Hello, everyone. I'm Jay Goldberg from Barclays. We're very pleased to have with us today Tom Wilson, who's the Chairman and CEO of Allstate. Allstate is one of the largest insurers of vehicles and homes in the U.S., and it also has a life insurance and retirement savings business. Tom will begin with some opening remarks, and then I'll moderate a Q&A session. With that, it's my pleasure to turn it over to Tom Wilson.
Thank you, Jay.
You're welcome.
Well, good morning. Thank you for investing your time to learn more about Allstate. Joining me today is John Greig, who leads our investor relations. If you don't know him, he's very knowledgeable about our company. Before we begin, this says I'm going to be making some forward-looking statements and references to non-GAAP measures. I'm also going to use some examples of hypothetical catastrophe losses as a way to illustrate the components of our risk management programs. This presentation, and more specific information, is obviously on our website. The first thing, of course, on everybody's mind these days are hurricanes, let's begin there. Hurricane Harvey and Hurricane Irma are significant catastrophes that made landfall in the United States as category 4 storms.
Hurricane Harvey made landfall in a less densely populated area near Corpus Christi, and it dropped a tremendous amount of rain on Houston, which caused, of course, widespread flooding, which I'm sure you saw on TV. Auto insurance covers flood losses under the comprehensive coverage option, but flooding for homes is covered by the federal government's National Flood Insurance Program. Irma is primarily impacting Florida, but it's really too early to determine the extent of its damage. What we do know is it's pretty much widely across the entire state. What we do know is that at moments like this is when Allstate shines for our customers. We're prepared, we're out there early, we're in force, we're out there delivering the Good Hands promise. Of course, we're experienced at this. We've resolved over 2.6 million catastrophe claims over the last five years alone.
We currently have deployed over 3,000 people on Hurricane Harvey, and we pre-positioned people and technology last week for Irma. Our help goes well beyond just the insurance claims. It includes employee and agency owner volunteer work, fundraising efforts. We're doing some programs now on educating all consumers, not just our customers, on how to navigate the rebuilding process, get the most out of their insurance and the government support, and then, of course, we support nonprofits. We're also prepared financially with a comprehensive reinsurance program. This includes a nationwide program that's covering $3.9 billion of losses over a $500 million retention. That excludes New Jersey and Florida property, which have separate programs. Florida property is written through a separate company, Castle Key, which has reinsurance with a $20 million retention and $638 million of total limits.
We also have reinsurance for auto losses in Florida and several Southeast states, which is $200 million of coverage over a $300 million retention. That nationwide program I mentioned also includes property and auto, not all programs do that. These are complicated programs. They're explained fully on our website. We repositioned our homeowners business, as many of you know, after incurring significant losses in 2004 and 2005, by raising prices, tightening underwriting standards, reducing the number of homes that we actually insure, and using much more reinsurance. The results are that this line has generated $5.6 billion in underwriting profit over the last five years, despite $7.6 billion of catastrophe losses. This slide shows two hypothetical examples of how Allstate's risk management program works. If there was a hurricane that had $50 billion in total insured losses, that's not total losses, that's total insured losses.
Each scenario is independent of each other. It includes a number of simplifying assumptions, and it's based on Allstate's statewide market share as of year-end 2006. Of course, as you know, hurricanes go in different places in the states. The first example is a Florida-only event, and the second example impacts the Carolinas. Example one assumes a Florida-only event, over which one half of the insured losses, or $25 billion, is associated with personal lines insurance. Property losses are assumed to be pro rata to our 1.9% market share, which equates to $380 million of losses, of which $360 million would be recovered through reinsurance.
Our market share of auto insurance is, of course, much higher at 10.7%, as you can see from the middle column on that left example, which on a pro rata basis results in a little over $500 million of losses, which would trigger recovery of $200 million of reinsurance. The net of this hypothetical event in Florida would be a pre-tax loss of about $333 million and a loss after tax of $216 million. The second example assumes the losses occur in the Carolinas. Those don't, of course, access the Florida reinsurance programs, but instead access the nationwide program, which has a loss retention of $500 million and 5% of losses above that level. This would trigger a recovery of nearly $1.7 billion. The net loss is about $400 million after tax, given the higher retention level of this program.
The point of these examples is that Allstate proactively manages our risk and return for our shareholders, making sure we control risk from catastrophes. We make sure we take care of our customers at the same time. We, of course, disclose catastrophe losses every month if they're over $150 million. Our August results will be announced on September 21st, and September's will be on October 19th. I'd point out that determining catastrophe losses is complicated. It takes a long time to really determine exactly what the remediation costs will be. Those loss estimates change significantly over the course of time. Let's move past the weather channel and discuss how Allstate's broad-based business model generates attractive returns. In total, we have 75 million policies in force, $81 billion in investments, which generated $37 billion of revenue in the last 12 months as of June.
Operating income over this period was $2.4 billion, which was a 13% return on common equity, depending which income measure you want to use. We manage our business for both near and long-term value creation and have multiple initiatives underway across various time horizons so that we can achieve our strategic priorities. First, in the near term, we're focused on driving results and guided by our 5 operating priorities for 2017. The first 3 priorities, better serve our customers, achieve target economic returns on capital, and grow the customer base are intertwined. Our biggest challenge over the last several years has been doing all 3 of these in the face of significant increase in the frequency of auto accidents. On a slightly longer time horizon, we have a set of initiatives focused on continually strengthening our businesses to create sustainable returns.
We're further improving the value to customers through local Allstate agencies with our Trusted Advisor initiative in response to aggressive competition from direct insurance companies like GEICO and Progressive. At the same time, we're expanding Esurance, which is our direct writer, to include underwritten homeowners insurance and enhancing its customer experience by leveraging technology. We're building an integrated digital enterprise to substantially improve the effectiveness and efficiency of our operations, particularly in the claims resolution process. I'll give you an example of that in a minute. Enhancements in customer connectivity are also giving us the opportunity to be a leader in telematics. The Drivewise and DriveSense offerings improve the accuracy of auto insurance pricing, enable us to give some customers lower prices, enhances the customer value proposition. Additional shareholder value is also being created by building long-term growth platforms such as Allstate Benefits, Arity, and SquareTrade.
Let's touch on each of these horizons to set some context for our discussion with Jay. Auto insurance margins have been and are a key driver of Allstate's returns. On the left graph, you can see that Allstate, GEICO, and Progressive have consistently outperformed the industry in auto insurance profitability with a combined ratio below the industry average, which is shown by that dashed line. Allstate brand over the last 10 years has averaged nearly $900 million in underwriting income annually. When compared to our primary competitor in the local advice and branded customer segment, State Farm, this difference is over $20 billion over the last five years. I would point out they are bigger than we are as well, they shouldn't be exactly additive.
In early 2015, in reaction to a sudden and large increase in auto insurance loss costs, we instituted an auto insurance profitability program that included raising prices, lowering growth, reducing expenses, and implementing process improvements. That makes it hard to achieve all three of those objectives in your operating priorities. As you can see from the graph on the upper right, auto insurance profitability is now back to the levels we achieved in 2014. As a result, we're now instituting a comprehensive program in the Allstate agency business to increase policy growth. In addition to our current results, we're focused on making our business stronger. Our Trusted Advisor strategy will improve the value delivered by Allstate agencies, and our initial focus is on relationship initiation, so the process by which we start a relationship with customers.
In the last two years, we've provided four and a half million personalized insurance proposals, which increases our close rates versus control groups. Another key medium-term initiative is creating an integrated digital enterprise, which means we use technology, data, analytics, and importantly, process redesign to improve our effectiveness and efficiency. Our virtual estimating platform, QuickFoto Claim, replaces drive-in facilities and adjusters driving around in cars with digital operating centers that settle claims based on customer-generated photographs. Claims are now settled in hours instead of days and at a lower cost. The auto insurance business is also being strengthened through the use of telematics, which improves the customer value proposition. Allstate's approach is to use a continuous connection with customers through either OBD port devices or using their mobile phones.
This gives us the ability to provide customers with a more accurate price, and more than half the time, this means a lower price. In addition, we're actively developing a broad set of offerings to improve the driving experience. We were early into telematics, having initiated Drivewise in 2012 and investing heavily in building capabilities and are connected with about one million customers today. The same technology advances that enable us to use telematics to strengthen the auto insurance offering also provides us the chance to benefit from a restructuring of the personal transportation industry. There's a lot of discussion today about autonomous cars and the impact on auto insurance.
We like to say that autonomous cars are really just the tip of the iceberg, that the bigger part of that change is below the water, and there's a large economic opportunity for investors in the restructuring of the personal transportation industry. Today, the personal transportation system is expensive and highly inefficient. In the U.S., the total hardware value, that's cars and trucks, is $4 trillion. Direct cost, $2 trillion, and indirect cost of $1 trillion. Capacity utilization, even at peak hours, is only about a third. If this were a manufacturing plant or the manufacturing plant of a company you invest in, you'd tell them to redesign it, shut it down, and rebuild it with much higher capacity utilization and lower cost. The challenge, of course, is this is an economically and geographically dispersed system, so change is going to be a little harder.
That said, the opportunity is huge. Improving the efficiency of the system by just 20% would raise household income in the United States by 5% or $3,000 per year. That's going to necessitate industry shifts in car ownership, fleet size, number of parking lots, liability laws, safety regulations. It will go through that entire system of economic system will have to be changed. And it is going to be gradual, but the constituents that lean into this and change the business models and technology will be able to realize some of these gains that are delivered to customers. So we're working on how do we participate in that and leveraging our capabilities and customer relationships to participate in the opportunity.
Our strategy factors in these changes, not only defensively, so a lot of times we get asked about what's the impact on auto insurance, but also offensively, given our existing relationship with consumers and cars. In 2016, we created Arity, which provides hardware, software, and analytics to create value from connected cars. For insurers, this is a ready-made telematics platform that can improve pricing sophistication and basic connected services for their auto insurance customers. Arity serves Allstate and Esurance brands and is pursuing opportunities with other insurance carriers as well. We believe that the network effect can create value for all companies while enabling each individual insurance company to enhance their own individual competitive position by leveraging their business model and their capabilities. Arity is also building a suite of consumer offerings that will improve driving experiences. Connected car technology can also create value for governments and transportation-related businesses.
While this business is still in its infancy, based on our experience with 1 million connected customers, we believe it's a good risk and return proposition. So Allstate represents an attractive investment opportunity because we're proactive, we're disciplined, and focused on creating economic value for our shareholders, both current, medium, and long-term. That multi-horizon strategy of delivering near-term results, strengthening our businesses, and creating long-term growth platforms is consistent with our balanced approach to capital management. We're financially strong. We provide a dividend yield of 1.6% and have repurchased over 20% of our shares outstanding in the last three years, which equates to about $6 billion. With this context, why don't we have a conversation? Do you want me to get rid of this thing, Jay?
You might want to put up the.
With the logo up. I like the logo.
There we go.
Where do you want?
That's fine wherever you like. Thanks very much for the overview, Tom. That's really helpful. We're also joined by John Greig, who's Head of Investor Relations. With regard to the hurricane events, is there any additional insight you can provide around Harvey? I know on the slides you mentioned, you discussed Florida and Southeast, but just want to see if there's anything else you want to talk regarding Harvey.
I was down in Houston on Labor Day, visually it's a much different. I've been doing this for 23 years, I've flown over lots of hurricane sites. It's just a different kind of event from a hurricane. Typically, you'd fly over a hurricane event, and you'd see a lot of blue tarps. FEMA gives blue tarps out to people, they put them on their roof. I flew over Houston for an hour and 15 minutes. I saw three blue tarps. What that means is there's very little structural damage. Hurricanes cause damage from wind, storm surge, and rain. Wind, of course, knocks stuff down, trees down, blows out windows, takes off roofs and shingles. A storm surge is just the water coming blasting in from the ocean, there's rain. In this case, there's not really that much of the first two.
It's mostly just rain, the water comes up and the water goes down. You don't really have that much structural damage, which means the losses in total should be less. As it relates to us, of course, we don't do property insurance. We are a Write Your Own flood carrier, we adjust those claims on behalf of the federal government. We're working with our customers to make sure they get taken care of. The remediation should be relatively easy. What you do is you throw out all the wet stuff. You cut the drywall off at three to four feet, rip out the insulation, bring in dehumidifiers. It gets the water out, you put new drywall up, put a new floor down, you're done. That's different than having to put a roof on with trusses and trees falling on.
It's just much cheaper. The losses, I think, will be less, Jay, than a big windstorm would be. They are significant. Then, of course, we have auto losses. Most people buy comprehensive insurance. What's always hard to tell is whether people got their cars out of the way. I think the difference between Hurricane Harvey and Irma will be that in Harvey, some people got their cars out of the way. They'd park them up on, they'd put them on their neighbor's lot. I met one customer who put it up on their neighbor's house because they were two feet higher. I was like, "Yes, thank you for protecting your car." A lot of people didn't get out of Houston. I think in Irma, with mandatory evacuation, what we should expect to see is fewer cars as a percentage of cars out there.
It's too early to tell.
One of the controversies going back to Hurricane Katrina in 2005 was not only the wind damage, but also the storm surge.
Yeah.
Any thoughts on that?
Well-
With respect to Irma, given some of the footage?
Yeah. It's hard to tell since I'm going down there. I probably won't go down. I was going to go down Thursday, they're still kind of getting their act together. We're good at getting our act together, it just takes a couple of days. We had somebody drive from Jacksonville down over to Naples yesterday, and of course, we have people on the site. I think too early to tell. There's a lot of flooding in certain cities. There does appear to be a lot of downed trees. What we've done is one of the things that companies will have to contend with is those that are not prepared for this can't necessarily get adjusters. We have our own adjusters. We have great relationships with vendors and long-standing relationships.
Some of the carriers are paying extra now to get adjusters to come look at the stuff. What we did was we used, for our homes, we did satellite photography of the houses over the last couple of weeks, so we have the recent before pictures. We have our drone vendor is out in the air today taking pictures of houses so that if it's just shingle damage, we can adjust those claims with a computer. Just two pictures, put them right next to each other, adjuster can tell exactly what needs to be done. We're using technology to try to speed up the resolution process. It's cheaper too. You don't got to pay someone to go climb on a roof.
Right. Okay. Thanks for the disclosure on the reinsurance program. When I look at the Florida catastrophe reinsurance cover on the property side, and then also the Florida and Southeast auto reinsurance cover, the question that came to my mind was will that reinsurance protection be exhausted for Allstate from that storm?
Of course, it's hard to tell since we don't know what the losses are. They're each different. The losses would have to be bigger than we talked about here in terms of gross industry losses. Remember, we only have about 2% share. Whatever our number is, you could gross it up, times it by 50 before you'd get to go through it. That would be a pretty large property event for us to blow through $600 million of reinsurance. In the nationwide program, we have a reinstatement on if we want.
We can buy that by layer. That program is very complicated, very robust. It's working the way we thought it would work. I feel comfortable we have plenty of protection. I'm not worried about running out of reinsurance.
Okay. I appreciate that. The range of industry insured losses we're seeing out there seem to be extremely wide for both storms, Hurricane Harvey and Hurricane Irma. Have you given any thought to how large the industry insured losses might be for both storms?
Mostly we're working on our own, as you'd expect. What I can tell you from our standpoint, we don't have a number yet. Claim counts have started to go down in Hurricane Harvey. They kind of peaked at the end of the month, 28th, 29th, and 30th. They started to come down since then, down to, it's not small, but it's relatively few claims. We're getting a good handle on the number of claims. The next question is, well, how much, and how many are actually claims? We're figuring that out. The industry in Hurricane Harvey, I think the harder one to figure out is flood is covered by commercial policies. I think that's going to be a lot harder to figure out.
I'm not sure each company's different, what inventory they have on the floor, what they don't have on the floor, what they did at that particular moment to remediate any losses. I think that one's going to be harder to tell. On Hurricane Irma, it's of course, way early. I think there'll be more wind and storm surge in Hurricane Irma than Hurricane Harvey.
Okay. All right. Before I move on from this, anything else you'd like to add?
No, I think that's pretty straightforward.
I think we covered it. Okay. Moving on to the core business. Allstate is among the largest auto insurers in the U.S. The Allstate brand auto business has shown a sharp recovery in profitability, but policies in force are still declining. Can you talk about the prospects for further margin improvement in the auto business as well as a potential recovery in top-line growth?
Yeah, sure. Let me maybe talk about auto insurance first, and I'll do it by brand, and then we'll talk about growth. Auto insurance in total, we had this dramatic increase in frequency and severity that was 2015, 2016. As I talked about, we did everything we know how to do, and we're pretty good at making money in auto insurance, and it's a very strong muscle for the company. We did that. We're there. The Allstate brand is clearly in the right zone. Esurance and Encompass still have some room to go. I think they're making good progress, but I think their combined ratio should be lower than it is today. Allstate, the combined ratio is, I like where it's at. Could it be higher, and we still create shareholder value if it was higher?
Of course, it's way above what our cost of capital is. You'd say you'd like to grow that. Growth really becomes a series of, it's three factors. One is what you choose to do, second, how well you execute, and three, what the competitive situation is. I would say in the Allstate channel, we're now choosing to grow, so we have a very comprehensive program. The way in which we will drive that growth and execute it is, first, the easiest way to grow is by retaining more customers. By not raising prices by 6% or 7% a year, you chase less people off. They shop less. You should see our retention come up, Jay, that should help with growth. There's new business. You're starting to see our new business levels start to tick up.
There's expanding our distribution. That takes a little longer because we have to get our existing agencies to spend more money on marketing, and they have to hire more licensed sales professionals. We got to hire more agency owners, or we got to recruit them, and they got to buy out and rent offices, that takes a little bit longer to grow. I think what you saw in the first quarter is we were down 2.9%, I think, John?
First quarter.
Yeah. Second quarter was 2.6%. I don't think smaller decline is growth, let me be clear about that. We need to grow that business. We should grow that business. We can make more money for our shareholders growing that business. It's going to take a while for that to turn, because you got to get through the retention takes a couple of cycles with the renewal of your customers. It takes a while to expand your new business. It should go up, and start to go up from there. Esurance is easier to grow faster because.
The direct.
This is our direct. Yeah, right. It's easier to grow faster. It's a little behind in its getting its profitability improved. We're choosing to start to grow it, but a little more selectively than there. Encompass, we're about ready to grow the auto business, but we're not executing as well there. It's a $600 million premium business. It should be bigger. We should execute better. We need better comparative raters. We need better agency relationships. That one will take. I would say the Allstate brand will drive most of the growth. Esurance should have a higher percentage growth over the next couple of years. Encompass, I'm more interested in fixing the foundation than 10% growth in that $60 million of premium. It's important, but it's not going to drive the overall company. Then there's the competitive position.
You saw the losses from State Farm in auto insurance. State Farm's a good company. They're a sophisticated company. I would like to believe they're not going to continue to run that kind of underwriting loss. That should give us the opportunity to grow. One of the things that sometimes people, even inside the industry, they think growth is all about them. Sometimes the growth is all about everybody else, because it's not a high growth business, and it's not a high customer engagement piece. The fact that we shrunk over the last couple years has made it easier for other people to grow. I'm not taking anything away from them. They grew, they got the business. I'd like to have it. The challenge is now, if they have to do what we did, I think there's an opportunity for us to grow faster.
Good to hear. Allstate's homeowners insurance business has generated attractive margins, both including and excluding.
Yeah
catastrophe losses of third quarter is probably not that case, including cat. Would you expect that trend to continue on an underlying?
Yeah, we really like the business running an underlying combined ratio in around 60, which excludes catastrophes, just so everybody gets the math right. It's worked well for us. I talked about we've made a bunch of money over the last 5 years. We took that business from being a sometimes profit generator and sometimes big loss generator to being always making some money, sometimes a lot of money, sometimes less, but always good, and it gives us a good return on economic capital. We actually just lowered our economic capital requirements for that business. I think it could and should grow faster and does good returns for us.
The challenge for us there is we're risk and return focused. About 40% of the people live in harm's way along the coasts, and we're not crazy about growing along the coast, which has worked well for us, as you can see from our examples. Our share in Florida is 2%, and the reason our share is 2% is because we've made our share 2%. When you make your share 2%, you sell less auto insurance. We're good with that. We don't believe in subsidizing from line to line. I think there's less potential growth in the homeowner business than the auto business.
Okay. Before we go to the audience response system, can you update us on your capital management philosophy? Share buybacks and dividends?
Yeah. Maybe the headline, since other people have made announcements, we're not actively thinking about what we do with our share repurchase program as a result of the catastrophe. Maybe I just say that we're good. Part of that is because we look at.
That means the buyback stays in place.
Yeah, our $2 billion buyback program, we have no reason to be talking about changing it. I think it gets really to what your overall enterprise risk and return management is. We obviously have substantially reduced our risk for catastrophes, as you saw. We've also substantially de-levered the balance sheet by getting the Allstate Financial down from $75 billion to about half that in terms of assets, so just assets to equity. We also have a low debt-to-equity ratio. We feel good about the risk in the investment portfolio. We have lower interest rate risk than we used to have. We have a little more equity risk, we're comfortable with that.
You tie that together with our capital management program, the risk philosophy on that is we kind of do our share buyback knowing we could buy it back even if something bad happened in that year. We have $2.1 billion of cash at the holding company. We could buy shares back with that if we needed to. We could certainly handle more leverage in the balance sheet. We just don't think we need it. We're getting good returns anyway. I don't see any reason why we would stop doing it. It's because it all ties together, right? If we didn't have a lot of reinsurance cover, if we didn't have $3.9 billion, over $500 million of Florida covered, if we didn't have the investment portfolio, these things might interact differently. Because of the way we do it, I'm comfortable that what we have.
We tend to do our share buyback about one year in arrears. If you just look over time, the program is about what we could have generated. We generated last year, we kind of keep one year. If we don't earn anything this year, which I'm not thinking that's the issue because we have so much reinsurance, we would still be able to do what we want to do from a capital standpoint.
Great. All right. Buyback stays in place. Let's go to the audience response system. Okay, first question is, if you don't currently own shares of Allstate or happen to be underweight, what would cause you to change your mind? We can start the countdown.
I always feel we should have music going along.
five seconds to key in. The audience response, nearly half saying a lower valuation. Wow.
I'm not in favor of that one. How about that? Yeah, we do sometimes get the question of, geez, you've had this great run this year, you've had this great run over the last five years. Where does the additional value come from? First, you can look at that kind of valuation a couple ways. You can look at how much it's gone up and decide, oh, well, the train left, and I missed the train, and I want to get on it. You could also just look at the valuation and say, where's the valuation of our company relative to peers and other people in the industry? How much cash do we generate? What's been our track record of performance over the last 20 years? I would say on that basis, it looks cheap.
That's helpful. Next question. What is your confidence level in Allstate's ability to achieve a higher return on equity above 12% over the next several years? Of course, we're not taking into account the most recent catastrophe losses. Level of confidence. Year to date, ROE has been.
Last 12 months, 13 and change, depending what measure you want to put it on.
All right. A third saying high confidence, a little over 40% saying medium confidence. What's your perspective, Tom?
Well, I get paid to make sure that I think we have high and very high confidence. You can never tell what, when you're doing your projections, of course, nobody ever thinks things are going to go wrong. In 2015, I didn't think frequency was going to go up as much as it did, nor did I think severity was going to go up as much as it did in insurance. I think it depends what period of time you look at. What I can say is we know how to run our business. We're good at it. We're precise at it. When auto insurance margins got out of whack, we fixed it. We have a preference for return versus growth. We do not believe in subsidizing losses.
Some of our competitors have different approaches to that, and I'm not saying it's right or wrong, I'm just telling you what you get if you invest in us. We are highly focused on delivering returns for shareholders in terms of ROE.
From a directional standpoint, though, you're not saying that ROE should be headed lower over any reasonable-
No, I think one of the things that some investors are not factoring in is higher interest rates. We've shortened the duration in our investment portfolio, particularly the property liability portfolio, intentionally. We give up triple digits, millions in operating income as a result of that on purpose. Interest rates move 100, 200 basis points. We reposition that portfolio, that money goes right through the bottom line. That is, if you're looking for what else could be on the, first, I think we can deliver, and I think we'll build our business out well. If you're looking for some external event that would help, I think the market's undervaluing the potential upside from higher interest rates.
That certainly benefit from those realized gains would add to book value for sure, even if you're not.
Well, what we've done is we actually took some of those realized gains off the table, put them into, because we didn't think it was a good risk return trade-off. One of the things you should know about is we try to run the company for economics. When we're looking at risk return, it was about three years ago, interest rates were low. We liked all the gains we had in the portfolio, we said rates are more likely to go up than go down. We didn't want to try to call rates because I don't really know anybody over time who's made money calling rates. What you can do is say, from a risk return standpoint, it wasn't a great risk in return.
We decided to harvest those gains, sold those bonds, shortened up the duration, gave up over $100 million of investment income annually because we thought it was the right thing to do. Some people are hooked on operating EPS. That was not our gig. We decided not to do that. We still got a bunch of gains in that portfolio, we would lose some of those gains if interest rates spiked up. On the other hand, we would make so much more income, I believe it would be more than offset.
That's a great trade-off. Next question. What should Allstate pursue more of? We can start the clock. Acquisitions, divestitures, option 3 being share buybacks and dividend increases, none of the above.
I'm not sure. Is none of the above like go home or what? Maybe they have some other idea.
Okay. The audience-
I guess none of the above. Yeah.
Nearly half saying share buybacks and dividend increases, around a quarter saying acquisitions, and 13% saying divestitures.
Yeah, let me maybe talk about the smaller one, divestitures, just to tell you where our head is. Oftentimes, we get the question about Allstate Financial. If you look at our strategy in Allstate Financial over the last, I guess, decade really, John, that we've basically downsized it in a whole bunch of ways. We sold the VA business. We got out of the fixed annuity business. We've just been whittling it down. There's really a couple of businesses there. There's Allstate Benefits, which gives good high teens returns, 9% compound annual growth rate over 17 years at top line. We like that business. It's in the voluntary benefit business. Allstate Life Insurance, which works well with our Allstate agencies, and we get a sort of low teens return on that business.
The other business that we still have to sort out is what to do with the $12 billion block of payout annuities. These are structured settlements, people getting paid over 10, 20, 30 years. We have chosen, back to our philosophy again, to put the investments behind that block into equities. Because much like a pension fund, you want to be invested in equities. It's actually lower risk in return to be invested in equities versus bonds in that portfolio. That costs us a bunch more capital. I know a bunch is a technical term, but a lot more capital. The returns are crap. We have to figure out, we're constantly working on how can we do that? Some people have said, "Why don't you just sell it?
People like Athene and other people want to buy it." Our challenge is that with that $12 billion of policies, those are written on Allstate Life Insurance Company. It would be like me giving John $12 billion of cash, and I still owing Jay the $12 billion, and say, "John, hope you do a good job." There's investment manager risk there. It's not really credit risk, but it's investment manager risk. If we could find a way to mitigate that, we would try to do that. That would make Allstate Financial look like it is, which is really Allstate Benefits, because you've got great returns in Allstate Life, which has got decent returns, and fits well with our customer base.
Excellent. Do we have one more audience response question? All right. Final one. This is more of a longer-term question. Which statement do you agree with on the potential for long-term effect from autonomous, essentially driverless vehicles on auto insurers? We can start the countdown. Would it be favorable, unfavorable, immaterial, or too soon to tell? The impact of autonomous vehicles on auto insurers. Almost half saying too soon to tell. Not too much of a surprise, I guess.
Yeah.
evenly split, essentially, between people saying it's either favorable or unfavorable long term for auto insurers.
I would say, in part, that depends how you define yourself. If you define yourself as an auto insurer, and you're going to do business the way business is done today, I would say it's most likely going to be negative. If you define yourself as somebody who knows how to protect people from various ways, and you can sell them lots of different stuff, it has lots of potential and opportunity. Some of that's Arity, but also in the auto insurance business, we have a product today we're testing in two states called Milewise. Milewise is enabling you to buy insurance by the mile if you want. You don't have to buy for six months, you can buy it for 100 miles if you wanted to. That technology could be very useful in providing insurance to people when they share cars.
When you share cars, we're probably going to split the liability coverage from the asset coverage. Today, it goes together. You have to have insurance when you drive a car. If I'm going to let you use my car, I'm probably not going to let you use my liability insurance. I'm happy to take care of my hardware. I think there's more opportunity for people who are willing to use their capabilities and take the risk to make a difference.
Excellent. Well, please join me in thanking Tom Wilson and John Greig from Allstate. We'll have a breakout session in the Madison Suite. Next up is Athene. Great job. Thank you so much.