I believe so.
I probably say this every single time, except for the first time, this is always one of the easiest companies to introduce. I am happy always to introduce them. It is Selective Insurance Group. We have Greg Murphy, the CEO, and Mark Wilcox, the CFO, with us. One of the reasons it is always so easy is while the industry evolves, and even the company evolves, the main aspects of the story, or the cornerstones, pretty much stay relatively focused and consistent. One of them being the commitment to the agency channel and how important it is to their growth and their strategy. As always, it is always best to turn the podium over to them to discuss it. All yours.
Thank you.
I did it at 3:30 P.M., so we are good.
Perfect. Thank you very much. Go ahead. You hop down. All right. It's great to be here this afternoon. Hopefully, we can keep your attention for 30 minutes to just give you the Selective story in terms of what we're doing in the marketplace. To let you know, Selective is the 36th largest property casualty company in the U.S. We have currently a 25-state footprint. We've got a 90-year history. We did $2.4 billion in premium last year. That was growing at 6%, well above the industry expectation. Our market cap, just to give you some sense, is $3.4 billion. Overall, our combined ratio was an excellent, on a GAAP basis, 93.3%. I'll stack that up against any presenters that were here today or that will be here tomorrow.
This seems like it's cutting in and out, or is it just me? It's good? After-tax investment income at $190 million, up 20%. Our non-GAAP ROE, one of the very few companies that printed an ROE north of 10%, did an 11.4% and met the target, or in line with the target of 300 over our weighted average cost of capital. That's a goal that we do not change year to year just because it's more difficult to reach. When you think about our sustainable competitive advantages, there are really four, I'm going to go into them in a little bit more depth. Number one is true franchise value with Ivy League independent agents.
We have 1,250 agents that represent approximately 50 agencies per state, when you do the math, in the number of states that we're at 25. Our agencies score us an 8.8 on a 10-point scale in terms of customer service and what we do for them. The second main advantage that we have is our unique field model. I'm going to get into that in a lot of depth in a minute, it is truly very different than any other model that you would see in the industry. Omni-channel, in terms of how we're pushing our agents to think about the future of the business, is a very big lift for us internally in terms of the resources that we're spending on that.
I think that's absolutely critical to separate yourself from, in my mind, the disruptors that will start to appear in the market as we move forward. Some of our above-average return comes from the fact of our leverage, our leverage is very simple to understand. Every one point of combined ratio is 110 basis points of ROE. Our investment portfolio, that's about $3.32 of invested assets per dollar, yields an ROE of right around 8% overall. Good company in terms of our leverage capability. Here's the geo footprint of Selective, the 25 states that we represent us for commercial lines, which represents 78% of our overall premium. We are in 13 states from a personal lines standpoint. You notice that we try to stay out of states that we view as non-regulatory friendly states in terms of personal lines distribution.
The E&S, we're pretty much everywhere from an E&S standpoint as well as flood. We have about 80 partners on the wholesale side on our E&S business. When you think about our sustainable advantage of our field model, they are the linchpin or the centerpiece of how we conduct our agency relations. Generally, we have a 10-to-1 relationship of field underwriters to agents. As a field underwriter, I'm responsible for that agency's growth, profitability, and all of their new business. I have underwriting authority within a certain level. Things outside my band of control need to be kicked up the line. Generally, I'm in your office 1 day, 2 days, 3 days a week, looking at opportunities, going out and seeing business overall, helping you as an agent write new business so you can grow as an agency, and we grow as a company.
That center then is around also our field safety management people. We do safety management right down to smaller-sized accounts. We have field claim people. Sometimes a producer that goes out, "Hey, I want to sell this account," and we would send the field claim person with them to go out and go on the prospect, say, "I'm going to be handling your claims. I'm a local person." That's an important part of the relationship sale. They're also surrounded by small business people that are in the regional offices but are there to support the small business issues that come up on the agency side, so we can turn it around where speed is absolutely critical from that end of the perspective. When you think about where we are on omni-channel, I think this is a critical investment for us.
We spent 2017 as our base year, building out master data management, getting the golden record, tying together all of our customers. Whether it's a commercial business to a personal insured, seeing all of the policies, now we've got that underway, and we're surrounding that now with our CRM system. We can see all of the interactions that a customer has had, whether it's on billing, whether it's something that happened in their commercial policy, whether it's something in their personal policy, we can surround that. Ultimately, our goal is to build out that capability to where our agencies can see into that presentation panel and add to that the information that they want to add into that, about that account.
When we sit there, no matter however and whenever a customer wants to interact, whether it's text, telephony, in person, whatever, they can contact, and we can all see and share the same information going forward. We feel this is absolutely critical as we move forward as an organization, because we've got to be able to respond in a 24-hour-a-day environment. Because of our franchise value, we have a unique opportunity to be able to deploy that capability throughout our agency plan. Ultimately, the goal is to increase switching costs, become much more proactive in communication, increase our retention, both as an organization, but also to get our agency's retention higher as a result of ongoing proactive communication and other things that we're doing with our agencies, with their end customers.
When you think about our performance for the 2017 year, I think the best thing to walk away is, whether you're looking at 2015, 2016, 2017, all fairly similar performance. Combined ratio overall was a 93.3 on a GAAP basis. Very stable, but that's being driven by underlying price increases and a number of things that we're doing from an underwriting standpoint. Great core performance and our forecast for the 2018 year, one year ahead of myself already, is for a 91 combined ratio on a GAAP basis, and that's underlying. Our expectation for cat losses is about three-and-a-half points. All in, that would give you your GAAP combined ratio in total. When you think about growth rate and opportunity, we look at it, and we scale it in terms of risk.
The least risky way to grow is to grow in your existing footprint states, continue to increase what we call share of wallet with our agents. Our target share of wallet is 12%. That means for all the business that you produce as an agency for commercial lines, we target, we want to have a 12 share of that wallet, and then we measure and monitor that, and it's the AMSs, our field underwriters' responsibility to get that growth rate to get to that 12 share of wallet. If you think about that overall, just getting to a 12 share of wallet would be something around a billion and a half dollar premium opportunity for us in the 25 states. We're going to continue to add new agencies in the existing states that we're in, and that's about another $1 billion or so opportunity.
That's the next most least risky way to grow. We're going to continue to build out our geo footprint, not totally, but we've got about 30 states overall that we're going to target. Looking about that geo footprint, you get a sense of where we are and our Southwest and ultimately our Northwest expansion. We have Colorado opened 1/1 of this year. Arizona's been a great expansion state for us. We just also added New Hampshire to the inventory. I'll tell you, among Arizona, Colorado, and New Hampshire, we have 37 agents, and those 37 agents that we have already represent about 20% of the premium available in those states. That's the kind of concentrated greenfield strategy that we have as an organization. Commercial lines, 78% of the business, very profitable overall, continuing to drive it with rate.
This is how we've been pushing our rate through overall and extremely granular basis. This slide shows you from the best cohort to the worst cohort, and you can see that the size of the bars are the amount of rate that we're driving, and then the blue squares coming down is the retention of point of renewal. This is exactly what you want to see, high retention, lower premium at the above average, trailing down to bigger increases with lower retention at the worst end of the cohort. If you socialize rate in this marketplace, you're going to get hurt. This is our graph going all the way back to 2009. This is pure renewal premium, not renewal premium, pure renewal premium in terms of us versus the CLIPS index.
You can see how we've outperformed, and this is why our combined ratio relative to the industry is so much better. On the personal lines side, the only thing I can say is that we continue to drive our performance. Home is where we want it to be, which is about a 90 combined ratio on a normal cat year. Auto is still trailing down, have a little bit more work to do on that side, and have been aggressively managing the auto side. E&S is an area that we target our returns on and then let the top line float up or down based on market conditions, and we're willing to shrink that business if we're not reaching our goals overall. Now I'm going to turn it over to Mark, who's going to run you through some of the financial results. Thank you.
Great. Thank you, Greg, and good afternoon, everybody. I'll start with just an overview of our financial targets and objectives. Greg sort of teed it up, but would like to just provide a little bit more color. When you think about Selective, we're very public and transparent about our financial targets and objectives. It's 300 basis points over our weighted average cost of capital. Going into 2017, that was a target return for us of 11.5%. Going into 2018, we've seen an increase in our weighted average cost of capital. That's a function of a higher market capitalization for Selective, as well as a higher after-tax cost of debt, given the lower federal tax rate in the U.S. That rate, our weighted average cost of capital now is 9%, so the target for 2018 is 12%.
When you think about Selective and our planning process and compensation, it's all built around trying to achieve our target returns over the long term. We believe that if we generate our target returns over the long term, we're going to generate good growth in book value per share. More specifically, for good stewards of capital, we believe we can compound tangible book value per share plus accumulated dividends at a higher rate than our peer group. Over time, we'd expect that to generate superior total shareholder return. From a business perspective, Greg walked you through our footprint, the lines of business that we're in. Just spending a couple minutes on our business and operating model is, I think, time well spent.
When you think about Selective, if you focus on the chart on the right-hand side of the slide there and look at the bullseye, that really represents Selective's underwriting risk appetite. We're in the business of accumulating a portfolio of risks that has a lower to medium hazard type of risk. We're in the small to medium commercial space in standard commercial lines, in standard personal lines. We're not writing high-value homes. When we're in E&S, we often refer to it as E&S light. We generally have an average premium size or account size of about $3,000. 98%, 99% of the policy limits are $1 million or less. Our business starts with a lower risk profile, which we believe will result in lower volatility in our results over time. We back that up with a conservative approach to managing the balance sheet.
When you think about the investment portfolio, reinsurance, and a disciplined approach to reserving, it's a conservative approach to managing our capital and our balance sheet. Because of that, sort of what we would consider low risk profile, we take on more operating leverage than the industry as a whole. If you were to look across personal and standard commercial lines in the U.S., you'd see that the average company, the industry average is about a 0.7 to 0.8 times net premiums written to surplus, and Selective is double that. We're at 1.4 times. We have a target range, 1.4-1.6, that we're comfortable writing our business from a premium to surplus perspective. Today, we're kind of at the low end of that range. As Greg mentioned earlier on, for us, every point of underwriting margin is 110 basis points of operating ROE.
If you compare that to the industry as a whole, they'd have to have twice the underwriting margin to generate the same operating ROE, given the leverage for the industry compared to Selective. Same holds true from an invested assets to dollar of surplus perspective with $3.32 per invested assets to dollar of surplus. That's higher than the industry. For us, for every 100 basis points of pre-tax book yield, it generates 275 basis points of operating ROE for Selective. On the bottom of the page, you can see our financial strength ratings, which is sort of the price of admission for our business. From an investment perspective, just drilling down a little bit more into the investment portfolio, we believe we have a conservative portfolio. 95% of it is in fixed income and short-term investments.
AA- average credit quality, 3.7 years is the effective duration. Within that, there's an allocation to high yield. It's been running about 3%. When you add to that allocation to high yield, our equity allocation of 3% and alternatives of 2, we have about an 8% allocation to what we consider risk assets. That is up a little bit from where we were a year ago. We're about a 7% allocation to risk assets. We have a target of 10%. Risk assets are pretty expensive at the moment. We'd like to get to 10%, but we're going to take our time and let the market dictate the time period it takes us to get there. From a cat perspective, sometimes you can get lucky with cat losses, sometimes you get unlucky.
I think when you look at our results over a 15-year period, you can see we've outperformed the industry from a cat perspective. We've averaged, over this time horizon, 2.8 points on the combined ratio versus the industry of 5. If you go back to some of the heavy cat years, look at 2004, 2005, 2008, 2017 from a U.S. perspective, where you saw those big U.S. hurricanes making landfall, you can see our cat loss ratio much, much lower than the industry as a whole. That's a function of where we decide to deploy our capital, where we decide to write business. Greg took you through the geo expansion and our footprint. You don't see us writing standard commercial lines or personal lines in states like Florida, Texas, and California. We've largely avoided some of those cat losses.
Sometimes the CATs do come in an area that you're concentrated. If you look at 2011 and 2012, we had some events like Hurricane Irene and Hurricane Sandy and some Severe Convective Storms in 2011 that elevated the CAT loss ratio. Over that time series, you can see we're well below the industry as a whole. Part of the strategy, as I mentioned earlier, is the reinsurance buying. To take some of the sting out of the CAT losses, we are big buyers of reinsurance. We have a core CAT program that we renewed at like 1/1. We're a loss-free account. We had a very, very modest risk-adjusted price increase, and we're able to enhance our terms and conditions with our 2018 program versus 2017. We buy pretty far out into the tail, out to about the 1 in 250 return period.
For us, if we had a 1 in 250 year event, the net impact to equity is about 5%, so very, very manageable. In addition, for what we consider kind of our historical footprint outside of that, when you think about the ENF states that we're in, states like California, states like Texas, states like Florida, and when you think about the geo expansion, the new states that we've gone into, Arizona, New Hampshire, Colorado, and later this year, Utah and New Mexico, we buy the limit down or the retention down to $5 million. We cap our losses at $5 million. There's a little bit of co-participation up the side, but it drops our retention down to $5 million.
When you go back to August of 2017, when the big ones were heading towards the U.S., we felt very, very good about the reinsurance program we had in place, and it held up nicely. We also, from an individual risk perspective, we take some of the volatility out of our results by buying down the individual losses we can have on an individual risk, whether it's a property risk or a casualty risk, to $2 million per risk and buy our reinsurance protection above that. A key part to our overall operating model and business plan. From a reserving perspective, our goal is obviously to book our best estimate of the ultimate cost to settle all claims associated with the exposures we've taken on.
We take reserving seriously, you can see from over that time series for the last 11 years, we've had actual claims emerging come in better than expected, and we've had Favorable Development in each of the last 11 years. We do deep dives on a quarterly basis across all our major lines of business. We have a Big Four firm come in. They do an independent assessment twice a year in terms of reserve adequacy. They bring perspective and industry knowledge, which we find very beneficial. At year-end, they sign our reserve opinions. From a 2017 perspective, Greg teed up the Operating ROE 11.4%, very, very good result. We feel very proud about our results for 2017 and the last couple of years. When you think about 2017, you think about the level of CAT loss activity.
You think about the very weak or anemic pricing environment for standard commercial lines business. You think about the continued low interest rate environment to deliver a return that essentially was right on track with our long-term target we feel very, very good about. Again, going into 2018, we've increased our target to 12%, driven by the increase in our weighted average cost of capital. On the conference call just a few weeks ago, we laid out our expectations for 2018. We had a strong year in 2017. Our all-in calendar year GAAP combined ratio was 93.3%. When you back out 2.9 points of cat losses, 2.1 points of favorable development on an accident year underlying ex-cat, it was a 92.5% combined ratio. Our expectation for 2018 is for 150 basis points of margin improvement to take that down to a 91%.
Again, that's accident year ex-cat assumes no prior year development, then cat losses, our budget or expectation is 3.5 points, so 94.5% all in. When you look at the component parts of driving the underlying margin improvement year-on-year, it's really a pretty even split between focusing on expenses and driving the expense ratio down and also driving the loss ratio down. I would like to remind everybody, we're in a very volatile business. Actual results will differ from expectations, and as we have in the past on a quarterly basis, we'll update our guidance as we go throughout the year, and actual results differ from expectations. From a capital liquidity perspective, we go into 2018 with a very strong financial position. Our debt to capital ratio has decreased to 20.4%.
We've got headroom there if we need to raise some debt for any strategic reasons. We're at the low end of our leverage ratio from a premiums to surplus perspective. Again, that's 1.4 to 1.6. We believe the most attractive way to deploy our capital is back into the business. We grew, as Greg mentioned, at 6% last year, 8% in 2018. With the new states that we have in front of us, we have a little bit more headroom for growth. We believe that's the best use of shareholder capital at the moment, is to put it back into the business. We have what we call our sustainable growth rate, and that is at 9%. That's really if you think about our forward ROE and one minus our dividend payout ratio, our payout ratio is around 25%. We increase the dividend in 2018.
It puts a sustainable growth rate at about 9%, that's really if we want to keep the premiums to surplus ratio the same year-over-year, that's the rate of growth we can grow keeping the premiums to surplus at a 1.4 ratio. From an expense perspective, you saw that the margin improvement for 2018 versus 2017, our expectations, half of it's driven by a lower expense ratio. For us, when we think about expenses, we think about the long-term. We're making investments in our business. We're putting money into geo expansion. That takes time and money and costs. We're putting money into the omni-channel customer experience, into underwriting tools and technology, into RPA, operational efficiency, you name it. We're making significant investments in our employee base. We're also trying to be more efficient.
We're trying to leverage our infrastructure over a higher premium base. Our expectation is we can continue to drive the expense ratio down. On a statutory basis, our target's 33%. We think we can get there over the relatively near term. The ratio in 2017 was 33.5. We've also restructured our long-term incentive compensation. We had a higher allocation to a variable-based plan, which resulted in some volatility with some mark-to-market accounting on a quarterly basis. We've restructured that to take some of the volatility out and hopefully reduce the overall cost associated with long-term compensation that comes through the corporate expense line item. Last of all, from an expense perspective, we couldn't finish the discussion here without a discussion about U.S. taxes and the effective tax rate.
Selective is a U.S. primary domestic insurance company with no offshore operations, no affiliated quota shares, just a pure U.S. company. We're big beneficiaries of tax reform. The Tax Cuts and Jobs Act of 2017 cuts our corporate effective tax rate by 10 points from 28% to 18%. There's obviously a little bit of volatility in what the actual number will be, depending on our allocation to taxable versus tax advantage securities. We're expecting a full 10 percentage point decrease considering a consistent investment portfolio going into 2018 compared to 2017. Lastly, I won't go through all the bullet points on this slide. This is our investment proposition. We've sort of hit all the points on here. It's really about leveraging our competitive strengths. We believe we can generate continued sustained financial outperformance.
We have good opportunities for growth, lower risk growth, through our greenfield geo expansion strategy that we have in place that we're executing on. Good margins, our position for the firming market. Again, we're very focused on generating a return that beats our long-term targets over time, and a conservative approach to risk selection and managing the balance sheet. With that, why don't I stop, and we can maybe turn it over to Jay and Allison for any questions.
Yeah. I've got certainly a couple questions. Maybe for Greg. You obviously take your pricing is obviously by segment. It's not a broad brush. What I am hearing in the market, I want to see if you're seeing it, is that there are some companies out there mandating rate increases in certain lines of business or in certain regions.
1, are you seeing that? Secondly, if you are, can you take advantage of that to some extent?
All right. Let me start with some of the commentary of the market. In our presentation, first of all, you saw Selective printed a 2.8% pure, let me focus on the word pure renewal price increase for the month of January. That's not renewal price. Renewal price in my mind has exposure in it. You start parsing renewal price, that's a path that we don't want to go down. We talk about that. I will tell you, generally what we see at the start of the year is companies have a tendency to be a little bit more aggressive to meet their growth targets. In some cases, you got to understand, our renewal inventory is new business opportunities for some of our competitors.
We have to defend a little bit of our beachhead relative to that, we're very granular in how we defend that beachhead. I have not heard a lot of, from our regional folks, you had them all in, I haven't heard a lot of wholesale price increases going up. I'm starting to hear pockets of it, my belief is that as companies start to look at their performance in 2018, they'll come to the realization that arithmetic has no mercy, unless they're getting renewal pure price, unless they're getting rate overall, then they're not getting any rate. We plan on taking advantage of it. When I showed you the slide of our performance relative to CLIPS, there's no company that you're going to look at that has performed as well as we have quarter after quarter relative to beating CLIPS.
We will take as much advantage of that as we possibly can, within the context of commercial lines pricing, which has a tendency, even within cohort, to be a little bit of a scattergram. I hate to be a long answer to that.
No, that's fine. The other question, when you enter a new state, I assume there is a little bit more operational risk. You have new agents that you haven't dealt with. Have you gotten better over the years as far as identifying and integrating those newer agencies into your system?
That's a great question. I would tell you absolutely yes. Part of it's the tools that we have, but part of it, let's use Arizona as an example. Arizona was open with three field underwriters, two of them which came from one company, one came from another company. They already had existing relationships. At least two of the 17 appointments were existing relationships with Selective in other geo locations. These are shops that we knew but had operations in Arizona.
When we went to the meeting in Arizona, I will tell you, Jay, that the look, and you know our agents, you've been at a number of our agency events, the look and feel of the 17 agents that I met in Arizona was no different than going to any of our other meetings that we would go to in terms of their commitment, in terms of what they're looking at.
How we've improved a little bit internally is our modeling and our Insights product, which I think you saw some of the Insights tool that we have as an organization, allows our field underwriters to get a very good handle on how they feel this account's going to perform when they write it new. The metrics that we have are very complete, which we try to lessen the risk by getting a better grip on the new business that you're writing. Because everything's new-new, the actuaries obviously put in a higher acquisition cost in a new geo state than they would necessarily new business in an existing state that we're in. In our modeling are higher acquisition costs associated with that endeavor. We feel greenfield is absolutely the right way to go because in these three states, as I mentioned earlier, we have 37 agents.
They represent close to 20% of the volume in those three states, and we're handling it with three, four, five, six field underwriters in that. That's the kind of opportunity that we have as an organization, and that's exactly how we did Colorado and exactly how we'll do Utah and New Mexico.
I had a question for Mark. Prior year reserve development, you showed the graph, it's been positive really for some time. In your projections, in your guidance, you obviously don't assume any, I understand why you wouldn't. Having said that, is there anything in the environment that you're seeing, claims trends, that suggests a change? Where as an analyst, where I can put in that number in our model. Are you seeing anything to suggest, gee, that might really slow down?
It's a good question. I think when we look at our planning model and we look at our targets for the upcoming year, everything's about the current accident year. It needs to be clean, and we're looking to achieve our financial targets based on the current accident year as is not factoring in any development from prior years. I think going into every quarter, you have no expectation of favorable or adverse development, and you'd look at the data. Over the last 11 years, quarter in, quarter out, actual claims emergence, as I mentioned earlier, has come in better than expected. I think what you've seen is over the years, there's been shifts. Workers' comp was adverse for a while, and now the last couple of years has been very favorable.
The commercial auto used to be favorable, now it's sort of turned around on us. We had a little bit of adverse development in commercial property in 2017. That was a little bit unusual, but net-net, it's all been favorable. I would say there's nothing that's new or different than what we've seen over the last 11 years. It's going to move around from line of business to line of business. I think when you look at the overall statistics from the industry as a whole, you do see a reduction in the amount of favorable development that's come through the books on a calendar year basis, and I think there's going to be some pressure.
That's generally a leading indicator of price increases going forward, so there could be some benefits to that, but obviously, some pain if you're one of the companies that's having to work that through the system. I'd say there's nothing specific or different going into 2018 than we've seen over the last half dozen years or so.
Since we've been here all day taking questions, let me just get this one out. I mean, I know many of you have asked about, "Hey, well, Mark, what about headline CPI overexpectation? Are you worried about it?" I want to make sure that everybody understands that what drives our business is medical and medical inflation. It's hospital, physician services, Rx. That's what our medical payments trade off of, and that's what our liability suits trade off of. All of this about, well, headline inflation's been close to zero in 2017 and 2016, but our claim inflationary trends in both those years have been more like 3%. Really, it's the medical aspect that you need to pull apart when you look at it.
When you look at actually how many buildings or how much repair work that we're doing as part of our overall cost of goods sold, it isn't that much of our dollars being spent. Headline inflation does not drive our cost of goods sold. Our cost of goods sold principally are driven by medical and then the liability suits that trade as a multiple off medical. They've been in the 3.5%-4% range. Although lower than what they've historically been, they still are way elevated over headline inflation.
Great clarification. Thank you. Going to wrap it up here. Guys, thank you very much.
Thanks.
Thank you.