearly $11.5 billion in the fourth quarter and $44.7 billion for the full year 2019. Net income was $1.7 billion in the fourth quarter and $4.7 billion in the full year. Adjusted net income was $1.02 billion, or $3.13 per diluted share in the fourth quarter. For the full year, adjusted net income rose 11.1% compared to the prior year, to $3.48 billion or $10.43 per share.
It reflects excellent underlying profitability and lower catastrophe losses. Returns were also excellent with an adjusted return on equity of 16.9%. If you turn to slide four, Allstate delivered on all five 2019 operating priorities, which focus on both near-term performance and long-term value creation. The first three priorities, better serve customers, grow our customer base, and achieve target returns on capital. That is to ensure profitable long-term growth. Customers were better served as the enterprise Net Promoter Score improved at most of our businesses. Total policies in force reached $145.9 million, which is an increase of 20.7% compared to the prior year. Property and liability policies, which are bigger dollar amounts, increased by $428,000 for the prior year to $33.7 million, as Allstate and Insurance brands grew 1.3% and 2.3% respectively.
Allstate Protection Plans, which of course was formerly SquareTrade, continued its rapid growth through the addition of a major retail partner, with items in force reaching 99.6 million. Returns remained excellent, driven primarily by strong property and liability results. The underlying combined ratio of 85.0 finished 2019 at the favorable end of our revised full-year guidance of 84.5%-86.5%. You'll remember as part of our second quarter earnings release last year, we had improved this annual outlook range due to excellent operating results. As you know, Allstate's no longer going to provide underlying combined ratio guidance since return on equity is a better measure of performance. Mario's going to provide some additional context on this measure. The $88 billion investment portfolio generated $3.2 billion in net investment income in 2019, which reflects higher market-based portfolio yields, which was offset by lower performance-based results.
Performance-based results were below expectations for the quarter. Longer-term results have been strong. Total portfolio return was 9.2% in 2019. Shareholder value is also being created by building long-term growth platforms. We announced new features of a transformative growth plan, which we'll discuss next. They already continued to expand capabilities. Allstate Identity Protection is growing and launched its new Digital Footprint offering. Avail, a car-sharing platform, initiated operations. If you move to slide five, let's discuss the transformative growth plan to increase property and liability market share. The plan builds on our strengths and reflects current competitive conditions. Allstate has a significant number of competitive strengths, as you know. Particularly in property liability, we have the Allstate brand, we have pricing sophistication and claim expertise, product breadth, and a broad distribution platform that goes from Allstate agents to Esurance's direct capabilities to Encompass' independent agents.
As a result, we're growing. GEICO and Progressive are growing auto insurance market share faster through massive advertising spending and low cost structures. Our plan also recognizes that customer needs are changing due to increased connectivity and advanced analytics. Our leading positions in telematics and digital auto collision estimates are two examples of how we're embracing these changes. At the same time, a majority of customers prefer an insurance agent, we hope an Allstate insurance agent, when purchasing a policy. Are comfortable with self-service. We're increasing mobile application capabilities and building low cost, centralized, integrated service capabilities. We're now accelerating these efforts with a transformative growth plan, which has three components: expand customer access, improving the customer value proposition by lowering expenses and redesigning property liability products, and investing in technology and marketing.
Expanded customer access will be provided by utilizing Esurance's direct capabilities to sell Allstate branded products. Esurance has strong, direct capabilities, having more than doubled in size since it was acquired a little over eight years ago. As a result, we can further leverage these capabilities by selling Allstate branded policies directly to consumers. This will require us to reposition the Allstate brand, and the advertising previously deployed for the Esurance brand will be shifted to the Allstate brand. Then the Esurance brand will be phased out in late 2020. Expense reductions will improve affordability while funding investments in technology and marketing. We will also strengthen the independent agent platform by merging the Allstate independent agent offering into Encompass. This is a comprehensive plan that will make us a stronger competitor and lead to increased market share.
Now, let me turn it over to Mario to go through the property liability and investment results.
Thanks, Tom. Let's go to slide six, where we highlight strong results within our property liability segment. Policy and premium growth continued, and all of the brands had strong underlying profitability. Underwriting income of $1 billion in the fourth quarter was significantly higher than the prior year, driven by lower catastrophe losses and the continued improvement in the expense ratio, despite increased marketing investments and the write-off of the remaining Esurance acquisition intangible related to the brand name. Moving to the table, net written premium increased 4.4% in the fourth quarter and 5.6% for the full year, driven by policy and average premium growth in Allstate brand auto and homeowners insurance and the Esurance brand. Total policies in force increased 1.3% to 33.7 million in 2019.
Underwriting income of $1 billion in the fourth quarter and $2.8 billion for the full year were substantially higher than their respective prior year periods, driven by continued progress on improving our cost structure, average premium increases, and lower catastrophe losses. The underlying combined ratio, shown in the bottom left, which excludes catastrophes, prior year reserve re-estimates, and the $51 million pre-tax impairment charge related to our decision to utilize the Allstate brand for direct sales, was 84.9% for the fourth quarter. The 2019 result of 85.0% was at the favorable end of the full year guidance range of 84.5%-86.5%. Focusing on the table on the bottom right of the page, you can see our recorded combined ratio trend over time by line of business, as well as total property liability. Auto insurance profitability remained strong with a combined ratio of 92.8%, excluding the Esurance impairment.
Increased average earned premium and lower property damage and bodily injury frequency offset higher severity in 2019. Property damage severity continues to be impacted by higher costs to repair vehicles, which increases the number of total losses. Bodily injury severity increased at a rate above medical inflation indices, but we factor these severity trends into our pricing algorithms. Homeowners insurance continues to generate excellent returns. Lower catastrophe losses were significant in achieving a combined ratio of 88.4% for the year. In total, Allstate has industry-leading combined ratios for our property liability businesses. Let's go to slide seven, which highlights investment performance for the year and the fourth quarter. Our investment portfolio total return for 2019 was 9.2%. Net investment income contributed 3.7% to total return, with a stable contribution from interest income on fixed income investments, but a lower contribution from our performance-based portfolio.
Higher bond and equity valuation contributed 5.5% to total return in 2019. The chart at the bottom shows net investment income for the fourth quarter of $689 million, $97 million lower than the fourth quarter of 2018. Market-based investment income, shown in blue, increased to $735 million and benefited from proactive actions we've taken, including a duration extension in our property liability portfolio. Performance-based income, shown in gray, was lower than recent trend and included lower valuations of $74 million pre-tax on two private equity limited partnerships. Given the performance-based portfolio's impact on reported income, let's turn to slide eight to review the purpose, makeup, and results of these investments. Our performance-based strategy delivers attractive long-term risk-adjusted returns and is well-suited for long-dated liabilities and capital. The strategy diversifies our portfolio through idiosyncratic equity returns that complement our market-based strategy.
The portfolio is broadly diversified by asset type, including private equity and real estate, as well as geography, sectors, and partners. We invest in funds, co-invest, and have direct investments as well. The $8.7 billion portfolio had a one-year return through September of 7.6% and three and five-year returns of around 11%. While idiosyncratic investments generate higher returns, they do create income volatility, as you can see on the right. The portfolio generated $469 million of investment income in 2019, but all of that occurred in the first three quarters. In 2018, the portfolio generated $716 million of investment income and $145 million in the fourth quarter. We continuously monitor the portfolio and do not believe the lower valuations on two private equity investments in the fourth quarter are indicative of the broader portfolio. Slide nine highlights results for Allstate Life Benefits and Annuities.
Allstate Life, shown on the left, generated adjusted net income of $76 million in the fourth quarter and $261 million for the year. The fourth quarter improved by $7 million compared to the prior year quarter, driven by higher net investment income and lower operating costs. The full-year income of $261 million was 11.5% below 2018 due to the third quarter write-down of deferred acquisition costs in connection with the annual actuarial assumption review, which reflected lower interest rates. Allstate Benefits adjusted net income of $16 million in the fourth quarter was $10 million below the prior year quarter. This decline was primarily driven by the write-off of acquisition costs related to the non-renewal of a large underperforming account. For the full-year, adjusted net income of $115 million was $9 million below the prior year.
Allstate Annuities, shown on the bottom right, had an adjusted net loss of $33 million in the fourth quarter, driven by lower performance-based income. Adjusted net income of $10 million for the full year 2019 was below the prior year, reflecting lower performance-based investment income in the first and fourth quarters. Let's turn to slide 10. The service businesses continued to grow the number of customers protected, with policies in force increasing 42% to 105.9 million. This is largely due to the increase in Allstate Protection Plans. Revenues grew 21.9% to $434 million in the fourth quarter, reaching $1.6 billion for the full year 2019. Adjusted net income was $3 million in the fourth quarter and $38 million for the full year.
Adjusted net income increased in 2019 compared to the prior year period, as improved loss performance in Allstate Protection Plans and Allstate Dealer Services were partially offset by continued growth and integration investments at Allstate Identity Protection. Slide 11 highlights the excellent cash returns provided to shareholders. In 2019, we returned $2.5 billion to common shareholders through a combination of $1.8 billion in share repurchases and $653 million in common stock dividends. We repurchased 16.4 million or 4.9% of common shares outstanding over the last 12 months and increased our book value per share by more than $15 to $73.12. The $3 billion share repurchase program announced in October 2018 was completed in late January 2020. Yesterday the board approved a new $3 billion share repurchase authorization to be completed by the end of 2021.
We also reduced the cost of capital in 2019 by redeeming multiple series of preferred stock and issued two new series at lower rates, which will reduce preferred dividends by $17 million annually. Let's move to slide 12 and discuss our long-term return on equity goal. Starting this year, we will no longer provide annual property liability underlying combined ratio guidance, but instead discuss returns on equity. As we've discussed over the past two quarters, this is a better measure of our performance for multiple reasons. It is a broader long-term measure of performance than the underlying combined ratio, which focuses on just one part of one of our businesses and excludes significant items such as investment income, catastrophes, and other protection businesses income. It factors in capital management and is more correlated with stock price. It fosters a better comparison with our peers.
On a long-term basis, our adjusted net income return on equity goal is 14%-17%. This range also aligns with executive compensation as discussed in our annual proxy statement. I will open up the line for questions.
Certainly. Ladies and gentlemen, if you have any questions at this time, please press star then one, and as a reminder, please limit yourself to one question and one follow-up. You may get back in the queue as time allows. Our first question comes from the line of Elyse Greenspan from Wells Fargo. Your question, please?
Elyse, you might have your phone on mute.
Hi. Yes, I'm sorry. Good morning. Can you guys hear me?
Yes.
My first question is on your ROE target, the 14%-17%. Just a couple of questions there. If we can get a sense of what you define as the long term, and then how you see 2020 coming in reference to that 14%-17%. My last question on that front would be, are you going to go into any particular details on the returns that you're targeting within that, within property liability, versus your life-related businesses or services?
Elyse, thanks for the question. First, let me deal with the second question first. We're not going to give out a number for 2020, which goes to your first question, which is, we look at this like a rolling three-year basis. Over a three-year period, cats go up and down, things go up and down, we should be able to earn a 14%-17% return. We got there by really looking at a couple things. First is, what's the level of economic rent we can capture by how good we are in the market we operate in? As you know, in the property liability business, we run a substantially more favorable combined ratio than the industry, that's worth a substantial amount of economic return.
We also look at it from what's the appropriate return for the risks that our shareholders take, we feel like 14%-17% is a good goal for us to have over that three-year basis. I don't think you should try to use this to decide what 2020 forecasts are. There's a lot of better ways to do a forecast. We're not going to break it out by component because our goal here was to get up to the corporate level, say, what are we doing in total? With the underlying combined ratio, we basically zoomed ourselves in, including you all, into how well was auto insurance doing, we've neglected to focus on the other opportunities to do. Obviously, the 14%-17% includes a whole bunch of components. The annuities, as we've talked about before, are a large drag on that.
If you were to intuit between the 300-plus % drag that has on the ROE to the other businesses, you would say the other businesses are higher return. Of course, most of that is property liability. That will change over time, obviously. The acquisitions, like the purchase of Allstate Protection Plans, obviously, that business is doing extremely well from our standpoint. It's well above its acquisition targets. It does put a drag on ROE. Interest rates would go the other way, particularly in the property liability portfolio. What we did was stand back and say, "What do we think is right over time on a rolling three-year basis?
Okay, that's helpful. My second question, you guys rolled out your new growth-related optimization efforts in December. Just want to get a sense, as you change your commission structure a little bit, how you envision that playing out in terms of both a pickup in new business growth and any kind of drag you see that having on the underlying loss ratio, I guess specifically, to your auto book.
Let me maybe go up for a minute, ask Glenn to talk about the commission changes you're talking about for 2020, which I think those specific ones you're referring to. Transformative growth is a multi-year effort. We expect it has both expanding customer access, improving customer value, and investing in technology and marketing to help us grow market share. The improving value will be done two ways. One, reducing expenses, and that's expenses across the board, and we can talk more about that later if you want, but that's really across the board, and redesigning the property liability products. Glenn did make some changes to the commission structure for 2020, which he can talk about.
Yeah, thanks. Elyse, we've got a few things going to stimulate growth in what is a very competitive environment right now. We're pleased that we're able to keep growing and, to the second part of your question, deliver returns. I'll answer that part first. Do we expect it to, as we grow, to deteriorate returns? I would say no, because we've evidenced that we're pretty disciplined about that, and as we've grown, we've kept our eye on the returns. The specific actions, one, you referred to the change in compensation. We moved some of the compensation variable comp from agents' renewals to new business to incentivize growth. We want to pay folks for growing their business. We're also putting more into advertising and managing our lead management platform.
Okay, that's helpful. Thank you very much.
Thank you. Our next question comes from the line of Greg Peters from Raymond James. Your question, please.
Good morning. My first question will be around the expense ratio improvement that you guys generated in 2019. I think that's probably the biggest surprise, at least from an outside observer perspective. In some of your previous comments, you said there was some benefit for incentive comp. I guess my question is around how much of that 90 basis points was sustainable considering the strong results that you posted last year. Is that going to boost up the incentive comp for 2020? Dovetail the expense ratio improvement with What kind of anticipation or expectations should we have about continuing improvements? Should we factor in some modest improvement every year? I know you guys are rolling out the integrated services platform, et cetera.
Greg, let me take incentive comp. Mario can talk about expenses. We set incentive comp targets every year. We have two programs. One is the annual incentive plan, which is our management's compensation program, and that's set every year based on how well we did the year before. It's not like you get a couple of years there. We also have, obviously, a longer-term plan, which is our performance stock awards, which are tied to return on equity. Those you shouldn't expect fall over from one year to another as it relates to expenses. If we exceed the targets established by the board, obviously, we have to put that into the P&L. Presumably, that's because we're making more money than everybody thought.
Yeah, Greg, thanks for the question. It's Mario. I just would add a couple things. First, I think we're continuing to be pleased with the progress we're making on expenses. As Tom talked about earlier, it's a core part of our transformative growth strategy and one of the ways that we're going to further improve the customer value proposition and create the capacity for the growth investments that he alluded to. Our work in this area is not done. We're going to keep focusing on expenses and look to drive down the expense ratio. One thing I'll add with the fourth quarter specifically, the impairment charge related to the Esurance brand was worth about six tenths of a point on the expense ratio.
If you look year-over-year, you see some real meaningful improvement in the expense ratio in the quarter that gets a little bit masked by that non-recurring charge.
Mario, is that where the Esurance expense ratio is? I think was up 27.5% in the fourth quarter. Is that what you're talking about?
Yeah, it's in the expense ratio. It's worth about six tenths to the overall property liability number. Obviously, more of an impact to Esurance specifically.
Got it. Great. Thanks. The second question, let's pivot to the performance-based investments. Obviously, it's generated a strong result for you over a long period of time, but there's volatility. This past year, there was volatility. For outside observers, can you provide any guidance for us on how we should think about that volatility as we think about the 2020 outlook?
Greg, thanks. Let me make an overall comment about investments, John can go through the more color on performance-based, recognizing we don't like to give next year's projections out there. We proactively manage a portfolio. John's created and uses a really comprehensive capital allocation framework that helps the investment team decide what and when to invest. It keeps track of the decisions we made, not just the investments we made, so we can track those, but also any opportunities missed or any losses we avoided. We watch our money well, and that's on all parts of the portfolio. John can talk about what we do, how we deploy that in performance-based, and what it means.
Yeah. Thanks for the question, Greg. When you go up and think about the whole portfolio, it helps put performance-based in context and why we like it. The portfolio is about $88 billion in size. Majority of it's invested in high-quality fixed income instruments, about 18% of that is equity, and roughly half of that is performance-based. Just to kind of size it within the overall context. We like performance-based because it focuses on private markets and really expands our opportunity set of investments that we can play with. It allows us to create, in our belief, a more efficient frontier or a more efficient portfolio that's well suited to our overall liability and capital structure. What comes with that is increased volatility on income. We expect volatility. We'll expect it in the past, and we'll expect it going forward.
We believe that we're compensated for that volatility of income. In this particular quarter, that volatility, especially if you look at it year-over-year on a quarterly basis, the difference was $97 million, and about $74 million of that was tied to two investments. That obviously was a disappointment, we've gone back in, we recognize that when you look at the underlying investments here, they range from timber and agriculture to real estate to private equity to a number of different investments in there. It's hard to pin down with great precision in any given quarter how that's going to play out. I would expect continued volatility.
We have also gone through as part of our regular risk and surveillance processes, looked at all of our holdings as we normally would do and believe that this was a bit of a one-off type of quarter. We remain confident in the effort going forward despite the performance in this quarter. We believe we have a strong team. When we look at measuring this, we really look to longer-term measures. We look to 3-, 5-, and 10-year, and some of those measures are in the investor supplement. Those tend to be holding true, we really aren't changing our overall commitment to it and expect while there may be bumps in the road in any given quarter, that long run, it's an important part of what we do.
Got it. Thanks for your answers.
Thank you. Our next question comes from the line of Paul Newsome from Piper Sandler. Your question, please.
Good morning. I was hoping you could just give us a little bit of an update as to how you see the current pricing and competitive trend in auto. Obviously, last year was a pretty competitive year. If you think that things are stabilizing industry-wide or not, love to hear your thoughts.
Well, this is Tom. Let me make a couple of overall thematic conversations, and then Glenn can talk about what he's seeing in the market by competitor. Our philosophy has been to, as you know, profitable growth and long-term profitable growth. We raise prices when we think we need to or when the trends show we should do it, we don't get behind on it. We're not big in terms of reducing price to try to grow. I'd like to say around here, anybody can lose their way to increase market share, and that's not our plan. The other thing I would say is, you always have to be careful of the percentages, because it always depends where you start.
We have some competitors who are higher price than us, and they're reducing their prices, and other carriers who are lower than us on certain risks, and they're raising the prices. Glenn could talk about the percentages, and there are obviously some overall macro trends at work in auto insurance pricing. We have to be careful not to be too specific as to what it means for next quarter sales.
Yeah, thanks. As Tom said, I'll lead you to one set of numbers that I think tells the story. Our average premium in auto was up 3% year-over-year, and our average loss in expense was up 2.5%. We were disciplined in managing to a positive return and actually made a little progress during a time where, as you said, it was very competitive out there. We did see some competitors take sort of broad-based rate reductions, which is why the overall CPI is negative. That, as Tom said, not a trend we'll follow or something we subscribe to. We look state by state. To your question about what we're seeing more recently, is it stabilizing? I would say, I'll give you a qualified yes. There is a little bit more rate activity starting to flow into the market.
You're seeing less price reductions or really no price reductions in the recent past, and you see some increases out there. I'd call it more stable.
Fantastic. A second question. Obviously, on the commercial line side of the world, there's a lot of conversation about basically casualty-related inflation trends. I recognize that that's not a huge part of what you do as an auto and home insurer, but could you talk to whether or not you're seeing any of that in the fairly modest pieces that are casually a part of your business?
Paul, are you talking about bodily injury costs? Is that where you're zooming in?
Well, actually, I'm thinking more of pure liability in the home insurance business, for example. Slips and falls and lawsuits and things of that nature.
Yeah. Not seeing much of a trend on the home side on that. Our trends that are more meaningful for us on the home side really are the mix between frequency and severity of our property, first-party losses. We're not seeing really anything on home.
Thank you.
Thank you. Our next question comes from the line of Yaron Wiener from Goldman Sachs. Your question, please.
Good morning, everybody. First question is around net premiums written growth. I think it slowed down a little bit, and I guess in particular, you see a little bit of a slowdown in renewal ratios and in new applications, especially in auto. Can you maybe talk about that and maybe how that ties to the Transformative Growth Plan?
Sure, we can. First, as I said earlier, it's all about profitable growth for us. The Transformative Growth Plan should bring our expense ratio down and enable us to improve our competitive position, which should drive to higher growth. In 2019, throughout the year, there was different stories in the Allstate branded business than the Esurance business. Glenn, maybe you can talk about the Allstate branded business, and Esurance for 2019 was still under Steve. It's now, of course, under Glenn as part of our Transformative Growth. Steve, maybe you can pop in on Esurance.
Yeah, I'll start us off. Just at a high level, I'd tell you, we feel good about the fact we were able to grow in a really competitive market. As has been pointed out before on the call here, negative CPI, premiums up 5.5%, policies in force grew across all major lines. It's hard work to do that in a competitive environment, so it's challenging. To your point, that moved down a little bit in the fourth quarter, we're still in the upper bounds in that sort of, call it the upper third of our long-term trend on both retention and new business.
We're coming off of a high base year, we're taking the actions short-term as we build out transformative growth, taking the actions short-term to grow, which I referenced before, of changing agency comp, where we're shifting more towards new business, putting more into marketing.
For 2019, Esurance had about 8.5% increase in net written premium. For the year, it feels good, but that trend declined throughout the year, the fourth quarter was only about 2.6%. What we did is we grew fairly rapidly last year also, so we had that new business penalty, in some states, we took some fairly significant rates more in the second half of the year. That really slowed down our growth as we went into the third and fourth quarter. Overall, we felt good about the growth we had. We feel good with the business' position going forward. We have to grow profitably.
That's helpful. Thank you. Then my second question just goes back to the ROE guidance range for the next three years. I guess, what would be the largest components that would drive ROE up to the upper end of that range or the lower end of that range? Does that range also contemplate the possibility of this disposition of the annuities business?
Obviously, there's a number of components here. The obvious piece is what's your combined ratio and underwriting income. That's impacted by what your underlying is, also what catastrophes are, which as you know, bounce around. There's what happens to investment income, we've talked about that at length today. There's also the amount of equity you have at play. So, as we've been building equity, we're still buying shares back, that has an impact. As it relates to the future, this is kind of an as is business model approach. We didn't assume anything was going to change. As I mentioned to Elyse, things will change in the future, we still feel like 14 to 17 is the right range. Lots of stuff could change.
The interest rates could go up, and if interest rates go up for the property liability business, most of that income falls right through to the bottom line. If you look at our investment income over a long period of time, we are down substantially. Now we've made that up by reducing the combined ratio and improving underwriting income, but that will change over time. Obviously, the Annuities business would be a positive. To the extent we decided to invest in growth in an acquisition, that would be a negative. We felt like in total, the 14%-17% was a good goal for us. If it changes as the world changes, then we'll let people know. It's a three-year kind of rolling average for us.
Thank you very much.
Thank you. Our next question comes from the line of David Motemaden from Evercore ISI. Your question, please.
Hi, thanks. Just to follow up on the transformative growth plan. Just trying to get some sense for when you think we should start to see the benefits of those changes start to roll through in terms of more competitive product and then an uptick in new applications, better retention, and higher PIF growth after we've seen a bit of a deceleration in all of those metrics.
Yeah. David, first, it's a long-term plan. This is going to take us multiple years to do. There's obviously components of it, and some parts will have a bigger impact early, and other parts will have a larger impact later in the game. We expect early in the plan that expense reductions will play a significant role in us improving our affordability for customers and in getting to that second part of improving the value proposition. As you go forward, we would expect that redesigning our property liability products will drive more differentiation. We were first to market with declining deductibles, first to market with new car replacement. There are other ideas and things we have in place that we think can drive more value, but it's going to take more time to put those into the marketplace.
In terms of improving the customer value proposition, expenses first, redesigned products later. The reason redesigned products is later is in part it requires a new technology platform, which is the third piece. We have to invest more in a new technology platform, particularly our product management platform and our customer experience layer. That will take us a while to do. We think we can still cut expenses and reduce the overall expense load while we are investing in those, and that's our goal. At the same time, we're going to invest more in marketing earlier in the process because we believe that will drive growth. Glenn talked about what he is doing to enhance the Allstate agent value proposition, whether that's increasing new business commissions or freeing them up by taking low-value work out of their offices and putting it into integrated service.
The other piece is to expand the access, and that's happening as we speak. The access to people who called for the Allstate brand or got onto our website and Allstate brand had some restricted rules on it, which we've started to open those up now. For example, if you called during working hours, you could get a different experience than if you called after working hours. You should have the same experience. You're getting online, it shouldn't matter what time of day you do that. We're changing those rules now. What Glenn is also doing is taking the Esurance capabilities, which are exceptionally good in direct. He's putting those together with the Allstate branded capabilities, which do some direct today. That will then be utilized for the Allstate brand. That will take us some time to do.
One, obviously, you've got to put all the operations together and call centers and get the technology to work the same and get the order, the flow on the website all to be the same. We also need to reposition the Allstate brand. The Allstate brand, of course, is extremely well known, great unaided awareness, high trust marks. People, it is one of the great brands of the world. That said, it's not viewed as modern as Esurance is. Esurance is built for the modern world, a little hip, a little cleaner. We are working hard on repositioning the Allstate brand so that it brings both of those together, so that we both serve our customers with agents when they want them, and we serve them direct when they want.
It's a long-term plan, but you can see it's front-end loaded on some parts and back-end loaded on others.
Got it. Just with the customer value proposition and some of the increases in the bodily injury severity that popped above medical CPI this quarter, which is the first time I think in a bit it has. Is that already embedded in your pricing, or is that something that would, I guess, go the other way against some of these expense saves that may be able to pass through a more competitive rate?
Yes, I'll take that. This is Glenn, and we've seen some of the trends come through that you've heard about from others on bodily injury. It is embedded in our reserve position that we have, we think through these things as we set our reserves, as we set our pricing strategy. You've seen over the last couple of years where we've run hot on physical damage severity, and yet it really hasn't flowed through to the margins because we're pretty quick at adapting from a pricing standpoint and on a market-by-market basis. I would tell you, we're watching it. We've got good process. We have a great claims department that looks at both the technology and the process that they follow to handle claims. Where it goes beyond what claims can manage as they're paying claims accurately, we get it into pricing quickly.
Okay, great. Then if I could just sneak one last one in, just in terms of the high end of the ROE guide. I guess just thinking about potentially, there've been a few transactions in the fourth quarter on the life side, and it seems like there's a pretty vibrant interest, a lot of interest in some of these runoff books of annuities or life insurance. Just wondering, in Illinois, I believe there's a division of company rule that can allow you to carve out specific reserves out of that Illinois legal entity and divest to a potential acquirer.
Just sort of wanted to get an update on your view in terms of your thinking in terms of whether you see any opportunities there to potentially reinsure or divest that book and maybe get to a more sustainable high end of that ROE guide, given it has like a 300 bip drag on the ROE.
It is a third question, this is different than the other ones I will take. There are other people in the hierarchy.
Sorry, yeah. I apologize, yeah.
The answer is we've been working on annuities for a whole bunch of years trying to get it better. We've made a lot of progress. The extent there's something else we can do, whether it's use a division statute to get a different solution reinsured. If we can do it in a way that is economic for our shareholders and is a good risk-adjusted return, we'll do that. If we do that, I don't think anybody would be unhappy if you looked at if we were above the upper end of the guidance.
Great. Thank you.
Thank you. Our next question comes from the line of Amit Kumar from Buckingham Research. Your question, please.
Thanks. Good morning. Two quick follow-up questions. The first, I guess, is in response to the previous question on the shift in the new plan. Can you give us any early sort of feedback from agents, number one, on the integrated services platform? Number two, also what is the feedback on shifting the bonus and the commission structure from existing and new businesses?
Amit, I think we got it. We were all leaning in towards the speaker to make sure I heard. It was a low light. I think Glenn can talk about integrated service and the commissions. Maybe let me go back to Transformative Growth, and make a comment really on behalf of the customer. The agents are part of our team. We want them to be successful. We're leaning in hard to make them successful. They drive a bunch of our business, and we're going to work hard at making them successful. That said, there's a range of opinions as to what people think about what you're doing. Let me come back to the customer. What we are trying to do with Transformative Growth is stick with our customer philosophies. One of our philosophies is you should pay a fair price for the risk you create.
We're very sophisticated in the way we do pricing. We have telematics, which is the new end of it with Drivewise. We also have a philosophy that customers should get what they pay for. As it relates to the auto insurance, those same things apply. If you get extra service, and extra advice from the help of an agent, that should be something you should be prepared to pay extra for. If you want to do all the stuff yourself, that's something that's a different price value proposition. We are working hard to make sure that agent value proposition is as effective and efficient as it can be. People, as we said, I think we said when we talked about this That over half the people want an agent, and that's higher as it reflects our customer base.
We want to help them be as successful as they can, and that's why Glenn is working on the couple programs he is.
Yeah, I'll just add to it. We talk to our agency force all the time, talking to agents every week. We've got national advisory council, regional boards. We spend a lot of time with our agency force. I'll give you the range of reactions to things.
On the overall, the big thing out of transformative growth, we're going to pull the marketing in from Esurance and add to the Allstate brand. As you might imagine, wildly popular. Everybody feels really good about the fact that there's going to be more marketing in the pipeline, and we're focusing on the Allstate brand. On the commission change, you've got folks that they were growing fast, and they're built to grow. They like it a lot because it leans right into their strength. You've got folks that really weren't growing or weren't built to grow. You have a range of folks who either are excited about making that transition and retooling their business, or ones that say, "I wasn't really ready to retool my business." They need to in order to earn the commission that goes with the new business, the higher new business commission.
As it goes to integrated service, as Tom said, lots of different opinions out there. It really is about freeing people up for growth, and what I can tell you is the folks that are in it are growing faster than the folks that aren't. We think we're creating something that's going to be really valuable for the agency force.
If you go back to the customer, and with Glen's higher new business, when you ask customers, "Where does an agent add the most value?" It's almost always on either the purchase of the product or doing an insurance review, helping me sort out stuff. Very little of it is on changing addresses. That's what Glen's working on.
That's a fair comment. Very quickly on the loss cost trend. I know you briefly talked about it, net-net, can you talk about when you look at the frequency and severity trends, is the expectation that this is how it sort of remains for the foreseeable future? What is your outlook on the loss cost trends here? Thanks.
I'll take that, Glen. Let me go up a little bit, because I think if you look at the whole system and you say there's the premiums that come in, and how we may change those over time. There's expenses that are going out the door, you split your loss cost by frequency and severity. Over the past year, three out of those four have been going well for us. Our premium, as I mentioned, auto's up 3%, average premium. Home is up 5% average premium. The respective loss and expense on those lines are two and a half and four. We've made up a half a point and a full point in terms of actually improving over the time period. It's hard for me to say.
I can't predict that severity will be flattening or up more over a year or two years from now. What I can tell you is we've been really disciplined about managing it quickly into each market on a market-by-market basis. Our pricing sophistication, our state management team has gotten better and better, and we're able to react in real time to those changes and manage to the types of returns that we've been able to deliver. While I can't give you a perfect prediction on where the severity will go, I feel really good about our ability to react to it in real time.
Got it. Thanks for those answers.
Thank you. Our next question comes from the line of Brian Tunas from Autonomous Research. Your question, please.
Hey, thanks. Good morning. A couple for Glenn. Tom mentioned that over half of Allstate brand customers want an agent. I thought that was interesting. In terms of just the pathway of becoming an Allstate brand customer, what percentage of people, I guess, start on the internet, Google insurance or something like that, and is that the lead generation mechanism? I guess just digital or the internet versus an Allstate agent is a little league coach or something like that.
Brian, Glenn can give you through some of the percentages we give out, but this is not an or conversation. A lot of times people are like, well, they want to buy direct or through an agent. It's an and. We'll sell to people any way they want. If you want to start online, buy from an agent, if you want to call an agent, buy online. We want to sell as much as we can to as many people as we can through as many ways as we can.
It's a challenging question, I'll tell you, because there's a lot of stops in the process where, for example, people go online, they start to get a quote, and the agency that's nearest to them pops up with that quote, and they decide to drop off, make a phone call, and that's one path. There's multiple paths people take through. The vast majority of our business is written by a local agent. That said, all the data and research out there tells you that the vast majority of people start their process online and at least get some understanding and then want to talk to somebody about it. We've got a mix of that and imperfect data for me to give you any more specifics on that.
I was hoping you could give us a bit of a peek on some of these new product initiatives. I guess, in particular, I'm trying to understand the price sensitivity or whatever. Obviously, you're a captive agent. Are you thinking new business pricing down 10% or something like that? I'm just trying to understand how much of a driver do you think that is for potential new issued applications?
Brian, I'm not sure we understood the question. Maybe you could rephrase it.
Yeah, no, I'm sorry. I'll ask it this way. Two or three years ago, Travelers, who's obviously an independent agent player, they just announced that they were cutting new business prices by 10%. I could see how that could have more of an impact in an agency distribution channel versus the captive. I'm wondering if that's how you see it as well, or if you think that, am I missing something or is there a similar level of price elasticity in captive as what you might see in agent? Could that be a big driver for growth?
Okay. Let me maybe start with the last piece and work my way up, and anybody else wants to jump in. First, it's of course, always hard to tell how much price elasticity there is within an agency. In general, you would think that an agency that had multiple companies and could share those multiple quotes with you, there would be more price elasticity than in a captive agent. That said, people can call around to three or four people and get their points of view. Auto insurance in particular is price sensitive. Although we've talked at length at us trying to have profitable growth. That's what we're trying to do. The redesign of the products is on all of our products, auto, home, and a variety of different products where there are things we think we can do to make them less complicated.
For example, this year we rewrote the renter's policy in one state and took its size down by 40%. That gets to less complexity, helping customers understand it. We also have a variety of other products in the circle of protection we're trying to get to. We don't sell identity protection today through our Allstate agents. Makes sense that we have customers that come to Allstate agents probably would like to have their identity protected. How do we do that? How do we get the systems to work? How do we get the technology platform to it? How does Glenn factor it into compensation and everything else? It's something we're working on. That's part of transformative growth on a longer basis. We have many products we sell through many channels, but we don't sell all of our products through all of our channels.
That's part of transformative growth, is how do we sell more of what we sell already in more of the places that we sell.
Thanks.
Sorry. Okay. I think we probably have time for one more question. How about one more question?
Our final question for today comes from the line. Michael Zaremski from Credit Suisse. Your question, please.
Hey, good morning. First question, the P&C investment portfolio, the duration has increased fairly measurably over the last few years up to 5+ years now. I was just curious, do you continue to see that elongating, or are you at the high end of the range you're comfortable with versus the duration of liabilities?
Yeah. Hey, Mike, it's John. Thanks for the question. Look, you never know exactly what the future holds, so we're always going to respond to economic conditions, overall risk environment of our firm, and structure the portfolio that way. I would say that what we've seen in terms of the duration extension in recent time, and a lot of that's taken place in the last 18 months, was more of a reaction back to strategic norm. As interest rates had been rising prior to that with the cycle that the Fed was in, we thought that it was prudent to stay a little bit short and look to recapture higher yields when there's a better opportunity. We coupled that with, a little bit if you go back 18 to 24 months, the economic situation in the U.S. and the world was different.
It was softer at that time, it wasn't clear that policymakers were going to engage. They were in a tightening cycle. We thought that they needed to engage. We also wanted to put a little bit of recession protection back in the portfolio. When you have more duration, if the economy falls off, it tends to help your portfolio and balance it out more. We did actively, highly coordinated internally with all our risk processes, extend the duration. It had a number of benefits that we're pleased with. It did increase income, which is a positive. It also did build in a more balanced portfolio, balanced out our risk-based assets. This isn't perfect math, but if you look back at where rates were when we did most of the extension relative to where they are today, it's been a great price increase in our fixed income securities.
Rates are roughly market yields when you include where treasuries were and where corporates are, roughly about 100 basis points lower than that time. Going forward, we'll continue to monitor. What I can assure you is that it will be highly coordinated inside. Tom talked earlier about our numerous processes, our capital allocation process with investments. We also work very closely with our enterprise risk group to make sure that all our moves are well understood. We believe that proactive management investment portfolios is a way to add value to the enterprise.
Okay, that's helpful. Yeah, definitely, it's been a great call. Lastly, clearly a lot of questions about the new commission structure. I may be curious, has policy growth, is it more and more weighted towards maybe a disproportionately smaller percentage of agencies? Has that changed over the years? I know commission structure did change a number of years ago, too, it seems like it worked out okay. Sales didn't fall apart. Maybe you could any color there would be great.
I mean, Glenn, what our strategy is to have all of our agents growing and all of them be productive. As you point out, not all of them get there every year. Sometimes that's what we do in a local market or what the local market conditions. Some of that's what they do or they don't do. The commission change wasn't to address a problem. It was to seize an opportunity. We move commissions around frequently. It happens all the time. People do it in all the businesses. What our goal is to make sure that their objectives are aligned with our objectives. Thank you all. Let me just close by saying thank you. We had a good 2019. We're focused on transformative growth in 2020. We'll talk to you next quarter. Thank you.
Thank you, ladies and gentlemen, for your participation in today's conference. This does conclude the program. You may now disconnect. Good day.