Welcome to the Progressive Corporation's Investor Relations Conference Call. This conference call is also available via an audio webcast. [Operator Instructions] In addition, this conference is being recorded at the request of Progressive. If you have any objections, you may disconnect at this time. The company will not [ph] make detailed comments in addition to those provided in its quarterly report on the Form 10-Q and letters to shareholders, which have been posted to the company's website, and will use this conference call to respond to questions. Acting as moderator for the call will be Matt Downing. At this time, I will turn the call over to Mr. Downing. You may begin.

Thank you, Candy, and good morning. Welcome to Progressive's conference call. Participating on today's call are Glenn Renwick, our CEO; and Brian Domeck, our CFO. Also on the line is Bill Cody, our Chief Investment Officer. The call is scheduled to last about an hour. As always, our discussions on this call may include forward-looking statements. These forward-looking 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 this call. Additional information concerning those risks and uncertainties is available in our 2012 annual report on Form 10-K and the quarterly reports on Form 10-Q issued during 2013, where you will find discussions of the risk factors affecting our businesses, Safe Harbor statements relating to forward-looking statements and other discussions of the risks, uncertainties and other challenges we face. Each of these documents can be found via the Investors page of our website, progressive.com. Candy, we're now ready to take our first question.

[Operator Instructions] Our first question is from Vinay Misquith, Evercore.

The first question is on capital management. We've seen a lower level of share repurchases in the last few months. Curious as to what your views are on a special dividend going forward in light of your excess capital at the subsidiaries?

Sure, I'll let Brian take that. But recognize that if we had any significant statement to make, we would have made it in our release. These things that we study, I certainly understand the question. I don't think there's any issues there. To the extent that we have anything to release, we will discuss that with our board in December. But there's really no specifics. Brian can maybe comment a little bit on share repurchase and any other things relative to the dividend, but we have no statement today.

Sure. I continue to characterize our capital position as strong. We ended the end of the third quarter with $8.8 billion of capital, that's including both debt and equity. Ended October with closer to $8.9 billion of capital, and that includes the retirement of $150 million of debt during the month. So even in a month where we retired $150 million of debt, we had over $200 million of comprehensive income. So strong month in October. Certainly, for the year, our capital creation has been strong with comprehensive income of over $1 billion so far. I would characterize third quarter capital management activity as pretty modest. We did repurchase $54 million of our debt securities. Share repurchases in the quarter were pretty modest, 1.5 million shares, about $38 million worth of shares. So far this year, about $165 million year-to-date. So given all of that, our capital creation for the year has been pretty strong. We've mentioned that we do have the variable dividend that will be payable in late January, early February. If we were to quantify that, just using through October's results, that variable dividend would be about $0.43 a share or about $260 million, approximately those numbers. Obviously, that will change based upon the results of the next 2 months. So we have that as an upcoming activity. But currently, our capital position is strong. And certainly, we always want enough for operations and to grow the business. We feel comfortable with that. We always want enough for contingencies, and it looks like we have escaped an active hurricane season for this year. So we're beyond that. So it gives us more confidence in our capital position. And I think, as Glenn mentioned, we continue to remain committed to what we say, is that we will return under-leveraged capital on the timing of it and the form. It's always under discussion to be determined. But in terms of consideration of a special dividend, yes. In terms of what we will do, more to follow.

So just as a follow-up to that. Should the fact that we have not heard anything about a special dividend mean that it's less likely to happen near term?

Well, we would only announce it and talk about it after it has been considered by our capital investment committee and the full board. So it's only after discussions amongst them that we'd ever comment on it or make note of it. So it could be about the timing of when or if anything would occur.

Sure, okay. The second question is on competition. We've heard a large competitor recently announced meaningful rate decreases, and all new [ph] business and the rollout of a new product. Curious as to what impact you think it's going to have on your business. And do you think that also leads to other competitors trying to reduce their cost structure and reduce their commission structure?

I assume I know who you're referencing there, and I think that's probably way too early to totally speculate on it. And I don't mean to dodge the question. Do I expect others to reduce their cost structure? I would we've been very clear in including it at our IR meeting of the importance we place on cost structure. And to us, what we believe is a fairly direct correlation between low cost and some of the results we've been able to achieve. So it won't surprise me if we see others take actions that suit their needs, whether that be commission or otherwise. And I understand relative to certain segments of the market, others will get competitive or even more competitive over time. We are always looking to do that. I think I put in my shareholders letter that, sort of, product design, we don't talk about product design quite as discreetly as maybe others because we're always doing product designs. And at any given point in time, we have rollouts of different degrees of product enhancements, and thus, we are working actively. I even gave you some indication of the things that we'll be in, the ones that we're working on right now. So at any given point in time, we're very dynamic. We're trying to work on rate. And if there is rate opportunities in the marketplace that others see, there's a good chance that we have seen them as well. So it's very hard to sort of just go -- yes, one will be X result versus Y result. I think we'll be extremely competitive. And some of the books that are out there that perhaps are not as well priced are always great opportunities for us. So not a particularly insightful answer to your question because there's really no way to give anything more color at this point in time. That may or may not be a strong player in the marketplace. I think we will continue to be a very strong player in the agency marketplace, especially.

The next question, Mike Nannizzi of Goldman Sachs.

This is Mike Kovac on behalf of Mike Nannizzi. Through the end of the year, with more modest rate action and lapping easier comps, you guys mentioned building on increased demand. And with other direct competitors growing some of them as high as 30% and the uptick in loss cost trends that we saw in the third quarter, relative to earlier in the year. At your current rate level, how do you guys expect to grow? Do you think it will be more Snapshot based or something broader? And if it's broader, does that imply that your rates today are below that of some of these other growers?

Again, always hard to sort of predict exactly on one-for-one. But our demand, I think I've been pretty clear that demand has actually been picking up all year for us. So I'm focusing on the direct side here. Our advertising seems to be working. It seems to be bringing not only good numbers, but good mix of customers. What's really happened, and I highlighted this in the letter is that, that's turning into solid conversions. So on our size and the percentage that we're getting, we're actually very, very comfortable with our growth. I've said before, between GEICO and ourselves, we capture a very significant part of the new market, the new business marketplace, somewhere over 50% of that. So I think that, on absolute numbers, we are in a very good position to see, not only the growth that we're experiencing, but the absolute numbers that are coming in. Do I expect others to compete and increase their intensity from time to time? Yes, absolutely.

And if I could, just a couple follow-ups on Snapshot here. Many of your Snapshot customers have now gone through multiple renewals. What kind of retention data are you learning about this business? I think you talked about -- a little bit about that in the past. And second, of the now over 1.2 million Snapshot customers that you have, what percent of those are sort of new to Progressive? And I know you're finding that your rates for the non-Snapshot customers that you have today are competitive before any UBI-related discounts.

Yes, I think that most of the customers, and I say most, there are some exceptions, most of the customers that actually start with Snapshot are new to the Progressive, because it's really offered as a new business, rating enhancements [ph]. So in fact, recently, for reasons that doesn't matter right now, I started a new policy with Progressive, and that vehicle actually was on a Snapshot basis. With regard to retention, it is absolutely true that our experience is that Snapshot customers are retaining longer than an equivalent mix of non-Snapshot customers. I'm not going to give out a lot of data points simply because, clearly, we know the interest in this is high and some of what we know is probably what we should know. But for those that get a substantial discount, we are seeing retention that can lift as much as 40%. And those numbers are not that shocking when you realize that you're probably taking someone to a price point in the marketplace that is simply not available at a new-for-new comparison. So once you've established yourself with the renewal, it's somewhat, I sometimes call it an absorbing state, you have a rate that truly reflects your driving behavior. No one else can know that in the marketplace on a new quote. So the retention benefit of Snapshot has been one of the #1 issues that we've always been focused on. If we could get people to a rate that was reflective of their pure premium and perhaps lower than what the market would be able to recognize without the availability of that rating variable, then there's every reason intellectually to believe that, that becomes somewhat of an inert state that, why would I shift because I simply can't find that rate or better anywhere else. And that certainly is proving to be true at the overall level and even more dramatic at the higher discount level. Having said all of that, we still have very few customers who have been with us on a Snapshot basis or having had the discount with Snapshot for 5 and 6 and 7 years. So we actually look forward to understanding what the long-term dynamics of those customers will be. And as we look forward, all the interim signs are very encouraging. Did I get all your questions on that one?

Great. The last one was just, are you finding that your -- the non-Snapshot rate that you're offering today is competitive? So when customers are basically not opting, what does it look like for you.

No, that's a good question, too. Recognize that in the marketplace, let's just keep it simple. You go shopping today, you get offered effectively the non-Snapshot rate. So you're getting offered $500 for 6 months -- or let's make it simple, $100 a month, $600. Then you're getting offered the option to take Snapshot with the ability to improve your rate. And after 30 days, we give you the first recognition of that improvement and then a further recognition of the improvement after 6 months. So the actual decision you're making at the point of sale is as, what you might refer to as, our non-Snapshot rate, plus an option. So the rate at the market level is strong and about 36%, slightly up from, I think, the last number I gave you of customers who are shopping on a direct basis are actually opting to take that option, which currently is a free option. It's an option to say, "I'll take the price, but I may be able to do better." As we evolve that product like we evolve all of our product designs, we will likely do things that will be different 5 years from now than what we're doing today and probably sooner. Does that describe the dynamic of the shopping experience?

Yes. So there's not as much -- so you're not seeing really -- I thought there was a Test Drive in there a little bit, too, where people could potentially get the rate before they sign up as well. So just in terms of what most of your customers are basically, you're saying, getting a rate and then using it.

Fair comment. Test Drive was very much designed to do pretty much what you're describing. I would tell you, while very personally hoping Test Drive would be a great opportunity for us to reach out to a different set of customers and not have to have them commit. For the most part, those numbers have proven to be relatively small. And as a result, we will continue to try things. But in the spirit of Progressive, some things we try work really well and some things don't work as well as we expected. I would tell you Test Drive has not produced the numbers that we might have liked. Doesn't make it bad, just means that it's not something I talk about as much when we're talking about Snapshot.

Next question, Mike Zaremski of Crédit Suisse.

This is Crystal Lu in for Mike. I had a question about the reserve development. We've understood that the reserve development was going to be adverse for the first half of the year and favorable for the last half, as management had mentioned in July. But it's running higher than we would have thought in October for the quarter. And so how should we be thinking about the reserve development going forward, given the update you gave in July?

Brian, why don't you comment on the reserve development for this release and any other...

Yes, I'm not entirely [ph] referencing the comments in July, but I'll try to lay a color of where our development has been so far through third quarter, and then again, after the October results. So at the end of the third quarter, we had approximately $80 million of unfavorable development. Call it, approximately 50% in Personal Lines, primarily in Personal Auto, and then another 40% or so in the Commercial Lines products, and then some unfavorable development in our runoff of the professional liability group business that we sold several years back. So the $80 million was fairly comparable to what we had at the end of this first half of the year. So development in the third quarter is actually fairly modest, slightly favorable, if I remember it right. Just a few million favorable for a third quarter. So the third quarter activity in terms of the development pattern was better than, say, we had in the first half of the year. Then in October, we've had favorable development, much of it from actuarial reserve changes in terms of reviews that were done during the course of the month, such that now, at the -- through October, we have roughly $40 million of unfavorable development through 10 months of the year. After the October change, Personal Auto is relatively flat in terms of development for the year. We still have Commercial Lines having some unfavorable development, primarily in our higher limits bodily injury severities where we've had increase in frequency and severity, particularly on the IBNR on that. We still had a runoff in the professional liability group business. That hasn't changed much. But actually, as of October now, Personal Auto was relatively flat in terms of development. A little bit of difference between the channels. Agency has had some unfavorable development, whereas direct auto has had some favorable development, and that really some of the [ph] basis for the reserve changes in the month of October.

My other question is about the investment income. Could you give us an update on how we should be thinking about new money yields versus the portfolio yield?

Bill, why don't you take that one?

Sure. Certainly, the higher rates we've seen since the springtime have helped somewhat. Our overall strategy hasn't changed. So if you look at the marketplace and the type of the securities that we own, hopefully we'll have an opportunity to continue to add some securities at slightly better yields than they have been over the past couple of years. So no material changes. And it's just -- take a look at where the market is for the securities that we own and the type of securities that we have in our portfolio and project that forward.

Next question, Josh Stirling of Sanford Bernstein. Josh Stirling - Sanford C. Bernstein & Co., LLC., Research Division: So 2 questions. The first, obviously, we'll talk about Snapshot, but wanted to talk about your reserve development just a bit more. And from outside looking in, you're talking about service severity decelerating, you're seeing some favorable reserve development. I'm just trying to piece it together. It sounds like it's mostly, if I understand, at least in the Personal Auto side, it sounds like it's mostly coming from more favorable severity playing out than you originally assumed. I'm wondering if you can sort of clarify that, if that's right. And then if that's true, can you give us a sense of the color -- a sense of what severity trends you were assuming when you booked your '12 and '13, like early '13 reserves, and sort of what the magnitude is now of sort of what experience you're actually having and sort of -- as we can sort of think about -- I know, if this is an inflection which -- where we'll end up seeing favorable development play out for another number of months as we go into '14?

Sure, Josh. This is Brian. I'll provide sort of the color on the actual severities that we're seeing. And I characterize severity trends as modestly positive. So in the property damage covers, call it 2% to 4%, and collision more in the 4% to 5%. So property coverage is a little bit higher. Our bodily injury incurred severity, so this would also include reserve changes, is in the 1% to 2% range. And then PIF severity is actually minus 4%. In the PIF severity, obviously, we weren't necessarily projecting a decrease in severity in PIF. Much of that is driven by Florida, maybe some due to the new regulations. And also, in New York, we're seeing a decrease in severity in New York. So those 2 drivers drive the PIF severity. And we certainly were not necessarily in pricing or anything else, pricing for a negative PIF severity. And I would also characterize our estimate of future bodily injury severities was -- particularly when booking reserves and the like was probably -- is greater than 1% to 2% that we've actually experienced so far this year. And that's how I characterize it, principally driven by a little bit lower on BI. And certainly, a negative severity in PIF isn't what we had anticipated. Gary, do you want to add anything to that? Gary Traicoff, our actuary, is here.

Sure. I think Brian gave a good summary. This is Gary Traicoff, Corporate Actuary. The favorable development, for the most part, has been on the case reserves. And of course, that's going to be a severity driver. To Brian's point, when you look at what our severity was, say, a year ago for BI in particular, it was a little bit higher than where we're at today, looking at the 1% to 2% range. And we were booking a number closer to what we have seen historically relative to a year ago. So that's going to be making up a good chunk of that difference on favorable development. Josh Stirling - Sanford C. Bernstein & Co., LLC., Research Division: And I'm sorry, this is really helpful, but you guys reminded me that your statistics, your incurred severity statistics include the reserve changes. So is -- are you seeing a similar deceleration on paids, paid severity changes?

Yes, the paid severity for BI, and think of BI being the coverage where more of the reserve changes factor in, our paid severities are in the 3% to 4% range for BI year-over-year. Josh Stirling - Sanford C. Bernstein & Co., LLC., Research Division: Okay. Got it, got it. And so if -- I don't want to hog the call. I'll ask just one more question on Snapshot. You guys are making progress in a bunch of fronts. I'll ask about your licensing deals. It sounds like you've announced a few more, or at least are willing to announce that you have succeeded in a few more. I'm wondering if you can give us some color on -- if you can't tell us who the companies are sort of -- what sort of companies are signing up with you? What the -- or what the ultimate sort of terms that are being negotiated are looking like? And also, on the technology front, if people who are signing up with you are actually going on and using and sort of executing deals with Xirgo to use the technology that you guys have pioneered with Xirgo?

Hey, Josh, you're right. Some of the companies that we're doing business with, it's not a lot, as I told you before, for right now prefer that we don't use their name, but we did a release with USAA. So it gives you some sort of flavor. These are companies that are very credible. And when they bring their product to marketplace in '14 -- or '15, excuse me, or whenever, then we'll know more about it. So we pretty much told you the terms of the licensing arrangement. We don't get into exactly what they plan to do and what they don't plan to do and other licenses that they have. So other than what I've really said, there's not much more to add there. This may be just totally editorial, but this is perhaps a little bit of just a personal view. When we talk about severity trends in BI, at any time that you run below 5%, that always has my attention, I just sort of -- historically, sort of that's my threshold point. I would just say that as an industry, and certainly, for Progressive, we are all looking at what might be -- might turn out to be one of the biggest exogenous changes in terms of the Accountable Care Act and how that ultimately forces price changes and so on and so forth. So just as all of us are interested in observing the industry, you and us alike, we don't necessarily know what that will bring. But don't be surprised if severity changes are a little bit variable for the next few years as we sort of see how that settles out. And it won't surprise me at all to see things that go from relatively low to a little bit higher and back. I might be wrong on that, but something that we will have our eye on and recognize that, that clearly has some implications to reservings of claims that are not yet paid or influenced by that. But certainly, on a going-forward basis, the incurreds could be interesting. So just keep that in mind as you observe the industry in general that, that might be a significant driver that we haven't seen before.

The next question, Meyer Shields, KBW. Meyer Shields - Keefe, Bruyette, & Woods, Inc., Research Division: In the 10-Q, Glenn, you talk about, I think, a very rapid rebound on a sequential basis in new business in both Agency and Direct, but a slowdown in the year-over-year growth rate of renewals. And I was hoping you could explain how you interpret that.

Yes, I think that's a fair characterization. I think I would be awkward if it was the next quarter, first quarter of next year, when we have this call if I can't sort of report that the renewals are now showing the sort of same trend. I'm not necessarily suggesting double digits. But as we start to see our primary measure, a measure that we have as a primary measure, we have lots of subscore supporting measures. Our policy life expectancy is a lag measure. So it just takes a while for that one to turn. Clearly, we look at other measures that are a little bit more of an acceleration vector. So right now, I feel much more confident in what I can see and report as absolute higher data, which is what I did with regard to the new business, and then gave both a very clear indication. So I think, for gross purposes, 20% business is new, 80% is renewal. So this really needs to flow through to renewal. There's no logical reason why it won't. But I'd like to see that before starting to report on it. My commentary was this is where it must come to roost, and we are starting to see the signs that would make that a true statement. I would tell you, and I can't necessarily explain it with perfect detail, that it's just a little slower than I would have projected. But we're starting to see our renewal rates now head in the right direction. And the other measure that we use somewhat as an earlier proxy, our NPS measures, and those of recent times have started to head up in both channels. So fourth quarter if, as expected, that flows through to renewals, then we'll be in very good shape.

The other thing I'd add to that is in terms of, say absolute renewal application counts, they're somewhat a function of the policies in force, call it, 6 months earlier, right? 6-month policy term. And so if you were to look particularly, say, at Agency where our policies in force 6 months ago were close to 0 or slightly negative, then how many renewal apps you're going to have on a year-over-year basis is somewhat a function of that, pending changes in retention rates and renewal rates. So as our policy in force growth year-over-year was declining, more so in Agency but also in Direct, the absolute growth rates in renewal apps just follows that. Now as we are starting to see some sequential increases in policy force growth year-over-year, that's when we anticipate that 6 months further down the road, we'll start seeing renewal app growth higher and you'll see it. Since our Direct policy in force growth rate is now over 4%, you should see it in the renewal application, absolute counts, growth rates in Direct more than you will in the Agency. Meyer Shields - Keefe, Bruyette, & Woods, Inc., Research Division: Okay. No, that helps. The modest decline in PLE in Agency, is that a function of more new business or the type of new business?

It's not really a function of the new business. I think, for the most part, the decrease in PLE that we have experienced throughout this year is really a function of the rate changes we had made towards the end of last year, which adversely affected retention rates. So that was -- has been what the driver of the decrease in PLE is. And I think Glenn mentioned in his answer to a previous question in terms of the calculations of the PLE, some of it can be a lag measure. And what we are focusing in more and more on our current renewal rates, which we're starting to see increase on a year-over-year basis. But I would say most of the PLE decrease that we experience and comment on is more a function of a rate change as opposed to a mix change.

[Operator Instructions] Next question is Greg Locraft of Morgan Stanley.

Wanted to just understand a bit more on new applications picking up. Any intelligence as to where they're coming from, why are folks engaging with you more today versus before and who's losing market share is what I'm trying to figure out.

Two thoughts there. One would be an idea that I tried to express in the letter, that our rate, when we take it, we're always taking estimates of future rate. And it turns out that perhaps our loss cost did not go up as fast as we had expected, which makes the competitive point of a rate revision. So think of a rate revision lasting, just not officially, let's just say a year, it's a little less competitive right after you've taken rate and somewhat more competitive at the expiry point of the rate revision. Well, the most competitive point is somewhere in the middle. That's a little hypothetical. Lots of things change around that. But we are now approaching sort of what I'll call that more competitive position. So our conversion rate, the demand has actually been very strong. What we're seeing is, as we move through the life of a rate revision and recognize we're talking here about many hundreds of rate revisions, not one, so it's hard to characterize, but we'll keep it simple, that we're in more of a sweet spot of the rate revision. A second reason that we're getting it is, I also referenced in the letter, we have not done this in wholesale. But in some cases, where those estimates were off sufficiently, then the product manager has actually taken actions to adjust. And Florida is an important state for us. It's our largest state, and we've actually taken rate adjustments in Florida. But it all turned out, as far as we can see right now, to be exactly the right type of adjustments. And those have resulted in a good new business flow for us in the state of Florida, which is an important one. So both the overall competitiveness in most states and a little bit more direct action in some states.

The other point I'd make is in the direct channel. In the first half of the year, our advertising spend was basically on par with the first half of 2012. But on a year-over-year basis, our advertising spend in the third quarter and continuing in the fourth quarter is very much greater than what it was in the last half of last year. Because last year, given some of the profit concerns, we actually reduced some of our advertising spend from anticipated levels. Now given the demand that we're seeing, the conversion and the yield, our advertising spend is significantly higher than a year ago, which is aiding the quoting activity and demand in the direct channel. As to where it's coming from, we actually do have the data in terms of who their prior carriers are. I don't have it here with me. But we can follow up on that one for you. Maybe talk about it at the next conference call.

Okay, great. And then just an entirely separate topic and admittedly, incredibly far out. But our auto team here has done a lot of work on the driverless car, the autonomous car. 10 years out is what they say. The technology is there, but adoption is the concern. You guys are, I think, the leaders in the industry on the technology side. Any thoughts around that world, what it looks like, what it means for Progressive, where we're at, et cetera?

A lot of thoughts and sort of be irresponsible not to have thoughts. Unfortunately, sometimes, thoughts and knowledge or certainty are different. So we covered a fair amount of that in the IR meeting. But primarily, to try to give some sense of -- essentially, it sounds like similar to what you're doing, where the technology actually exists and then ultimately how it gets into the fleet. There is another level of thinking going on that suggests, and I'm not going to go into it in great detail, what might be insurable interest going forward in a different world. All of us are very familiar with many other industries, whether it will be film and digital photography. The fact is the world takes a lot more photographs today than they ever did, just you had to figure out what was important in that transition. So we're always looking to see what's important. I would tell you that purely, on a personal basis, I am very excited, not just about Snapshot for what it is and what we all have some knowledge of, but it gives us some very direct insight to the vehicle in gross terms. If today, the driver is the biggest single issue of quality defects in driving, accidents happen, there's quality defects, the driver is largely responsible for those. Knowing what's going on in the vehicle is an important aspect. And our telemetry and knowing that gives us opportunities to think about things differently, what might be insurable interest. There'll be a lot of technology in the car. Some of it will be plug-compatible replacement. Some will have dual fault-tolerant dual processing and so on and so forth. There's all sorts of things in there that might come about that represent insurable opportunities going forward. Those are appropriately somewhat blue sky. But as that sky gets clearer and clearer, I'm not sure that I think it will be exactly with anybody else's estimated time frame. But we're not in the business of necessarily worrying about whether we hit that on the nail or not, we'll just continue to refine estimates over time. But along with those estimates, we'll also be starting to ask, what does an insurer look like in the future? And I suspect that's a lot more to do with information that comes from the car than the classical segmentation with regard to characteristics of the driver.

Next question, Ian Gutterman of BAM.

I wanted to follow up again on PLE. I guess, Glenn, where I'm a little confused with PLE being down so much this year as I thought, Snapshot, as you mentioned, is supposed to help PLE. And if Snapshot is becoming such a big part of the mix, I mean, if we're -- it looks like we're into the double digits overall. On Direct, maybe it's as much as 20%, I'm guessing. Why hasn't that been more of an offset to the PLE pressure for pricing?

Well, it's certainly in the mix. I don't have that broken out. I'm not trying to avoid the question, just don't have it broken out that way. I'd also tell you that Snapshot customers are primarily sort of, or at least have historically been more new customers. So there is -- to the extent that we ever experienced a significant lengthening of PLE, our measures will ultimately have to experience that before we sort of incorporate it in. So to a prior question, I said we really don't have a lot of 5- and 6-year customers. It may be that we are under it, it may be, want to stress those words, that we're underestimating PLE on Snapshot customers. Any time you take on a very different mix, i.e. a customer who has a discount that wasn't previously available to them, we will only be able to experience what we experience. We won't know necessarily what the long-term state of that is. So I hear your question. I understand, and it is affecting PLE. But the effect is probably not as great as you think right now. Will it get greater over time? We may consider breaking that out a little bit. But I will tell you that it's certainly an encouraging vector for the change of PLE over time. The other thing with PLEs is, recognize the ones that we talk about are not necessarily mix adjusted. So we can talk about the PLE for certain classes of customers. We're not going to do that. But it really does require a little bit more of a diagnostic into the PLE to see what type of customers we're getting. Obviously, our non-standing customers are not producing a lot of extension of PLE, but we're very happy to have them, as long as they're paying their full freight in terms of loss costs. Other types of customers, our multiproduct customers, those that are getting a homeowner or those that are getting a Snapshot, are absolutely showing the changed dimension that we think we can incorporate into our overall PLE. And having said all of that, we are perfectly happy to have some short-term PLE customers, as long as they are ultimately profitable customers for us. So it would be a little bit more of a 'maybe we'll do it at a future IR meeting,' but we really need to break the PLE down to considerably smaller segments to be able to see the change. In the macro, you may not see the change as readily.

Okay, understood. Let me just ask one clarification on that. Would it be fair to say Snapshot has benefited PLE more than it's benefited your actual retention? I guess, maybe I'm being simplistic about it, but I thought PLE was policy life expectancy. So if you can incorporate expected benefits versus retention, which is more of an actual. So why wouldn't, if you expect those people to stay for 5 or 6 years because of Snapshot, why wouldn't that be in the expectancy of the PLE?

Expectancy is an estimate based on renewal events. So we -- it's not our -- we just don't make a priority estimate. It is based on the observable events of renewal at a 7 month, 6 month renewal. 7th - are they here at the 7th month after the 6th month. So we didn't go in with an artificial or -- we think these customers will last x long and therefore, price them to that. We are actually doing it on loss cost improvements, on Snapshot, not necessarily PLE. As the PLE is measurable, we will incorporate that in.

Okay. Now I think I finally got it. And then my other question, if I could, I'm just curious what initiatives you're working on for -- it seems there's an arguably permanent change in how younger drivers think about cars and more of the ridesharing and some of these other services, which lead to sort of new insurance challenges. Are you working on sort of new products for that type of young consumer that look different than traditional insurance? Are those on the market yet or are they still in pilot? Just curious how you're thinking about that issue.

Nothing that is overly specific to those characteristics. Again, at our most recent IR meeting, we talked a little bit about the different characteristics of those younger drivers. You're correct. They're getting licensed later in life, so on and so forth, which actually has some benefits from a loss cost perspective. We're very, very focused on understanding millenniums and what their characteristics might be and the media that they are responding to. So yes, it sort of permeates our thinking, but there's no specific product that we're planning to come out with at this point in time that is focused on that group specifically.

There don't seem to be any additional questions at this time. Candy, we'll turn it back over to you.

Thank you. That concludes the Progressive Corporation's Investor Relations Conference Call. An instant replay of the call will be available through Friday, November 29, by calling 1 (800) 945-6737, or can be accessed via the Investor Relations section of Progressive's website for the next year. Thank you.