Good morning. Thank you, Brian and welcome to our Third Quarter Investor Event. Today we will begin with an update on our efforts to penetrate the home auto bundle market. Our presentation will last approximately 40 minutes and will be followed by a Q&A session with our CEO, Tricia Griffith; our CFO, John Sauerland; and our guest speaker today, Heather Day, Preferred Marketing Process Leader. For Q&A, Bill Cody, our Chief Investment Officer will join us by phone. As always discussions in this event may include forward-looking statements. These statements are based on management's current expectations and are subject to many risks and uncertainties that could cause actual events and results to differ materially from those discussed during the event. Additional information concerning those risks and uncertainties is available in our annual report on Form 10-K, where you will find discussions of the risk factors affecting our businesses, Safe Harbor statements related to forward-looking statements and other discussions of the challenges we face. These documents can be found via the investors page of our website, progressive.com. And with that it's my pleasure to introduce our CEO, Tricia Griffith.

Good morning, and welcome to Progressive's Third Quarter Webcast. Before we get started at the topic at hand, I wanted to reflect a little bit about the last quarter, specifically with Harvey and Irma, two names I think all of us are very familiar with based on this very, very active hurricane season. As you know from my letter, several of us went down to both Florida and Texas just to get some insights into how our employees are doing, do they have the right resources to help our customers and get some insights into our customers and what they need from us. This year, at this CAT season, we took down one of our internal videographers just to capture some of the moments and really to have everyone at the company understand not only what happens in a CAT but how we react, which is so important and so critical to our customers. So we captured this a couple of minute video, mostly internally because although we had our CAT team available in both of those states and almost 1,300 reservists from around the country, not everybody at Progressive gets the insight of how we really handle CATs and to us, that is really something we find very important, both for our employees and customers. So I'd like to share that around two-minute video with you for you to get some insight into our culture and the importance of making sure we take care of our customers in their greatest time of need. [Presentation] As always, I'm very impressed with our response, both on the Progressive and the ASI side. And we know customers want us to be out there quickly, and we're known for our speed and that's why we take so much time to really understand all…

Thank you, Tricia. Good morning, everyone. The size of the opportunity for Progressive in the personal line space remains very large. We're very excited by the fact that we can now attract and retain virtually all personal lines customers, executing on our destination strategy. Most of you are familiar with how we segment the personal line space, but for those of you who aren't, let me spend a minute to review. We have four primary segments within personal lines, those are represented by the personas of Sam, Diane, the Wrights and the Robinsons. Sam isn't consistently insured, you can think of Sam on the nonstandard end of the spectrum. Diane, a renter, who is more consistent with her insurance consumption. The Wrights, homeowners, who don't bundle home and auto, and the Robinsons, who bundle home and auto. You can see on this slide, we've delineated the personalized market space across those four segments as well as the primary distribution channels in which they currently consume their insurance. As you can see, the Robinsons are over half of the personal lines marketplace. We've had some success penetrating the Robinsons segment, our growth year-on-year in terms of policies in force for households is 30%. We think we now have slightly less than 2% of Robinsons countrywide or around 0.75 million households. The predominance of the Robinsons in our current customer set, we have developed or as we say graduated. Meaning they started as a Sam, a Diane, or a Wright, and we have evolved them through their life to be a Robinson, bundling home and auto. Now today the new business, the new Robinsons coming into Progressive are a little more equal, 50-50, and that's driven predominantly by independent agents who are moving existing Robinsons to Progressive, but by and large our…

Thank you, John. So as John highlighted, there is significant market opportunity for Progressive in the Robinsons segment. In 2014, we laid out our vision for the Destination Era and plans for growing that book of Robinsons. Subsequent investments in ASI, but also in processes and systems across channels, have laid the foundation for that growth. In year-to-date, we are experiencing steadily increasing velocity in creating Robinson bundles. We are now accelerating away from the historical slope in both direct and agency channels, with about equal parts coming from our existing auto customers that are transitioning to bundled homeowners with us, and from new business Robinsons that are adding an - that are buying an agent - an auto policy as they bundle with us. So property is considered an anchor product in the Destination Era and we have extended our core distribution strategy to include property. As with auto, property insurance is now available how, where and when customers to choose to interact with us and that requires thinking about property comprehensively. It has been woven into how we go to market and throughout our funnel, from our branded messaging, across all of our channels for purchasing and servicing a policy and through engagement with our existing auto customers as they become homeowners. As part of that strategy, Progressive is well positioned to connect with the highest potential segment for home insurance shopping. The top graph here shows the overall U.S. population by age, the bottom graph shows the Progressive auto distribution by age. Merging the two, the orange line highlights Progressive's market share by age relative to the overall U.S. population. Progressive has an advantage position with the younger population. And homebuyers skew younger. Overall, 62% are under the age of 44. Not surprisingly, the skew is even…

[Operator Instructions] I'm going to start with a question that I've gotten frequently, which is how has the hard market contributed to the success of Platinum in the Agency channel?

I'll start with that, and Heather, if you could elaborate a little bit, specifically with Platinum. Overall, the hard market is really about being in the consideration set, especially on the Agency side. So when we think of that, we think of course our brand, your trusted brand. We think of service, especially times like this when you have catastrophes agents get feedback on exemplary service and service that needs work, so make sure you always service your customers. But more importantly, when the customers come in that agency, do you have the right product at the right price? And that's really the key to being successful in a hard market, to be able to have a great product and we are very happy with our [indiscernible] product at a very competitive rate. And to me, that's the overall picture of how you really win a hard market. Specific to Platinum, Heather, why don't you elaborate a little bit?

Yes, I think you're right. It is about being in the right position to take advantage of market opportunity. If I look back a couple of years, we would not have been in a position to grow, whether it is a hard or soft market. So it is really the work that has happened to reinforce our bundled product, to deepen those relationships with agents, that during this time where we've got the opportunity to grow, we're in a position with the right product, the right relationships at the right time.

And the agents really do talk about that. So I do a lot of big agency events, and you can tell just that in the last year as we've come together, especially with the agent representatives that go in, the agents really respect and honor that relationship and you can tell it's adding to our results.

Thank you. So Brian, can we please take the first question from the telephone participants please?

Of course, our first question comes from the line of Elyse Greenspan from Wells Fargo. Your line is now open.

Hi good morning. My first question, I guess, first as I think about the bundled customers versus non-bundled customers, do you assume a lower loss ratio on the bundled side, just because there generally have been more affluent customers or can you kind of just walk us through how you're kind of thinking through the margin profiles of the two different segments of the business?

Yes. We do assume – we'd look at all of our customers with segmentation and subsegmentation. So generally, we believe that bundled customers will be more competitive and so we bake in, obviously, loss ratio-expense ratio differences in those customers and we know that they stay longer, which also influences that.

Okay. And then something that's a bit more topical over the past has obviously been the prospects of tax reform. Given that you guys could be a pretty big beneficiary on the tax side, would you look to keep any potential benefit or do you think that, that - would that be passed through to potentially lower pricing?

Yes well, the devil will be in the details. Generally speaking for Progressive, we believe that if it comes to a 20%, we don't know what the outcome will be, that we will benefit from that. I think what I'd like to do is to have John talk about how we look at that from an operational side and then Bill Cody is on the phone, and he can talk about how we've been looking at that from an investment side. We talk about this all the time, at least in the last year or so, and model out what those changes would be. So John, why don't you start on the operational side?

Sure, obviously, a timely question. I'm sure most of you spent a lot yesterday reading a lot on potential tax code changes. Generally, it is our bias that we would allow benefits of tax change, federal tax change to fall to the bottom line. That said, we are - we would be considering the competitive environment in which we operate as we decide whether we would change that 96 target or not, our initial expectation is that we would not. We would also obviously try to understand the likelihood of permanence in any change. But I also would like to note when it comes to taxes that we actually pay as much tax above the line in an average year as we do below the line, so to speak. So for insurance, we have a state premium tax, generally not a state income tax, and that state premium tax is about $0.5 billion for Progressive, and that's included in our expense ratio. We do a lot to try to optimize our premium tax rates, meaning multiple legal entities. We work on getting tax credits as well. And more recently below the line for federal taxes, we have been working to be more proactive in managing our federal tax rate through the use of renewable energy credits. So generally, we will expect to allow it to fall to the bottom line unless the competitive environment changed a lot. And I think Bill would - can comment on the potential implications for our investment portfolio.

Sure, Elyse. The biggest implication or the direct implication would be in the muni portfolio, which is now, let's say, 8.7% of our total portfolio, which is down about 2% from last year, the end of last year. So that comes to a little bit less than $2.4 billion. About $900 million of that or roughly 40% is in housing bonds which - most of which, we would still find attractive at a 20% tax rate. A similar amount we hold for premium tax abatement purposes, which is something that John had mentioned and I would say the vast majority of that would still be attractive at lower tax rates as well. The balance, there are some special situations which will be also attractive. And then others that we would sell, given valuations and tax rates because we look at munis on a fully taxable equivalent basis compared to other products. And to the degree they're not competitive on that basis, we will sell them. We generally expect munis to do well under the proposed tax plan because individual demand should hold up pretty well, given the top tax rate remaining at the 39.6% and the elimination of state and local in the income tax deduction. And as well as the fact that we will see some supply-demand - supply drop, given the elimination of advanced refundings and private activity with bonds. Through the rest of the portfolio, corporate bonds should do better as cash flow is increased and leverage is reduced. We should also see some less issuance in that space as well. We also expect rates should move a bit higher on the boost to growth and some possible F1 series bonds [ph]. And we also think our equity portfolio should do well. Obviously, it's a big winner of reduced corporate taxes.

Thanks bill

Thanks Elyse. Brian, can we take the next question from the conference call line please?

Our next question comes from the line of Bob Glasspiegel from Janney. Mr. Glasspiegel your line is now open.

Good morning, Progressive. A question on the - on how we should think about the option execution. Is there anything magical about getting into a 100%? I mean if things are like they are, do you anticipate that you're going to want to get to 100%?

Yes, so Bob, I'll take that one. In April of 2018, we have the option to get to 80% or more and I expect we will take that option. We have the option to get to 100% at the end - or in April of 2021 and it is our expectation that we would take that option as well. The owners of those shares also have the option to put their shares to us at the same time at those two different dates. This is obviously intended to ensure that the key leadership team that I talked about remains to continue to build the business and be very engaged, which they continue to be. On that note I will, however, mention that we have also recently had a senior person from Progressive join the ASI team. So Dave Pratt, who most previously ran our Snapshot business and has a long history, probably 27 years or so with Progressive, is now in direct report of John Auer, the CEO of ASI. So working to get further integration, but yes, we would expect that we would be at, at least 80% come April of '18, and get to 100% in '21.

Sorry, Bob, I didn't know you were referring to ASI. I was like, options? We don't have that options anymore. Sorry.

A question on where the sort of current run rate is in homeowners. If we sort of take out amortization and investment for growth and sort of unusual CATs, where should we think of this sort of current year-to-date run rate?

You can talk a little bit more about that. I think taking out the amortization expense and the current CATs, we feel much more positively about the run rate. And of course, as we expand across the country, we learn more and more about - we have more development and we understand a little bit more of each state and the nuances. But we also believe that that's a great strategy because being across the country and not having so much concentration in the states we have now will actually be beneficial to us.

So just to understand your answer, you are earning your targeted return underlying? Or you need more scale to get there?

Yes, I can answer that, Bob. So as I mentioned, when we modeled the value of ASI, we obviously assumed a margin, operating margin and we are slightly above that inception-to-date. Actually, in 2017, if you take out that amortization, we're at a 101 year-to-date, and that is not where the model assumed we would be on an inception-to-date basis. But of course, homeowners is spikier, consequently the need for reinsurance, consequently need for greater capital levels. But on an inception-to-date basis, given the fact that we've had a high degree of hail as well as the named storms, we feel great. So we have a target combined ratio, as I showed you on that slide. For the property business, it is well below 96, reflecting the additional need for capital. And we think our underlying profitability is solid. Homeowners is a line in which we have been taking rates up some. We expect to take those rates up more throughout the rest of the year and into 2018 as well. So we are seeing loss cost trends in homeowners accelerate a bit, and the folks at ASI are responding.

Thank you.

Thanks Bob.

Brian can we take the next question from the conference call line please?

Our next question comes from the line of Kai Pan from Morgan Stanley. Mr. Pan, your line is now open.

Thank you and good morning. The first question is on the reinsurance side. If the reinsurance rates are going up, would that impact your reinsurance strategy and would you be able to pass on some of the increasing cost to your Home customers?

We don't know for sure if reinsurance will go up. That'll be dependent after everything. It's true in different companies. We believe that we have a great strategy in terms of having multiple policies that are staggered in different years. So we will - we always take a look to try to minimize our downside and we'll take action depending on what the reinsurance market does and what we feel like we need. But we want to obviously protect that spike in homeowners. So John just talked about that. And having two of the biggest storms in history in two of our biggest states for Home, we realized how important - not that we didn't before - how important to homeowners reinsurance is and we'll do, what we need to do to protect that downside.

If I might interject, I may have misspoken in my response to Bob. Inception-to-date, our combined ratio for homeowners is a little higher than we modeled. I think I shared that in my comments. But in case that was a misstatement on my part in Bob's response, I apologize. We are slightly above inception-to-date.

Kai did that answer your question?

Yes. My followup question is on the auto sort of loss frequency and we have seen pretty sort of like a lot of improvements so far this year. I just wonder, could you give a little bit color on the sort of what's like the puts and takes in there and how sustainable is that going forward?

I will start with the sustainable piece, Kai. It's really hard to understand necessarily frequency because a lot of macroeconomic trends can change with frequency. What we do see internally that we believe is favorable on that side is the fact that we have a much more preferred mix of customers and so we know that that's what lowered the frequency. We have some upfront underwriting we've put in, in the last several years that we believe is making a difference and when we look at our snapshot data, we have longer miles. So think of longer trips. So think of trips over 15 miles. And we're seeing that, and we believe that the Snapshot population kind of mirrors the rest of our Population. So when we think about that and create hypotheses around frequency, those are normally what we talk about. And we always have other underlying hypotheses that we continue to try to quantify. But we wouldn't want to try to predict frequency we react to it.

Okay, great. Thank you so much.

Thanks Kai.

All right, I'm going to take a question from the webcast at this time. So the question is from Austin Boaz from Principal Global Investors.

Do you think there would be better acceptance by independent agents of homeowners insurance if ASI was rebranded as, say, Progressive Home, for example? Why keep ASI as its own brand name? So a two-part question.

Great question Austin. Yes, and thank you, we're going to be doing that. So yes, we - so let's move back up. So we purchased ASI because in the states they were in, they were known as the preferred brand and agency channel. And for years because of our heritage of being a nonstandard auto, we knew that was a good marriage. So we've learned from - as the last couple of years have gone through, not just from agents but from ASI as a team, we knew there'd be a right time when we decide to name the product Progressive Home. So over the last year or so, all of us have had lots of conversations with agents. And these are agents in some of the agencies where they have been an ASI agent for a long time and they said, "You know what? It's time to leverage the Progressive brand and have it be Progressive Home and Progressive Auto." And so we've decided to do this. It will take some time, because of just systems and different things. But Heather's actually - will be running that project, and you can add in any more if you want, but yes, we think it's the time. We now are known as an auto home bundler. We have the right brand for that. It's a preferred product on both the auto and home side, and it's perfectly the right time, so great advice, Austin. Do you want to add anything?

I would just echo what Tricia just said. We tread very carefully here, knowing the legacy that ASI had built with agents in its existing states and how well recognized they were by those agents. As we expand it again, it was one of those, we've got those deep agent relationships through the Platinum program. We listened carefully and understood that as we are beginning to build out the Progressive Home brand generally, this was really the right time to make the switch, so yes, under way.

And I would tell you, the first person that would say we absolutely have to do this is John Auer, the CEO of ASI, and he's leading the way as well.

All right, Brian can we take the next question from the conference call line please?

Of course. Our next question comes from the line of Christopher Campbell from KBW. Mr. Campbell, your line is now open.

Yes, hi good morning.

Good morning.

The question on auto, do you think the reinsurance on your auto line would be useful given there is probably a benefit if you did a multiproduct treaty now does Progressive approach that type of question?

That's a great question. We've been talking about that for the last couple of years as we've integrated with ASI. We have it on some of our commercial auto. So the higher limit Commercial Auto, over $1 million, we have reinsurance. I think as we think about continuing to expand, we always consider and protect the downside and weigh that with the cost. So it's something we talk about often to determine does that makes sense, does it make sense especially like you said with the multi treaty. We haven't gone there yet, but it is - it's just something we discussed.

Okay, great. That's very helpful. And just the followup is how would you categorize the relative sophistication between your auto and homeowners pricing right now?

So from Auto, we have had years and years of segmentation, and a lot of variables go into our auto pricing and then recently, of course, we've added the upfront underwriting. On the home side, there's been a lot more underwriting in terms of acceptable risks. So coming in the door, knowing right away what we would accept and what we wouldn't. Now we're really starting to share data and best practices with both companies to really get the best of both worlds. So sharing the underwriting sophistication from ASI as well as our pricing and segmentation sophistication and we've been doing a lot more of our business reviews collectively together. That's one of the reasons why we had Dave Pratt come down to work with the ASI folks. And we work back and forth to really share the best of both worlds. So I would say each company has different levels of sophistication. So together, I believe we're going to have a really optimal approach to even understanding segmentation at a more granular level.

Well, thanks for all the answers.

Thank you.

Brian the next question from the conference call line please?

Our next question comes from the line of Adam Klauber from William Blair. Mr. Klauber, your line is now open.

Thanks. On Platinum, I think you mentioned 5% of the agencies are currently Platinum and also you're being selective. But as we think about a couple of years down the line, will that be multiples of the 5%?

Yes, we will obviously select more Platinum agents as we grow and expand and we really have always positioned it as a scarcity model. And we want to make sure these agents, when customers come into their agency, we're their first or second choice. And some agents actually won't have access to the customers that want to have a home and bundled package or maybe don't own a home. So we will expand when we believe it will increase our Robinsons population, our auto home bundle. We'll be expanding for sure in the next couple of years just because of our geographic expansion. And then we'll likely step back and say, "Are there opportunities in different geographic locations to increase our Platinum agency population?".

Okay, thanks. And then could you talk a bit more, I think you said in '18 that you're rolling out for Platinum a better interface and also better bonus program. Could you just talk about those a bit more?

Yes. I'll have Heather talk specifically about the bonus program. I brought up the overall commission schedule that we talked about during the last webcast, so we call it Paths to Partnership, and we are investing a lot overall in our agency platform with IT, just to make things smooth and easy for agents to be able to have that auto and home and bundle online in the same system. Do you want to give any specifics on that?

Well, I would - I think underlying this question is a sense of what kind of levels of investment are we looking at in terms of Platinum and how does that line up in terms of our cost structure. Let me know if I'm off-base on that. I think we feel very good that we are coming into the Platinum program with a very competitive cost structure in the agency channel. And as I think about the investments we are making in terms of ease-of-use with the overall platform as well as thinking about how we are going to compensate agents for both their bundled business and for other lines of business that they're writing with us, we are really rethinking about both the quality and the volume of the business we're getting, the kind of retention we expect to see with this bundled business. So overall, these investments should align us - should position us very well for growth, but we do not expect upward pressure on the combined ratio. We will keep that very same target in place.

Okay, thank you. And also the last on bundled, you showed the Robinsons population is growing pretty aggressively in bundled. Is that coming - is that growth coming more from the captive or independent agency channel or both?

I would say the majority of the Robinson bundle is coming from the independent agents channel. That's not to say that it's captive, Robinsons aren't shopping as well. And of course, the other part of the Robinsons bundle, not with Platinum but on - in our in-house agency on the direct side, really comes from the overall population in terms of what John had talked about with a lot of graduations of our Dianes and other people in our segmentation, so it's really across the board. But if I had to stay on the agency side, the majority of our growth is coming directly from the independent agent channel.

Okay, great, thanks a lot.

Thanks.

Brian, can we please take the next question from the conference call line please?

Our next question comes from the line of Brian Meredith from UBS. Mr. Meredith your line is now open.

Yes, thanks. Hi Trish.

Hi Brian.

A couple of questions, first, back on the frequency. I'm wondering, do you see any benefit, or in any of your data, see any benefit when you do have some major hurricanes or catastrophic events like this did you see a decline in frequency?

Yes, we talk about that. We can't surgically say absolutely yes. Right after Harvey, I walked into John's office and asked him the same question. Let's look at this. So I wouldn't be able to say concretely absolutely. We look at that, try to address that, but it's really specifically an unknown. I think the philosophy and probably what you're getting at, Brian, is - so if you leave, is there less frequency in BI et cetera. We look at that. We can't exactly point to yes, that, that equates to frequency because again that is a couple of states and we have the whole country to contend with when we look at overall frequency.

Great. And also just on that topic, if you quantify or can you give us a sense of what benefits you're seeing on frequency from kind of mix shift to more preferred? And kind of what do you expect kind of going forward?

I can't give you very - specifics in terms of that, but we know when we have the more preferred customer that - and the ones that stay longer, so we're going to be making more money over a long period of time that we benefit from that mix of customer. And I think the exciting part about this is that as we talked about, we have so much momentum going on and so much more to capture and we're seeing this now. So we believe that if it continues, hopefully, we'll have those low pure premium customers that will benefit us, so we won't be able to tell. It's not something we can predict. We don't put it into our pricing, we watch for it. But having a more preferred customer, we believe, will benefit us.

Great. And just one last question, could you provide us with kind of what the competitive landscape looks like right now in the direct agency distribution segment?

Heather, if you want to add the agency part, you can. I think, overall, every or most companies increased rates a fair amount in 2016. Our philosophy has been for a long time and we tried to do it; we can't always do is take small bites of the apple and continue to stay ahead of trend. We are seeing a couple competitors raising rates higher than we are. And so we believe that the hard market will continue to persist, I think especially after the hurricanes, so I won't speak of specific competitors. We obviously watch that from a competitive landscape, but there's a couple of companies that clearly need some rate in the system, and we believe that will have the hard market continue. Competitors from the agency side, do you have any?

So as we think about the competitive landscape on the agency side, it clearly varies across the country in terms of which competitors are relatively well positioned in different regions. We actually even think about it at the agent level. So as we are thinking through where our opportunities lie, understanding the competitive set within particular agencies is how we are getting a sense of where we have upsides. So it's a state-by-state, are we - is the market hardening across both the auto and home and what opportunities do we have in that particular agency and that's where our state-based field and sales model is really beneficial.

Great, thank you.

Thanks, Brian.

All right, I am going to take a few questions from the webcast. The first question is from Mark Lane from William Blair. I'm going to do my best to rephrase this just a little bit.

Can you make money on providing third-party insurance on homeowners?

And I'll let you answer that first, and then I'll ask the second piece.

Okay. Yes. We - with our third-party insurance carriers, we receive a commission when we sell homeowners. And remember, what we really - the intent of having multiple unaffiliated carriers and ASI on the direct side was really to extend the auto PLE, so when we've talked about that for years, being able to say that bundled customers, whether or not we write it on our own paper, really the goal is to extend the PLE of the auto.

Right, and then the second piece of the question is…

Including the homeowners policies that we sell through our third-party insurers, our unaffiliated carriers, how would we think about estimating our market share in homeowners?

We don't really publicize that. We've put the income in our service revenue and - but we have - we don't share that data.

Okay. All right. Thank you. So the next question is from Matthew Goetzinger from Fiduciary Management.

So the question is essentially, as Robinsons grow as a contributor to the company, the assumption would be that Robinsons have comprehensive on their auto insurance policies today. So therefore, as comprehensive - as comprehensive grows as a percentage of the coverage that our customers choose, would that change our reinsurance philosophy on auto insurance for the long term?

Yes, we'll always look at our reinsurance. I wouldn't necessarily say all Robinsons would have comp coverage, I mean some would, especially if there's a lien. But we watch the line coverage codes across our insurers. And if we felt that, that was necessary to protect the downside, we would absolutely consider it.

I might add to that. Recognize that within our pricing in states where -- that are more prone to storms, there are what we consider catastrophe loads in the pricing. So we are looking back over history to understand what we think, over a reasonable period of time, is necessary to cover storm losses. So while it's not reinsurance, it is built into those comprehensive rates. And we think we're adequately priced in comprehensive.

All right. Brian can we take a question from the conference call line please?

Of course. Our next question comes from the line of Ian Gutterman from Balyasny. Mr. Gutterman, your line is now open.

Hi, thank you. I had a quick question on Harvey and then I had a question on the agency growth. Tricia, can you walk through a little bit sort of how you came up with your Harvey auto estimate? And I guess what I'm looking at is - I think you even said it in your video - about 500,000 cars damaged in Houston and your market share appears to be about 10% in Texas. Now maybe it's different in the Houston area, but that would imply close to 50,000 cars and your 200-something million implies only 5,000 severity, which seems very low to me. So were you underweighted in Houston or can you give us a sense of what kind of severity you're using per car?

I don't know that we've shared severity. We look at - we model it out. Every storm, we model it out and we look at the share of market within certain areas. And Harvey was a little bit unusual because there were areas that were completely flooded, and we may or may not have had a higher or lower market share in those certain areas; and some were force flooded and some people left. So we literally modeled that one out prior to it happening and ongoing, and the severity was a little bit higher because most of those cars were totaled. But I will tell you, we feel that we have been very adequate in our reserving for both Harvey and Irma.

Is your market share in Houston similar to your market share in the State of Texas?

Yes.

Okay. Fair enough. So on the agency side, obviously the growth has been very impressive and I just want to understand, obviously, some of it is the initiatives you've talked about on the call, but I was hoping maybe you had some data you can share on sort of where customers are coming from? Is most of the priors coming from other agency carriers that ask for too much rate and they're coming to you or do you feel it's sort of the captive guys are moving into the agency? Just anything you can share on that?

Yes. What I would say overall is that it really has to do with the agents shopping for a great product at a great rate. And so when you talk about that, when you think about that hard market, most of those customers are coming - that are within their agency are likely to come to independent agents. I wouldn't say necessarily from captive. I don't know that, but anecdotally I would say people usually go to the captive or go to independent agents, so it's really about us having a different product in the market. So a couple of years ago, when we put out really our first preferred auto product, 83 and now it's evolved to 84, that has really been the shift in our ability to get more customers and having that right price. So we feel like we've got - we were in a good position in 2015, we've continued to benefit from having better product models with better segmentation at a good rate, at a great rate, because we're able to get ahead of that trend. And that's really been so important to us not to rate shock our customers because that ends up upsetting the agents and then they shop around. And that's really where we've been seeing the growth.

Well, that actually kind of led into what I wanted to ask, was I assume - so part of this especially in the - again, I understand the direct is a different environment, but in the agency environment where it seems like we've gotten so used to the, whatever you want to call it, the power of the rater and low man win sort of thing, it feels like some of that business that other people won maybe from you maybe a few years ago ended up being bad business for them. So now it comes back out, they're asking for more rate. You're maybe lower than what they're asking for, but how comfortable are you or how much analytics do you have to sort of see how the performance of that book was for someone else. I guess I'm worried a little bit just that the rate of growth - if someone else was getting it at a 110, and couldn't get the rate they want to renew it and the agency put it back in the rater, can you really get it to a 96 right away or do you have to bring it on with a heavy new business penalty and assume it takes two or three renewals to get into the 96?

Yes, that's a really good question. We have very specific targets in the agency channel per state, new and renewal. So we - when we're looking at this and why I've bee, I think, elated with the 84 model is the fact that we're writing at or sometimes below target for those - that new business. So we do feel good about that. If we didn't, we wouldn't - we would slow down growth. And you've seen us do that because you don't want to have that come back. If someone leaves after six months, that's not a good business model. So we feel very good and very adequate with our rate in the agency side.

Great, thank you.

Thanks.

[Operator Instructions] We have a few more questions from the webcast. The first one is again from Mark Lane at William Blair.

It's essentially around the purchase price for the ASI minority shareholders. Has that already been a negotiated deal and how is that valuation structured?

Yes, I'll take that. Yes, that was part of the original purchase agreement and the price at which we will purchase shares, either next year or in 2021, is a function of the change in the tangible book value of the company, plus the original price per share. There's also a multiplier that is applied to that change in tangible book value that is determined by a matrix of growth and profitability. It can vary from 1 up to a total - a high of two. So in addition to the base share price that we paid back in 2015, you add the change in book value times that multiple from one to two, and that's how the price we pay is derived.

All right. The next question again is from the webcast from Matthew Goetzinger at Fiduciary Management.

The question is really two parts and the first is regarding the launch of UBI in the Commercial Lines market, so using ELD for truck drivers, how aggressive do you anticipate being on premium discounts given the small initial base of existing statistical loss data?

So how we're going to work discounts in the ELD is if a current customer has at least three months of driving data, you will get - you could get anywhere up to 18% discount, because we already have some knowledge of your driving data. If we don't have any knowledge of it, you'll get a 3% participation discount, and then after that three months of data, you could get upwards of an additional 15% for a maximum of 18%. So I'm sure things will change as we know more, but we believe we've gotten out ahead of it, working with some of the vendors. But that's our current structure as we head into 2018.

Great. Brian, can we take another question from the conference call line please?

Of course. Our next question comes from the line of [indiscernible] Asset Management. Joe, your line is now open.

Thank you. Just a kind of a 30,000-foot question, when you guys acquired ASI, I think Glenn Renwick was very clear in saying the only way a success of the deal would be measured is if you're selling more car insurance through agents. Can you talk about whether that measuring stick has changed or morphed at all at this point?

Yes well, you know having worked with Glenn for so many years, I'm sure that was a statement. But it was really - the acquisition of ASI was getting access to those agents who wrote those preferred homes and to be able to bundle the homes. So we've talked a lot about extending auto PLEs. So I think with any good business model, especially as we integrated and have such a good partnership and relationship with ASI, we always want to evolve. But ultimately we want to be able to have as many preferred Robinsons auto home bundled in the agency channel as possible while reaching our target goals - our target margins. So yes, it's evolved, but I think it's evolved with what Glenn would have had in mind as well.

Thank you.

Thanks.

It would appear that we have no further questions, and so that concludes our event today.