Alison Jacobowitz, and I work with Jay Cohen, covering the property casualty group. The next speaker, I get to say this every year, has not missed a year. We've been doing this forever. Up next, we have Selective Insurance. Presenting for the company is going to be, up first, is Greg Murphy, the Chairman and CEO. Has been with Selective for now 35 years? 36 years. He's had numerous roles with the company over the years, including COO, CFO. He's been around and seen it all. Also with him is Mark Wilcox, the EVP and CFO, he's going to be presenting as well. The company has evolved over the years to a key trading partner for their independence in their selected geographic regions, they've worked really hard on those relationships.
With that, I'm going to turn this podium over to them for them to talk about the strategy and tell you how they do it.
All right. Thank you very much, we appreciate us being here today to talk to you about what's very unique and special about Selective Insurance Company. We are the 41st largest property casualty company in the U.S., we have some very unique competitive advantages that we want to talk to you today about. The first is our franchise value that we have with agents. We're in 22 commercial line states, we have about, close to under 1,200 agencies that represent us. We have a very unique field model that you're going to hear about, whether it's on the underwriting side, the claims side, or the safety management element of our business. The fact that we have a very customer-centric organization focused around world-class or best-in-class customer service. Approximately, we have three major segments in our business.
I would say 78% of our business is commercial lines, 13% is personal lines, 9% is E&S. What I'd like you to think about as Selective as a carrier that has unbelievable relationships, but also has the capabilities of a national carrier. When we move forward and look at where we are, we have the 22 states that I identified. Think about it this way in terms of franchise value. On average, we have about 50 agencies in each state, if you take the 22 states into the 1,200 agency plant overall. I will tell you from a personal line standpoint, we're only in 13 states. We have a little over 700 agencies in total, again, built around franchise value. Where we are when we think about it is Selective's target is to be a 3% market share.
When we look at that, we kind of divide it into two buckets. The first is the amount that our agency premium or our agents write in a state. We call that the agency market overall, and we try to get that to be 25%. We want agents that represent Selective that write approximately 25% of the business in a state, and then our target share of wallet is 12. When you do the math of the 25 times the 12, that's how we get to a 3% market share. To put it in terms of opportunity, our commercial lines business is currently at about $1.7 billion. If you looked at increasing our share of wallet with our existing agents, that's about another $1 billion of premium opportunity.
When you look at adding additional states, adding additional agencies in the states that we're in, excuse me, creates another $1.8 billion of premium opportunity that over the long term gets you to commercial lines operation of about $4.5 billion in terms of total premium. I touched a little bit on the front part. The uniqueness of our model is centering our agency in the middle and covering them with a field representation. It's local claim handling, local safety management, a local underwriter that's empowered, that shows up in that agency's office, that creates and harvests relationships on a producer-by-producer basis that ultimately has allowed them to grow our new business and reach our premium targets overall as an organization. That's a critical part of our success. Our agents will bring out with them a safety person on a sales call.
They'll bring out the claims specialist, particularly in areas where local presence is important. That's what helps cement and sell an account overall. Customer experience is something I do want to spend a little bit of time on, and I will tell you, we as an organization are uniquely positioned to advance this. What customer experience means to us is omni-channel. Omni-channel means we do business however and whenever a customer wants, depending on the circumstances of that transaction. Single call resolution is paramount. Whether that customer is contacting the agency or finding us first, we want that call resolved right out of the box. That is something we're spending a lot of time on during 2016. We made a lot of baseline investments in master data management, so we now have a golden record of what a customer, all of the interconnectivity of a customer.
We're building a CRM system in front of that, so we can see every interaction that that client has had. What did they have on the claims side? What did they have on the safety management side? How do we bring that together and create more progressive outbound communications that are contextual in nature out into that customer base? We've got to find ways in our agencies to be able to deliver a level of service that no one else has experienced. What we want to be able to do is increase our collective retention levels, increase the number of apostles that are out there talking about the word-of-mouth opportunity, about what Selective brings, but more importantly, be able to provide services to that client that they're never going to want to leave Selective or our agent down the road.
That's a big part of what we do. It's a big expense effort, and I want you to think about this. Our competitive advantage to do this in the fact that we only have 1,200 agents, we are uniquely positioned to be able to execute this kind of strategy, because if you believe in the Pareto principle, in the 20/80 or 80/20, however you want to look at it, we've got to get 240 agencies on board, committed to this model to be highly successful in an industry that's very fragmented, that most of our large competitors would have over 30,000 agencies to try and execute this type of strategy. How many systems do they have internally in their organization that need to be able to be connected to be able to offer a golden record in total?
I know it sounds like a lot of words to you, but when you think about carriers, where they're positioning, how are they thinking about the future, how are they building a beachhead that's more defensible against whatever new competitor may be coming in? I don't care where it comes in from, but there will be new competitors in the marketplace, and if you own your customer and have a better relationship with that client long term, your ability to keep that customer is higher on a relative basis. When you think about it was the most profitable year for us, you look at any one of these metrics that we have on an ex-cat basis, we were in 89%, underlying statutory combined ratio of 92.2, and all in of 91.8.
I don't care which way you want to look at it, relative to AM Best, relative to everything else you're seeing, that's a superior combined ratio. I like to say arithmetic has no mercy. I love that line because you're looking at an industry that's going to generate an ROE from its investment portfolio of around 6%. They need to generate underwriting performance if they're going to meet their cost of capital or whatever internal targets they have. When you think about us, we provided on our conference call an underlying combined ratio of 90.5 that has no favorable development in it, nor does it have any cat load. You can add in your own cat load based on your own expectations. Ours has a tendency to be three and a half points, but that's 170 basis point improvement year-on-year.
About two-thirds of that is coming from rate versus trend and underwriting and claim improvement, and the other third is coming from expense improvements overall. It's a very methodical plan that we've laid out to generate those results. When you think about disciplined top-line growth, you can see when the market softened in that 2009, 2010 timeframe, we did not grow the organization. We didn't feel that that was the right time to grow. Obviously we're into a growth mode now today. We've got great execution and capacity with our AMSs in terms of what we've done. As I mentioned to you before, what we've done in terms of share of wallet opportunity, $1 billion. New agents in our existing states, which is the next most least risky way to grow, $1.8 billion. We've added Arizona and New Hampshire. We've got our E&S operation that's also growing.
We've got good growth opportunities internally to meet our long-term expectations. When you think about our profit engine, it's always been our commercial lines, which represents 78% of our premium. It printed a combined ratio of 89.9 or just under 90. However you want to look at it, that's a great combined ratio overall. I think it reflects the holistic activities that we've done in terms of claims improvements in our workers' compensation division, what we constantly do in rate overall in the book, and then our business analytics and how we improve our core book of business. When you think about this slide is something that I think you really need to focus on, and I think every one of our competitors should as well be able to articulate where their renewal price increases are currently today. This is a pure price increase.
There's no exposures in it. You shouldn't put exposures in renewal pricing because that's new exposure, and you can't count that as rate. Look at the consistent outperformance of us versus CLIPS, which in our opinion is the only actuarially sound mechanism out there to measure renewal price increases. You go all the way back to 2009, there were five quarters in there where we matched the industry all of the rest of the quarters. For the past eight years, with the exception of those five, we outperformed the industry consistently. If you compound that, our compounded outperformance is about 1,700 basis points.
When you sit there and say, "Hey Murph, why is your combined ratio in the low 90s and the industry's printing up in the hundreds?" You can sit there, wow, look it, you've got 1,700 basis points in more rate than the industry got over the last eight or nine-year time period. When you think about, well, how is it that you're able to do that and everybody else we talk to can't? Part of it is the relationship and uniqueness that we have in our agency plan, but more importantly, or as part of that, is the fact that we have the unbelievable technology business analytics that provide actionable decision-making capability to people at the front line. We have very data-driven people that work for Selective, but they're getting information that's actionable.
This actionable data shows you what it looks like by cohort in terms of underwriting. We've shown you some of the major pieces of what drives one cohort, what drive one cohort higher or lower based on where it's priced, what's the diamond score, so on and so forth. Those are just some of the things that work into the different cohorts that we have in terms of whether something scores high or something scores low. Based on that knowledge and the distribution, and we have this vertically through our entire organization, I could see it for a region. You could actually see it for a field underwriter or field inside person, and you could see it for an agency.
That's how it is deployed down to an agency level that we are having portfolio meetings with our agency about where their book of business is, what needs to happen in terms of accounts, and it is on a very granular basis. It is not socialized at all. That's why we are so successful at driving rate in the marketplace. Personal Lines is another area I know that catches a lot of attention. I would tell you that we have always told our investor base that for us, Personal Lines is a tale of two cities. We were going to be mercenary-like on the home book, driving very high rate increases over the past several years. We have done that. We have told you that our goal in home is to generate about a 90 combined ratio in a normal cat year.
This, for us in home, was fairly relatively normal in the cat year. We printed just under a 92 combined ratio. I will tell you, the heavy lifting is there and done in the home area, principally. Auto is something that we are still legging into. I think the market wins now are going to give us a little bit more of a tailwind to help us accelerate the amount of rate that we are going to get. Our expectation is around a 6% rate increase in Personal Lines. There is a lot of Personal Lines activity that is happening rate wise, and also I think from an expense standpoint, we will be able to drive. E&S is something that did print to 102 for the year.
It is not a huge book of business, about $200 million in premiums. A little reserve movement did drive that, and it does push it around. It is still a fairly immature book for us in terms of from that standpoint. The points that I would like to leave you with is that we are very comfortable with where we are writing new business. We are moving the pricing up on our renewal inventory to match where we are writing new. Not only that, we have re-architected the entire claim process that we expect will harvest benefits in the future in terms of lower loss and loss adjustment expense going forward. Not all that different than what you saw in the workers' compensation activities. With that, I am going to turn it over to Mark. We are going to open it up to questions in a minute.
Mark, thank you.
Thank you, Greg, and good morning. I'll talk about our financial strength and risk profile. I'll move relatively quickly through the presentation, and hopefully, that leaves us a few minutes to answer some questions at the end of the presentation. We have a very clear and transparent financial target, and that's to deliver an operating return on equity that's 300 basis points over our weighted average cost of capital. We believe if we do that, we'll be able to compound book value per share, adjusted for dividends over time, and by doing so, we'll deliver a strong total shareholder return over time. We have a strong balance sheet, and it's underpinned by three key factors, and I'll touch on each of these three key factors in a little bit more detail in the coming slides.
One is a conservative approach to managing our investment portfolio, a conservative approach to capping our tail risk from a reinsurance purchasing perspective, and a disciplined approach to reserving. It's also important to note from an underwriting perspective, since we're in the small to medium-sized commercial account business, we write low to medium hazard business. That means our policy limits or exposures are relatively small, and our average premium size is relatively small. A fairly conservative underwriting mix within the underwriting book of business. With that, being that we have a conservative balance sheet and conservative from an underwriting perspective, we believe we can take on significantly more operating leverage than the industry. We write from a net premiums written to surplus ratio at 1.4 times, and that's about twice the industry average.
Because of that operating leverage, as Greg mentioned, we believe that gives us a competitive advantage, in so much as our competitors have to generate twice as much underwriting profit for the same level of operating ROE for their business. For us, each point on the combined ratio, each point of underwriting margin, is approximately 100 basis points of operating ROE for us. Finally, on this slide, from a financial strength perspective, we're highly rated AM Best A rating, and recently upgraded to A from A minus from S&P in the fall of 2016. From an investment perspective, we believe we have a conservative investment portfolio. Double A minus is the average credit rating. 3.6 years is the effective duration for the total portfolio, 3.8 excluding short-term investments. The fixed income portfolio, 97% of it is highly rated investment grade.
3% of the portfolio within fixed income is allocated to high yield. If you take that allocation and you add to it from the left-hand side of the chart there, the equity allocation and the alternative allocation, that's about 7% of the portfolio that's today allocated to what we consider risk assets. That's an area that we're willing to take a little bit more investment risk within the investment portfolio. Our internal comfort level is to take that up to about 10%. Today we're at seven, but we have a little bit of headroom there, and we find some opportunities within the investment environment over the coming months. We generate strong cash flow from operations, about $300 million in 2016 that we're able to put back into the investment portfolio and grow it.
At the bottom of the slide, you can see our investment leverage at 3.5x, multiplied by our yield delivered 6.7 percentage points of ROE in 2016. A couple of weeks ago, we had our earnings call, we delivered our expectations for 2017, and you can see the bullet point there for 2017. Our expectation is when you think about our current book yield, the expected cash flows we're going to generate, our asset allocation strategy for 2017, and expected interest rates. We're looking at about $110 million of net investment income on an after-tax basis for 2017. That would deliver about, as Greg mentioned, about 6.75 percentage points of operating ROE for 2017. Moving on to our reinsurance program. As I mentioned, we believe we have a reasonably conservative approach to purchasing reinsurance.
From a per risk perspective, we buy per risk cover from a property and casualty perspective. We like to cap our large losses to about $2 million. From a cap perspective, we have a relatively low retention. $40 million is our retention. We buy out just past the one in 250-year return period. When you get out to the one in 250 or the 99.5th, if we had a significant cat loss, our major peril that we're exposed to is Northeast hurricane risk. It would be about a $53 million loss to us for about a 3% hit to equity, so a relatively modest hit.
You go all the way out on the tail to a very extreme event, say the 99.8, the one in 500, you can see that goes up a little bit to 16%, but still very manageable from a balance sheet and capital management perspective. We also pay very close attention to credit risk when purchasing reinsurance. When you look at our balance sheet, we've got about $600 million of recoverables on the balance sheet. The average credit rating for our reinsurers is A-plus. At the top level, from a contingent credit risk perspective, if we did have a significant tail event, we'd like to buy the very top layer of our cat program on a fully collateralized basis. About $200 million at that limit is fully collateralized.
Again, we believe that mitigates some of the credit risk in an extreme tail event scenario and protects our balance sheet. From a reserving track record, clearly our objective is to book our best estimate of the ultimate cost to settle all our claims based on the exposures we have at the end of each balance sheet date. We believe we have a very sort of disciplined approach to reserving. Over time, you can see we've had 11 consecutive years of favorable development, and in 2016, we also continued that trend of favorable development. We have had a few small pockets of adverse development. If you drill down on the $66 million of net favorable development in 2016, we did have $25 million of adverse development within our commercial auto book. That's included in that number.
That was more than offset by very favorable claims emergence trends within our workers' compensation line of business, as well as our general liability line of business. That $66 million of net favorable development improved our combined ratio in 2016 by about 3.2 percentage points. Looking ahead to 2017. Again, on our conference call two weeks ago, we laid out our targets and expectations for 2017. As Greg mentioned, we have a very clear strategy, a very detailed operating and tactical plan that's driven down throughout the organization and a very detailed ground-up budgeting process. As we went through that process, we came up with our guidance and plan expectations for 2017. As Greg mentioned, 2016 was a record year from an underwriting performance perspective. Our all-in statutory combined ratio was 91.8 percentage points.
If you back out the 3.2 points of favorable development that I mentioned and 2.8 points of cat losses that we experienced in 2016, the starting point for the waterfall slide is a 92.2 sort of underlying statutory accident year combined ratio for 2016. As you roll through the waterfall slide, you end up at the final number, the 90.5. Let me just spend a minute or two just going through the component parts. In the first part of that slide, you'll see that there's loss cost trends embedded in there. We're expecting about 2.5 points of inflationary increased loss cost trends within the combined ratio going into 2017. Because that only hits the loss ratio portion of the combined ratio, that's why you see 1.9 points on the slide.
From a pricing perspective, a lot of the earn rate that we generated in 2016 is going to earn into 2017. We're expecting about 2.5 points of price improvement. As that goes through the combined ratio, there are expenses associated with that, namely agents' commissions, and that's why you see about two points of earn rate coming through the income statement in 2017. Put that all together, earn rate is largely offsetting loss costs. The remainder and the majority of the 170 basis point margin improvement we're expecting in 2017 is really driven one-third by expense ratio improvements and two-thirds claims initiatives. Very targeted claims initiatives, which we expect to drive loss costs and loss expenses down, as well as underwriting mix improvements. On the expense side, I'll just touch briefly on that. Our focus is not deep cuts within our expense base.
We're continuing to make significant investments in technology, significant investments in our customer experience, significant investments sharpening our underwriting tools, and significant investments in our employee base. Our expectation is we're going to be able to increase our premium growth rate at a higher rate than our expense base. Our expense rate base will actually increase in 2017, but at a lower growth rate than our top line, and therefore, we're expecting about 60 basis points of expense ratio improvement going into 2017. The final point on this slide that I'll mention is outside of the investment operating income, the investment income generating operating income, outside of the underwriting results generating operating income, we do have some expenses at the holding company. They took off about 240 basis points of ROE in 2016.
Those are principally debt servicing costs net of tax, as well as some holding company expenses, principally long-term incentive compensation, directors' costs, and some other fees. Our expectation after restructuring our long-term incentive plan is to take about $10 million of long-term incentive costs, those holding company costs, out of the expense equation in 2017. I think the message that we'd like to leave you with from an investment proposition perspective is, one, we believe we have a long track record of financial strength, superior execution, and disciplined, profitable growth. We have a strong balance sheet, which allows us to take more operating leverage than the industry as a whole. Each point of combined ratio equates to about a point of operating ROE. From an investment perspective, as we mentioned, we're expecting about 6.8% of operating ROE from the investment portfolio going into 2017.
Greg talked about the premium growth opportunities we have. We believe we have significant opportunities to grow in a disciplined way, organically within our 22-state footprint. That's the $2.8 billion opportunity that Greg talked about, as well as we have an opportunity to grow as we expand our states going into the latter part of 2017. The last point I'll leave you with is our focus is we're looking to grow, but it really is disciplined, profitable growth. As you all know, it's very easy to grow in our industry. The challenge is to grow in a disciplined and profitable way, that's our goal and expectation for 2017. We believe we have sufficient capital when you think about our capital from a rating agency perspective, an economic capital model perspective, and from a regulatory statutory perspective.
We believe we have sufficient capital to support our expected growth going into 2017 and beyond. With that, I would be happy to take any questions that anybody has.
I've got a couple of questions. I guess one question, on your pricing, where you've exceeded the CLIPS survey for quite some time pretty consistently, is part of that, Greg, a business mix issue, that your business mix is different than what's represented by the CLIPS survey? If John wants to handle that.
Yeah, I'm going to have John answer that. John's our President and COO, so I'm going to let him come up and answer that question for you. All right, Jay? Thank you.
Yeah. I would say if you look at the results, it really isn't driven by a business mix issue. We get the results by line, but generally speaking, on an overall basis, we don't attribute it to a mix difference from what the industry's reporting. We would attribute it to what Greg took you through earlier, which is the approach we take and the relationships we have in the communication it results in with our agents to deliver that kind of performance.
Just getting that mic all set. One additional question, that is, you showed us the graph of the cohorts.
My question is, on the far right, the most challenged cohort, that was a relatively small percentage of your business. Question I have is, has that percentage changed over time?
The question is, I'll tell you. No, it doesn't really change that much over time because the cohorts, for the most part, are force ranked. What should change is the relative curve of performance. In other words, when we first started modeling, you could have seen a line like this, and your best performing business could have had a very low combined ratio, and your worst performing business could have had a very high. We're always force ranking every year and redistributing by cohort. At the end of the day, you would expect every year this line to tilt more like that and become a little, it's never going to be flat. You'd love to see it be flat, but it's going to slope less. That percentage-wise is normally it gets reloaded every year.
That's where the potential improvement.
Yes.
Could come from.
Constantly comes from that because our inside underwriters know. Actually, no. Our AMS is through some of the new tools we have, which I am very happy about what was done. Actually will show when a field underwriter writes a piece of business, then they go through their mindset, they document, they price it. They're going to know as part of that equation what cohort, absent any claim activity, would this.
Policy that you're writing right now today fall in at renewal, which I think is a great piece of knowledge for a field underwriter to have. "Hey, I'm writing this account. I think it's here. I'm going through my process. Oh, by the way, if it renews with no loss" Only about 8%-10% of your policies in force have claims on them. If that account renewed with no loss activity, this is the cohort that it would be in, which is a great piece of knowledge to have.
Got it. I have one question for Mark. Mark, your slide on your reinsurance program-
Yeah.
Was pretty amazing. You guys are ready for like Armageddon- Given how much you buy.
Yep.
Have you thought about shifting some of those dollars at the high end down to take out some of the more frequent smaller losses? I know those are more expensive-
Yeah. No, that's a very good question, Jay. The core of our reinsurance program is balance sheet focused, so we're there to protect the balance sheet over the long run. We do like to take a little bit of income statement volatility out by buying on a per-risk basis for casualty, capping the claims at $2 million per loss, and on a per-risk basis from commercial property at $2 million, $5 million within the non-standard footprint within the E&S line. We have talked a little bit about buying on an aggregate basis to take some more P&L volatility out. It's expensive, as you know, and a little bit harder coverage to buy. Those are discussions that I know our chief risk officer and myself and Nas have had some discussions about. It's something we'll definitely take into consideration.
It's just a matter of how much you're willing to pay and how much expected profit you're willing to cede to protect the income statement a little bit more. It's a trade-off.
Okay.
I'd say that's a great question, you think about risk of ruin in the industry. That is the risk of ruin that you get. Let's not forget that you had AM Best for a period of time in the stochastic BCAR that was seriously debating a 500- and 1,000-year return period in their BCAR results. Now they backed off that ultimately, and they settled in at the 250-year sized event, which is a point, I like to refer to that as a 0.4% probability because it's more in probabilities, not in the amount of years that have to pass. I will tell you, if they did go to that extreme event in their stochastic modeling, it would've created a lot of issues for carriers in the marketplace relative to how much surplus protection they had from their cat program.
Again, it's something that you need to be thinking on for the long term relative to what may happen and change because you certainly don't want to be scrambling in the market to create programs. We always like to talk to our This is our conversation with our investor base. "Hey, we're very conservative on our investment strategy. We're very conservative on our reinsurance buying." Where we take more risk is we're at 1:4 to 1 premium to surplus ratio. We write low to medium hazard style business, that's where we're getting twice the return that the industry's getting in terms of underwriting margin performance.
In this market, there isn't anything that I see that's going to move interest rates on a book yield basis high enough that would sit there and say, "Hey, you can now not have to underwrite at a profit." I've been in the business like you mentioned, 36 years. In my lifetime, right? I have never seen a stretch of interest rates this low. Our budget for this year in terms of investment income has a 275 core underlying assumption on the 10-year. The 10-year right now is at 240. That's not some huge lift in assumptions relative to how you get to our $110 million of after-tax investment income.
Even with that, you sit there and go, "Wow, do you know how long at a three and a half to four-year duration that the industry's in, how long it will take them to move book yields higher?" It's years down the road for that to happen. We feel very comfortable about not changing those risk metrics and keeping the focus around on the underwriting style opportunity on the leverage side.
Any last-minute questions?
All right. Well, thank you very much. Appreciate it. Thank you.
Thank you.