Allstate Corp (The) (ALL) — Q2 2017 Earnings Call Transcript

Thank you, Jonathan. Good morning and welcome, everyone, to Allstate's second quarter 2017 earnings conference call. After prepared remarks by our Chairman and CEO, Tom Wilson; Chief Financial Officer, Steve Shebik, and me, we will have a question-and-answer session. Also here are Matt Winter, our President; Don Civgin, the President of Emerging Businesses; John Dugenske, our new Chief Investment Officer; Mary Jane Fortin, President of Allstate Financial; and Eric Ferren, our Corporate Controller. Yesterday, following the close of the market, we issued our news release and investor supplement, filed our 10-Q for the second quarter, and posted the results presentation we will use this morning in conjunction with our prepared remarks. These documents are available on our website at allstateinvestors.com. As noted on the first slide, our discussion today will contain forward-looking statements about Allstate's operations. Allstate's results may differ materially from these statements, so please refer to our 10-K for 2016, the slides, and our most recent news release for information on potential risks. Also, this discussion will contain some non-GAAP measures, for which there are reconciliations in our news release or our investor supplement. We are recording this call, and a replay will be available following its conclusion. And, as always, I will be available to answer any follow-up questions you may have after the call. Now, I'll turn it over to Tom.

Well, good morning. Thank you as always for taking your time and investing with us and to understand the progress we’re making at Allstate. So let’s begin on Slide 2. Allstate delivered strong financial results in the second quarter. Net income was $550 million or $1.49 per share in the second quarter of 2017 and that is in comparison to $242 million last year; this really reflects improved auto insurance margins and strong investment results from our performance-based strategy. Operating income per share was $1.38 in the quarter. The improvement in auto insurance profitability is a result of us rapidly reacting to higher loss costs beginning in 2015 and it was aided by declining frequency in the first half of 2017. Investment income on our $81 billion portfolio also increased in prior year, as stable earnings from the quarter, market-based fixed income portfolio was supplemented with higher results from the performance-based portfolio. Operating income return on equity was 30.5%, as you can see at the bottom of the table, a significant improvement over last year. Go to Slide 3, which shows our operating priorities for 2017. We made excellent progress on the five operating priorities, but not all of this has yet impacted reported financial results. The first goal is to better serve customers, and we measure customer satisfaction by a net promoter score; there has been an increase for most of our businesses, although this has not yet led to higher policy retention. We do expect higher customer satisfaction to translate into higher growth, particularly as insurance auto price increases moderate. Allstate manages shareholder capital to deliver attractive returns which require us to achieve target economic returns on capital. As you can see from our results, we are excelling in this priority. Auto insurance margins have improved, reflecting the broad-based profit improvement plans initiated over two years ago. Profitability also benefited from a lower frequency of accidents, reflecting both our profit improvement plan and overall flattening in the market. The reported auto insurance combined ratio for all brands was 96.6 for the quarter and 95.8 for the Allstate brand. Auto insurance combined ratio for the first six months of the year was 94.1 or 6.5 below the prior year. The primary driver of the difference was an increase in that led to over $700 million in underwriting profit difference between these first two quarters and the first two quarters of last year. The homeowners insurance plan was also profitable in the quarter, led by the Allstate brand, which had a combined ratio of 97.2 despite $650 million of catastrophe losses. The property liability recorded combined ratio was 97.2 and the underlying combined ratio was 85.5 in the second quarter. Through the first half of the year, our underlying combined ratio was 85.1; assuming current loss trends continue, we expect to end 2017 at or below the low end of our annual outlook range of 87-89. Allstate Financial operating income increased to $153 million due to a strategy to increase performance-based investments in the annuity business, which generated good results. Long-term value creation also requires growth in the customer base. The acquisition of SquareTrade in January added over 31 million policies, and Steve will cover our key priorities with SquareTrade in a few minutes. Allstate Benefits continues with a 17-year track record of growth and policies in force exceed 4 million. The Allstate brand is now accelerating the trusted advisor initiative to raise growth by delivering a better value proposition as a local advice and branded product customer segment. A large component of operating income results from investment income, so despite the continuation of historically low interest rates, we’ve done quite well there as well. The portfolio is proactively managed and is primarily a high-quality fixed income portfolio which generates predictable earnings with modest growth. The value of the market-based portfolio increased this quarter due to reduction in corporate bond yields and higher equity values. The performance-based portfolio had a great quarter with strong growth in private equity and real estate earnings. We focus on delivering current results while investing for long-term growth. In addition to the trusted advice initiative, we have growth plans for Allstate Benefits, SquareTrade, Allstate Roadside, and Esurance. We're also investing in building the connected car platform at Arity, which continues to enhance its capabilities in telematics, data, analytics, and customer service. Slide 4 provides an overview of our capital strength and financial flexibility. As you can see from the back to the top, we delivered excellent returns, increased book value, maintained a conservative financial position while increasing shareholders' ownership in the Company by reducing the number of outstanding shares. We've returned $903 million to shareholders for the first six months of the year; that includes repurchasing 7.9 million shares of our common stock or 2.1% of those outstanding at the beginning of the year. Yesterday, we authorized a new $2 billion share repurchase program that will begin following the completion of our current $1.5 billion program. Our intention is to fund this new program through a combination of deployable capital, operating cash flow, and a potential issuance of preferred shares.

Thanks, Tom. Slide 5 shows property liability results by customer segment and brand. Starting with the table at the top, net written premium was $8.3 billion which was a 3% increase from the prior year and the recorded combined ratio of 97.2 was 3.6 points better than the prior year quarter. When we exclude catastrophes and prior year reserve re-estimates, the underlying combined ratio for the second quarter was 85.5, 3.1 points better than the prior year quarter. The underlying combined ratio for the second quarter includes 0.6 points or $52 million of restructuring expenses primarily related to the expansion of QuickFoto Claim, our virtual estimating platform. This expansion resulted in improved efficiencies and the closure of a number of claims drive-in offices. As you know, our strategy is to provide different customer value propositions for the four consumer segments of the property liability market. The Allstate brand in the lower left competes with the local advice and branded segments. We are the most prominent competitors of State Farm, Nationwide, and Farmers. Obviously, GEICO and Progressive target these customers as well, but do not offer the same value proposition provided by our 10,400 Allstate agencies. This segment comprises 90% of our total premiums written. The underlying combined ratio was 84.4, with the favorable prior year comparison being driven by improving loss trends in auto insurance, which had a 92.8 underlying combined ratio, 5 points below the prior year. Net written premium was 2.3% higher in the second quarter of 2017 compared to the prior year quarter due to a 3.3% increase in auto. Esurance in the lower right serves the customers who prefer a branded product but are comfortable handling their insurance needs. GEICO and Progressive Direct have a larger share of this segment than their overall market share. We continue to focus on improving the auto loss ratio, raising customer satisfaction, and rapidly growing homeowners policies in force. The combination of these initiatives will support long-term growth. The underlying combined ratio for auto insurance improves slightly and was below 100 for the second consecutive quarter. The homeowners business continues to grow rapidly with underlying profitability that reflects start-up costs. The underlying combined ratio of 125 in the second quarter of 2017 was significantly better than the prior year quarter as homeowners marketing spend was reduced. Encompass, in the upper left, competes for customers who want local advice and are less concerned about a branded experience and are served by independent agencies. We are making good progress in improving underlying margins, but the business has gotten smaller as we exit unprofitable markets and raise prices. As we achieve rate adequacy, we will initiate growth plans on a targeted basis in this segment.

Let’s go to Slide 6 to cover the results for Allstate brand auto insurance in more detail. Starting with the top left graph, the recorded combined ratio for the second quarter was 95.8, which was 5.4 points below the prior year quarter and benefited from increased average return premium, lower frequency, and favorable prior year reserve re-estimates primarily related to entry coverages. The underlying combined ratio of 92.8 in the second quarter of 2017 improved by 5 points compared to the second quarter of 2016, driven by a 5.7-point improvement in the underlying loss ratio. The chart on the top right shows the results of the broad-based profit improvement plan initiated in 2015. Annualized average premium shown by the blue line increased 5.6% to $999 compared to the prior year, while underlying loss and expense shown by the red line was nearly flat. This resulted in a favorable GAAP of $72 per policy compared to the mid-teens in the second quarter of 2015. While we continue to selectively file rate increases to keep pace with loss trends, the overall magnitude of rate increases will moderate if the gap between the red and blue lines is maintained. Gross frequency trends for bodily injury and property damage coverage are shown on the bottom chart. Frequency continued to show improvement across both coverages in the second quarter of 2017 and favorable trends were geographically widespread. The lower frequency in 2017 reflects good weather in the first quarter, the benefits of the auto insurance profit improvement plan, and moderating frequency trends across the industry. Slide 7 shows the underlying drivers of policies in force for Allstate branded auto insurance. As you can see from the graph at the top, overall policy counts have flattened down on a sequential quarter basis. This reflects an increase in new issued applications and a steady renewal ratio. We are beginning to accelerate the components of the trusted advisor initiative while expanding Allstate branded distribution with the objective of policies in force. This should be supported by fewer required price increases now that auto margins have improved. Slide 8 shows similar information for Allstate branded homeowners, which has had consistent profitability and is also being positioned for growth. Now, I’ll turn it over to Steve.

Thanks, John. Let’s go to Slide 9 and our investment results. Overall, investment results have been strong this year, reflecting favorable market conditions and the asset allocation decision to increase performance-based investments, which reflect the 10-year history of increasing commitments and building our capabilities. Today, we utilize third-party managers, co-investments through additional partnerships, and our own direct investing, and have created a broad portfolio of diversified investments. Total return in the upper left graph was 1.8% for the quarter, as our strategic positioning coupled with favorable market conditions, show strong results across our diversified portfolio. Investment income shown in the blue has consistently contributed approximately 1% of return per quarter with stable earnings from our market-based portfolio of primarily investment-grade fixed income investments. Total return varies based on the portfolio value at the end of each quarter as reflected by the valuation component shown in grey. As you can see, the value of the portfolio increased in the second quarter primarily due to lower corporate bond yields and higher equity prices. The Property-Liability bond portfolio, which totaled $32 billion, is concentrated in three to five-year maturities. If interest rates rise, bond valuations will be negatively impacted; however, net investment income will increase over time. Net investment income in total and for the market-based and performance-based portfolios is shown in the upper right graph. Net investment income for the second quarter was $897 million, $135 million higher than the second quarter of 2016. This increase was driven primarily by performance-based investment income of $263 million. While performance-based income is variable from quarter to quarter, long-term returns for Allstate have been attractive as shown by the bottom two graphs. The quarterly impact of performance-based net income has averaged $155 million over the last 10 quarters and is largely driven by private equity and real estate investment. The table beneath the chart on the bottom left shows the increase in maturing value of the performance-based portfolio over time. Our performance-based portfolio has generated attractive long-term economic returns, as shown in the bottom right. Internal rates of return are generally over 10%. The recent downturn in the 10-year measure reflects high valuations just prior to the global financial crisis of 2008 and 2009. Turning to Slide 10, Allstate Financial had a substantial increase in profitability as a result of the performance-based investment results. Premiums and contract charges totaled $591 million in the second quarter, an increase of 4.8% compared to the prior year quarter. Operating income of $153 million increased by 27.5% over the prior year quarter. Life insurance net income of $60 million and operating income of $63 million were both $1 million below prior year, as higher contract benefits and expenses were partially offset by higher premiums and net investment income. Allstate Benefits net and operating income were both $25 million in the second quarter of 2017. Operating income was $4 million below the prior year quarter as higher revenue was more than offset by increased contract benefits and investments in growth. Premiums and contract charges increased 7.2% compared to the prior year quarter, primarily related to growth in hospital indemnity, critical illness, short-term disability, and accident products. Annuities operating income of $65 million in the quarter was an increase of $38 million compared to the prior year, reflecting the continued benefit from our performance-based investment strategy. You remember we increased the amount of performance-based assets in this business to match the long duration of liabilities. This should generate increased shareholder value but does require us to utilize more capital in the business and lower interest income, both of which suppress short-term returns on capital. Slide 11 provides detail on SquareTrade. Last quarter, we indicated additional disclosures which we provided regarding this recently acquired business. SquareTrade has three primary objectives. First, to increase and broaden Allstate's customer relationships; this will be accomplished through the existing model of selling through store-based and online retailers, leveraging Allstate's other market-facing businesses, and entering new markets either from a product or geographic perspective. Second, SquareTrade is growing rapidly by utilizing innovative customer service approaches to reinvent a traditional product offering, and it'll continue to utilize this capability to enhance this competitive position. Third, as all of our businesses seek to do, SquareTrade will earn attractive returns on capital. We evaluate this by looking at long-term cash flows. For high-growth businesses such as SquareTrade, we use shorter term measures such as operating profit to ensure we are on pace to meet our long-term objectives, but we are willing to continue investing in growth even if it reduces near-term profitability. Moving to second quarter results as shown on the bottom hand of the page. SquareTrade has solid growth primarily through the U.S. retail channel, with policies in force increasing by 1.4 million from the first quarter to a total of 31.3 million, as shown in the graph in the bottom left. Over those last 12 months, policies in force have grown by 28%. Premiums written in the second quarter of $85 million, and both of the magnitude of product sales, while earned premium of $70 million reflects a recognition of that premium over the approximate three-year average duration of coverage. Underwriting loss totals $22 million in the second quarter, reflecting $23 million of amortization of purchase intangible assets related to the acquisition. Operating income, which excludes the amortization of purchase intangible assets, was positive in the second quarter, totaling $1 million. We also included a new non-GAAP measure this quarter called adjusted operating income to provide a run rate view of the business. This factor excludes purchase accounting adjustments made to recognize the acquired assets and liabilities at their fair value. During the second quarter, we executed a 100% quarter-share reinsurance agreement with our largest third-party insurer. As a result, it reduced the premium paid to that third-party insurer, which will increase underwriting income. Additionally, Allstate assumed approximately $200 million in funds held in trust for potential future claim payments. Invested income on these funds will be earned by SquareTrade in conjunction with this agreement; claims and claims expense benefited by a $6 million pretax favorable adjustment for loss experience. Slide 12 provides an overview of a new reporting structure that will expand our financial reporting segments from four to seven. We plan to adapt the new reporting structure in the fourth quarter. This instruction will provide enhanced transparency and operational valuation of our businesses grouped by like attributes. Allstate Protection will continue to include the traditional property liability businesses that address the four segments of the consumer property liability market: Allstate, Esurance, Encompass, and Answer Financial. A service business segment will include operations that have a larger portion of earnings from services and generally have less underwriting risk. This is likely to include SquareTrade, Arity, Allstate Roadside, and Allstate Dealer Services. Allstate Financial will be split into three segments. As you know, we will substantially reduce the breadth and size of Allstate Financial over the last decade. Allstate Life sells life insurance to Allstate agencies and supports the trusted advisor strategy to broaden customer relationships. This business earns a low double-digit return. Allstate Benefits with a high growth in mid-to-high-teens return business. So breaking these results out will highlight its value creation. Allstate Financial does not sell proprietary annuities given our view on economic returns. So the annuity is really a closed block of business. Returns are low in part reflecting current lower interest rates, in addition to the decision to maximize shareholder value by increasing allocation to performance-based investments; that had an additional negative impact on near-term return on capital. This new segmentation will have an impact on goodwill impairment testing and the irrigation of the Allstate Financial reserves with efficiency testing. More information is available in the Form-10Q, and we will provide additional detail later this year.

Now, I’ll ask Jonathan to open the line for your questions.

I just wanted to follow up. I have asked this before, but now that you have experienced two consecutive quarters with a significantly improving underlying combined ratio, I am interested in your thoughts on your competitive position in the marketplace, especially considering some reports we've seen about your competitors beginning to lower their pricing.

Greg, this is Tom. I'll start and then Matt can jump in. We believe we were proactive in implementing necessary price increases in auto insurance, which positions us well if these trends persist. We anticipate we won’t need to raise prices as significantly, which is beneficial since price is a key factor in our competitiveness, although not the only one. Our competitive strategy encompasses various aspects. The situation largely depends on the actions of other companies, and predicting their moves is quite challenging. Some large mutual companies facing significant underwriting losses may opt to remain in that position, limiting our opportunity to capture market share from those competitors. Other competitors might choose to offset losses in certain states by adjusting prices and maintaining higher margins elsewhere. We prefer our current stance, having improved profitability, which positions us well for growth. However, it’s important to note that just because others may not follow our lead doesn’t mean they will in the future, nor does it imply that it will be easy to acquire business. Matt, you might have additional insights.

Sure. Thanks, Greg, and it’s Matt. First, it all varies based on geographies, so we’ll start there. On a countrywide basis, as I said last quarter, we feel good that we have essentially caught up to the loss trends and are now within a more reactive mode where we’re monitoring loss trends and keeping pace with them. But that of course has some variability state-by-state, there are some states where we still have indications and we still need to take some rate. And there are others where we’re doing quite well and things have stabilized. Overall, we feel really good about our competitive position. In our business with our distribution model, one of the most important things for them is stability, and as we’re stable, as we maintain more normalized adjustments to rate and more inflationary rate adjustments, as we’re able to build multi-month and multi-year marketing plans and give them predictability, they are more willing to invest, and they’re able to do long-term business plans. We see increased quota activity and increased closing rates; as a result, increased new business production. So, we actually feel quite good about our competitive position; it's interesting that you point to rate reductions that you've seen. We still see lots of our competitors with very large indications that are not yet taken all of their indications in some states, and others they're still taking double-digit rates. So, we're seeing it all over the board while we are for the most part very stable in taking modest rate increases, which we believe is very good for our business model.

As a follow-up, with frequency, both BI and property, being favorable for you in the last two quarters I believe. Is there anything going on in the environment that might lead us to believe that that's going to continue?

Boy, if I could predict the future, Greg, I'd be a very happy man and Tom would be happier too. What we've seen is a moderation in frequency in the first half of this year; as we closed out last year, it was really unclear what frequency would look like. We were, as you know, monitoring miles driven, monitoring the unemployment rate, monitoring cash prices, and monitoring all those things that fed into miles driven, which fed into, in addition to distracted driving, fed into tax within frequency. And we've seen some moderation there which has been good for our business these first two quarters. And we have no way really projecting what is going to look like for the remainder of the year. We feel, as I said, good about the fact that it appears not to be making step function changes; that is, it appears to be moving in a more normalized basis up and down, normal variability with weather and things like that, which is much easier for us to react to and monitor and take appropriate rates for.

As you begin to transition to growth, should we expect the underlying combined ratio to stay around these levels, or should we start to see it rise a bit given maybe a new business penalty or less price action?

When we manage overall profitability, we look at the underlying combined ratio, which is lower than we expected this year. We have also considered the reported combined ratio, taking into account catastrophes and other factors. We're satisfied with our current standing. While it could increase, frequency and severity typically fluctuate by more than 1% annually. In response to your question about our goal to further reduce it, the answer is no. It might decrease, depending on market conditions, but we focus on balancing the combined ratio with growth. In 2015, when frequency and severity rose faster than anticipated in our pricing, we prioritized profitability over growth. I believe we are currently in a more stable position. Matt, do you have anything to add?

Yes, let me just add. This is Matt. Let me add just two items for you. You mentioned the new business penalty and whether or not, as we begin growing, whether or not that could influence the combined ratio. Certainly, the new business penalty is a fact of life that as you grow fast, the newer businesses not yet tenured; they tend to have a higher loss ratio than the more tenured business. One of these things, however, that mitigates against that is that in our reaction to the frequency spreads in the last two years, we took very segmented rate actions, and we took segmented rate actions against those worst performing segments of the book. As a result, the quality of our book increased. That led to more defections from the worst performing segments. It also led to us adding fewer new customers from those lower performing segments. So, we believe that the new business penalty will be dampened as we add on, assuming that we maintain the same high quality we’ve been seeing lately. So I don’t expect that to be a significant drag. The other thing I would point out is that there has been some commentary about whether or not this will cause us to dramatically increase marketing spend to stimulate growth. Really, the marketing spend is not the primary driver of growth right now, as we've discussed many times. Retention was actually a bigger influence for us than the new business on our total items in force. So, we are just as focused on trying to improve our retention within those balances of what we can control since a lot of it is uncontrollable, as it's based upon competitor actions. But for those things that we can influence, such as customer satisfaction and customer engagements and stability of pricing, we're just as focused on the retention component as we are on new business.

And then secondly, I was wondering if you could talk a little bit about your QuickFoto Claim and other claims initiatives and how we should think about the potential for savings there, as well as if there should be further restructuring costs? I think your LAE ratio is running around 11 points, but would it be fair to say maybe over time you could shave a couple of points off of that?

Sure, it's Matt again. So let me just start high first. So, we've talked about integrated digital enterprise and emerging technologies; one of those areas that we are deploying it initially and in quite some force is the claims area. We believe it's an area of prime for it because there it is some inefficiency in the way the model operated. In the past, there was a lot of windshield time; there was a lot of dead time, unproductive time as adjusters drove around, driving to cars, driving to body shops for both initial estimates and supplements. We looked at that and realized that emerging technologies, data, and analytics could rectify that and take some of the inefficiency out of the system. So we began the digitalization of the claims process. Last year, we launched a new immediate payment method which we called Quick Card Pay. Quick Card Pay is the fastest claim payment method in the P&C industry. We make payments directly to the debit cards within seconds. We’ve also been assessing roof damage from hail events utilizing drones, and we’ve been doing that quite successfully for the last several quarters. In the last quarter, we opened two digital operating centers. They handle auto claims on a countrywide basis by estimating through photos. Approximately half right now of all drivable vehicles are currently being inspected through our Quick Photo method of settlement. That has led to the shutdown of many of our driving claim centers and has also led to a reduced need for fuel adjusters, since we took a lot of that inefficiency out of the system. We now have our adjusters looking at enhanced photos, digital photos on the computers without having to drive to the sites. We’ve also begun utilizing video chat technology to review supplemental damage with auto body shops; it’s something we call virtual assist. The combination of all those things has led to a dramatically more productive and more efficient claim system. We've taken a cycle that used to take five to seven days in order to get eyes on a vehicle and get an estimate; we’ve done that now in hours, and we are literally doing that in under claim for an hour. So for supplementary, instead of having to schedule when adjusted from the back out to the body shop and look at supplemental damage, we now use video chat technology same day and we move it along; everybody's happy either the customer gets the car back sooner, the body shop gets the car up sooner, and they can take another car in. We believe this leads to a more efficient system, leads to cost savings as we take inefficiencies out of the system, and leads to greater customer satisfaction.

Sarah, this is Tom. I think you should revisit the ULAE, as your number appears to be higher than expected. I'm not sure how you're analyzing it, but you can check with John for clarification. Regarding Matt's Virtual Assist, this technology is excellent and we are offering it to other insurance companies. Essentially, as Matt explained, it's like FaceTime integrated with your claims system. It allows you to assess the situation without someone having to drive to the body shop for a supplement; this can all be done remotely. We are providing this to Arity, so if insurance companies want to adopt it, we have a competitive advantage since we don’t need to train a lot of body shops on this technology. We have already trained body shops and implemented it. Therefore, we are making this available to other insurance companies. That's my pitch for Arity.

Okay. And can you quantify the savings you've seen from these initiatives at all?

We will not specify the exact savings. When we took the $52 million charge, not all of it was related to claim fees; some of it was for restructuring in the legal department. However, we do expect to earn that back in a relatively short time frame, though it’s not five years or just five days. The savings are significant in absolute dollar terms, even though they may seem smaller when viewed as a percentage. We will outline the restructuring fees by component in the quarterly report.

The Allstate brand underlying combined ratio in auto clearly improved year over year, although it was higher quarter over quarter. I'm just wondering if there is some seasonality that would explain that or if there is some other cause?

That would make anything of it, Jay. It bounces around. You'll remember the first quarter; January and February were much lower, so really better look at it versus the prior year quarter because of weather patterns.

I thought there was some seasonality there. Has Allstate considered purchasing increased catastrophe reinsurance protection for the working layer catastrophes? It appears that the Company is projected to have about 8 points of catastrophe losses over the past two years. With reinsurance rates decreasing, I am curious if there might be an opportunity for increased risk transfer.

I'll start with some general observations, and then Steve may chime in. First, we consistently assess reinsurance on a broad scale and have contributed to the development of markets, including cat bonds and larger, long-term aggregates that have been established over the past three years, which were previously unavailable. We are continuously active in the market as we are one of the largest buyers, and people respond to our inquiries. In response to your question about using reinsurance to mitigate the annual impact of smaller events such as hail and windstorms, we believe that we can manage that volatility within our profit and loss statements. Over the last year, we have generated over $1 billion, and costs rise when considering higher-end coverage for rare events, which can serve to alleviate the impact of those events on our financials but also acts as a tool for relieving capital pressures. We are willing to pay a slightly higher rate because we believe that we can achieve a greater return on equity than reinsurers can, given their portfolio diversification compared to our specific needs. We remain focused on exploring various methods to optimize our capital allocation while maximizing returns. Steve, do you have anything to add?

I think you did a pretty good job, Tom. The only couple of things that I might add: We do look not only at buying more but how we structure the reinsurance. And so we look at annually; it’s kind of an annual buy for Allstate in the early part of the year, and for Florida, it’s going to be mid-part of the year. So if you look particularly this year, we did buy a $200 million cover in the Southeast for auto, which kind of filled in the hole we thought we had, in terms of bringing down retention dollars for that area. So we continue to look not only at how it is structured, but there are holes we think might be if a storm were to hit; we focus on that. And as Tom said, it's really economics. And generally what we've seen over the last ample of years, it hasn't affected economics for us at that time to buy more or to structure differently. But we are entering into what we consider our normal type period and probably when we look at so we look at again given your correct that insurance rates have continued to move down.

As you guys talk about looking to grow in auto and some of your comments you also did mention in home. Can you just remind us about the bundling within your book, those that purchase both auto and home coverages? And how you expect, as you look to improve the retention in auto, how you expect that to play out in home?

Matt will answer for the Allstate brand and then Don will make a comment on Esurance, because I think there is one of the analyst write-ups last night; there was a comment about Esurance that makes sense.

Sure, Elyse. It's Matt. So let me talk about the Allstate brands. The last couple of years have been bundling capabilities, and we have bundled primarily with auto and another product, auto and home, auto and life, auto and consumer household products. With all the disruption in the auto business, it has influenced and impacted the potential growth of the other business. We know that we watch the home area specifically, and we know that home is lagging auto, not only because of the 12-month policy in home but the different renewal periods. As a result, you tend to see a lag in the growth. So when auto kicks up, it tends to be a quarter or two before home takes up. And so we're still seeing some of the influence from that disruption in home even though auto has begun to stabilize internally. Our expectation is fully that as this works its way through the system, and I remind you that a lot of growth was taken just over 12 months ago, has now worked its way through the system as this stabilizes, and as we hope retention stabilizes and potentially improves, we expect to see our capacity for bundling to improve. We like that; obviously number one we want to serve customers holistically, we think it's 100% consistent with our trusted advisors strategy, but it also helps to leverage the single acquisition cost across multiple product lines, leading to a more efficient system and a more efficient use of marketing funds. So as you should expect to see continued emphasis on the part of the Allstate brand to increase the bundling, increase the number of products sold for the household, and increasingly meet the needs of our customers.

Much of what Matt mentioned about the Allstate brand also applies to the Esurance brand and other brands across the industry. Various companies might create bundles for their customers. We acquired Esurance six years ago as a company focused solely on auto insurance. We recognized the potential to market homeowners insurance to establish a different kind of relationship with various customer segments. Currently, we don't offer homeowners insurance as a corporation, but we've been able to utilize the expertise we have at Allstate to enhance Esurance's capabilities. Esurance now provides homeowners insurance in 31 states, and bundling remains crucial for us for the same reasons Matt discussed regarding Allstate. Our attach rates are significantly increasing as we roll out the product and make it accessible to customers. This will allow us to improve retention, build a different customer relationship, and truly expand the business. However, I want to be cautious here; the net return premium for the second quarter is only $20 million with 69,000 policies, which is still a relatively small segment of our portfolio, and there can be fluctuations in that number due to the impact of large losses on the loss ratio, as we saw in the second quarter. Nevertheless, we are pursuing this for strategic reasons, and it is yielding the results we anticipated. I believe this presents a substantial opportunity for further growth of Esurance.

Okay, great. And then in terms of capital return, you mentioned the new $2 billion buyback program and being financed with capital at hand in your earnings. How do you view M&A right now? I mean you just completed the SquareTrade deal, but any kind of high-level thoughts on how you think M&A is still part of the equation as you think about capital return here?

Well, of course, the thing we do is look at managing the overall capital, and we’d like to invest in our existing businesses to the extent that we can, which is organic growth that obviously leverages our real capabilities for higher returns. We do look at acquisitions, and as you mentioned, we brought SquareTrade to broaden our product portfolio, broaden our distribution we have, and give us additional products to be able to sell through the trusted advisor issue through Esurance. And so if you look at them, what we do after that, so if we don’t have an additional use of the money, we should return it to the shareholder and help increase their relative ownership in the Company. So in the last three years, we bought back 15% of the shares outstanding. So, your value should grow up by 15% just because of the return of capital to shareholders over five years of spend, a little over 25%. So, I would expect that same pattern to continue; if we see something interesting, we'll buy it. We paid a lot of money for SquareTrade, but we believe we can grow just as we did Esurance and Allstate Benefits.

First question to follow up on digital claim processing technology; I just wonder how do you measure the accuracy using photos to settle the initial loss reserve versus using adjusters in person?

Hi, it's Matt. That's a good question. Obviously, when you initiate a massive change in process such as this, the quality is one of the most critical things, so we did a lot of testing along the way. We still have testing that we do both secondary reviews as well as reviews in person. So we do selectively have people doing onsite reviews. As I said in the summary, 50% of drivable cars right now are being inspected through our Quick Photo method of settlement. And the reality is that quality has been exceptional; supplements are slightly higher but rather an insignificant amount, and the overall quality is quite good. We are continuing to develop our ability to enhance the digital photos and get better visibility into the damage. Obviously, as our adjustors get more familiar with how to use these photos and what angles to request and how to change the lighting on it, we're able to pick up more-and-more information; we're using some technology to actually help us learn from that and compare photos from similar autos and similar types of accidents to help us baseline. So overall, the quality has been quite good, actually slightly better than we expected, and the productivity and efficiency savings have been tremendous.

Kai, the reserves, remember these are relatively short-tailed. So within 90 days, you know what it costs to fix a car. So it cycles its way through, and we think we are good there.

Okay, that's great. By the way, how much do you spend on these R&D investments? It's a part of your expense ratio, right?

I won't give you a number or percentage. I will just tell you that we always look at these things and expense. We invest heavily in research and development, whether that be the things we talked about here or Arity for other things, and we believe we can handle with the overall P&L. So we do not resource constrain that. Capabilities and ability to execute might constrain, but it is not a money constraint.

My second question is on your sort of like the change of reporting structure. Is there any change in the underlying operating structure? And how big do you think the service business will become as a sort of meaningful percentage of the overall business of the Company?

The operating structure has been adjusted to align with our capital allocation and to enhance transparency. We do not have a specific revenue or profit goal for any of these businesses. For instance, in Allstate Financial, we present many underlying numbers. When Mary Jane oversees the overall business, she doesn't focus solely on the overall return on equity, which might appear to be around 6%, 7%, or 8% for the quarter. Allstate Benefits and Allstate Life have strong ROEs, but the annuities business pulls the overall figure down. We believed it was necessary to present this information differently. While some of this data can be gathered from our quarterly reports or investor supplements, our approach provides greater transparency by aggregating it in a manner that reflects how we manage the business.

Allstate. I'm going to apologize in advance because this is going to get in the weeds a little bit, but between Sarah and Kai's questions, I'm still having trouble understanding the expenses in technology versus what's in the expense ratio and what's not. Can we talk a little about why the $52 million maybe shows up in the expense ratio? Typically, you guys have like $10 million or so of expenses that are other that show up. How should we think about it going forward? And normally, just to understand, you guys are always investing in the future. Why does this show up in the expense ratio, and other stuff in the past has not?

It's important to clarify that this isn't just about technology expenses. The costs include severance for over 500 employees who are no longer with us, and we currently have 937 drive-ins. Many of these were structured with month-to-month leases in anticipation of these changes. However, we still need to account for the costs of shutting down operations and removing leasehold improvements. This situation is not directly linked to our investments; rather, it pertains to winding down other activities. The question about why this appears in the underlying combined ratio is something Matt and I have been discussing with Steve for the past three weeks. Restructuring charges have consistently been included in our calculations, so you should expect to see them moving forward. Even though these charges may not reflect ongoing expenses in our underlying metrics, I don't foresee us taking a $52 million charge for the quarter. Nonetheless, if we do need to recognize any charges, we will manage them in relation to cash flow and economic returns. If these charges impact the income statement, they will be reflected there.

Okay, that's perfect. I am on the same page now. And then the limited partnerships, I mean I'm not going to get a great answer out of it, but they were phenomenal this quarter. Were there any specific gains taken that were unusual, or was it just a great quarter for mark-to-markets? How should we think about this going forward?

John will give you that perspective.

Thanks, Josh, this is John. It was a combination of factors. One, we have been building up the book of business you expect that the return on the book of business would get larger. But two, favorable markets influence when you look at our performance-based assets that you have a correlation of about 70% to public markets, so that was a factor there. And then three, to answer your question specifically, there were a few idiosyncratic properties that performed quite well, and credit to the team that sifts through many opportunities; there were some down to the one that make most sense for this firm and investment.

If you need additional co-investment, just let me know; I will give it a try.

If I get good returns, right. Jonathan, we have time for one more question.

Thank you for fitting me in. I have a question regarding your net investment income on page 51. I noticed that your fixed income yields have increased year-over-year, even though you have shortened maturities. How have you managed to maintain the yield on fixed incomes in a low interest rate environment with shorter maturities?

Hi, this is John again. As you would imagine, the combination of events can cause changes in yields; not only where you are placed on the yield curve, but also what investments you buy in the market. So part of that is in response to increased holding in securities like high yield and other higher yielding securities. We also did, and it should be noted while it is a small move, in the first quarter of this year, we did extend the duration of our portfolio by about a quarter of the year, and that impacted yields by about 20 basis points.

Got you. And one other quickie. On the preferred that you say you're going to issue to help fund the buyback, what's the motivation there? What roughly yield and magnitude are we talking?

So, we’re looking at part of the potential funding for our buyback.

Right.

So, I think currently right now, the market seems fairly attractive, and this kind of goes on cycles. And so we will look at that as part of that financial; obviously, we don’t have to do it, because we have plenty of capital and cash. But when opportunities are there, we think our track available, we want to take advantage of that.

Rough idea of the size and yield that you would be paying? Ballpark?

Well, the yield should be as low as possible from our standpoint.

Right.

You saw a deal went out earlier this week and a couple of ticks over five. And so when you can issue perpetual equity that guarantees a 5% return, you can use that as buyback capital; we think that’s a good trade on behalf of shareholders. And as you know, we’ve done a billion in three quarters of it over the last, Steve, two or three years.

Four years.

Four years, and we are pleased with the results of that. It's important to note that we wouldn't pursue this for a smaller amount; it wouldn't justify the overall effort. While we haven't specifically quantified it, there are numerous opportunities for us to issue given our strong brand, credit capacity, and associated payment options. Let me close by, again, thank you all for participating. Overall, we made excellent progress on our five operating priorities, most importantly better serving the customers, achieving economic returns, and capital employ really managing our investments. We also are beginning to focus more on growing the customer base and then always again building long-term growth platforms, whether that's investing in integrated digital, enterprise, or new initiatives like SquareTrade and Arity. Looking forward, we’re just going to stay focused on the 2017 priorities and be precise in the way we execute our business, balancing both short- and long-term initiatives. So, thank you all, and we’ll see you in the next quarter.