旅行家集團 (TRV) 2016 Q2 法說會逐字稿

Welcome to the second-quarter results teleconference for Travelers.

(Operator Instructions)

As a reminder, this conference is being recorded on July 21, 2016.

At this time, I would like to turn the conference over to Ms. Gabriella Nawi, Senior Vice President of Investor Relations.

Ms. Nawi, you may begin.

Thank you, Tina.

Good morning.

Welcome to Travelers' discussion of our 2016 second-quarter results.

Hopefully, all of you have seen our press release, financial supplement and webcast presentation released earlier this morning.

All of these materials can be found on our website at www.travelers.com, under the Investor section.

Speaking today will be: Alan Schnitzer, Chief Executive Officer; Jay Benet, Vice Chairman and Chief Financial Officer; and Brian MacLean, President and Chief Operating Officer.

They will discuss the financial results of our businesses in the current market environment.

They will refer to the webcast presentation as they go through prepared remarks and then we will take questions.

In addition, other members of senior management are in the room including: Bill Heyman, Vice Chairman and Chief Investment Officer; Michael Klein, Executive Vice President and President, Personal Insurance; Tom Kunkel, Executive Vice President and President, Bond and Specialty Insurance; and Greg Toczydlowski, Executive Vice President and President of Business Insurance.

Before I turn it over to Alan, I would like to draw your attention to the explanatory note included at the end of the webcast.

Our presentation today includes forward-looking statements.

The Company cautions investors that any forward-looking statement involves risks and uncertainties and is not a guarantee of future performance.

Actual results may differ materially from those projected in the forward-looking statements due to a variety of factors.

These factors are described in our earnings press release and in our most recent 10-Q and 10-K filed with the SEC.

We do not undertake any obligation to update forward-looking statements.

Also in our remarks or responses to questions, we may mention some non-GAAP financial measures, reconciliations are included in our recent earnings press release, financial supplement and other materials that are available in the Investor section on our website.

Now, Alan Schnitzer.

Thank you, Gabi.

Good morning, everyone.

Thank you for joining us today.

This morning, we reported second-quarter operating income of $649 million or $2.20 per share and an operating return on equity of 11.6%.

Our underwriting results remained strong as reflected in our 93.1% combined ratio for the quarter, despite the impacts of higher catastrophe and non-cat weather-related losses year-over-year.

For the six months, we reported operating income of $1.35 billion or $4.52 per share and an operating return on equity of 12%.

Given the cumulative impact of eight or so years of declining interest rates on our investment portfolio and also relative to the fresh lows in risk-free rate, we feel terrific about the returns we continue to generate.

Our expertise in risk selection and pricing, our thoughtful and consistent investment strategy and our active approach to capital management all come together to enable us to achieve our objectives of delivering superior returns and creating shareholder value.

Turning to production, the second quarter was another example of the successful execution of our marketplace strategies.

We're very pleased with the record volume of premium that we wrote in the quarter.

We continue to be very pleased with the modeled returns on the business we are writing.

In our commercial businesses, retentions remained high and renewal premium change was positive.

Underneath that, consistent with our very granular approach to execution, in middle market, we retained 90% of the premium from our best-performing accounts with only a modest decline in renewal rate change on that business.

On our poor-performing accounts, we achieved renewal rate change in excess of loss trends.

You might have noticed that renewal rate change in domestic business insurance was slightly positive as compared to slightly negative in the first quarter and renewal rate change in middle market improved by a full point.

We don't forecast rate and we're not calling a bottom, but the significance of this to us is that the market remains remarkably stable.

We're generally successful in achieving our written return objectives.

In terms of new business, we are actively seeking and finding opportunities that meet our return thresholds.

We delivered strong new business growth in our commercial businesses, including 10% growth in new business and domestic business insurance.

In personal insurance, net written premiums grew 9% to a record of over $2.1 billion, with accelerating momentum in both auto and homeowners.

New business was up 21% year-over-year, driven by the success of the Quantum product.

Quantum Auto 2.0 continues to meet our financial expectations and is successfully enhancing our market position.

Our success in the marketplace across all our businesses really speaks to the value of the talent that we have in the home office and in the field and their ability to leverage our data and analytics and work with our distribution partners to deliver great products and services at fair prices to our customers.

Franchise value matters.

Before I turn the call over to Jay Benet, I would like to acknowledge and thank all of our colleagues in the claim organization for their extraordinary efforts in handling the claims of our customers affected by severe weather and fire so far this year.

These experts are on the front line for us delivering on the promise we extend with every policy that we write.

It has been a very active year so far for them.

Their responsiveness and expertise are great for our customers and an important competitive advantage for us.

In Canada, where we saw significant wildfires in the Fort McMurray area, both our customers and we, benefited from the work we've done to replicate our US claim model, which enables us to respond to large-scale events with our own highly trained professionals and resources, without resorting to independent adjustors.

So all in all, solid results for the quarter and the first half, particularly in light of the weather and wildfires and continued excellent execution in the marketplace.

With that, I'll turn it over to Jay Benet.

Thanks, Alan.

As I did in the first quarter, I'd like to start by saying that we were pleased with our results, operating income of $649 million and operating return on equity of 11.6%, despite their being lower than in the prior-year quarter.

These reductions did not result from significant changes in underlying business trends, notably, as indicated on Page 4 of the webcast, the quarter-over-quarter decrease in operating income was impacted by, among other things, PCS cat events and lower fixed income -- net investment income.

Losses from PCS cat events that met or exceeded our threshold for recording as cats was $79 million higher after-tax than in the prior-year quarter.

While losses from PCS cat events that fell below our cat thresholds, which we recorded as part of non-cat weather-related losses within underlying underwriting results, were $56 million higher after-tax than in the prior-year quarter.

Fixed income NII was $21 million lower after-tax, primarily due to the continuing low interest rate environment as we had anticipated in the outlook section of our first-quarter 2016 10-Q.

Everything else was pretty much a wash.

We continued to experience net favorable prior-year reserve development, which totaled $288 million pretax or $192 million after-tax this quarter, up from $207 million pretax or $133 million after-tax in the prior-year quarter.

In business and international insurance, net favorable development of $138 million pretax or $94 million after-tax primarily resulted from: better than expected loss in domestic workers comp for accident years 2006 and prior and accident year 2015; better than expected loss experience in domestic general liability for accident years 2011 and 2015; and better than expected loss experience in our European operations, partially offset by an $82 million pretax or $53 million after-tax increase to environmental reserves.

In bond and specialty insurance, net favorable development of $150 million pretax or $98 million after-tax, primarily resulted from better than expected loss experience in fidelity and surety for accident years 2010, 2013, and 2014, and better than expected loss experience in GL for accident years 2007 through 2011.

Year-to-date on a combined stat basis for all of our US subs, all accident years have developed favorably or had de minimis unfavorable development.

In addition, other than commercial multi-peril, all product lines have developed favorably year-to-date, while CMP has developed unfavorably by $42 million year-to-date.

You may recall my telling you, in the first quarter that there was an offsetting favorable development in the property product line, as middle market property losses that had been recorded in the property line at year-end 2015 were subsequently determined to be CMP related.

Operating cash flows of $443 million were very strong, particularly recognizing that they were net of the $524 million payment we made this quarter as final settlement of the PPG asbestos litigation.

We ended the quarter with holding company liquidity of $1.75 billion.

All of our capital ratios were at or better than their target levels.

Net unrealized investment gains increased to approximately $3.6 billion pretax or $2.3 billion after-tax, up from $2 billion and $1.3 billion respectively at the beginning of the year, while book value per share of $85.73 and adjusted book value per share of $77.61 increased 7% and 3% respectively also from the beginning of the year.

We continue to generate much more capital than we need to support our business.

Consistent with our ongoing capital management strategy, we returned $747 million of excess capital to our shareholders this quarter through dividends of $197 million and share repurchases of $550 million.

Year-to-date, we returned over $1.5 billion to our shareholders through dividends and share repurchases.

Before turning the microphone over to Brian, there is one additional topic I'd like to cover, our cat reinsurance coverage.

In addition to our corporate cat aggregate XOL treaty, which had renewed at the beginning of the year, our cat reinsurance includes cat bonds that are specific to the Northeast US along with the Northeast gen cat treaty.

With respect to the cat bonds, one that had provided up to $300 million of coverage for certain losses expired as scheduled in May 2016.

Given our current risk profile, we decided not to replace it.

With respect to the remaining cat bond, which will not expire until May 2018, the attachment point and maximum limit were reset as required annually to adjust the modeled expected loss of the layer within a predetermined range.

For the year beginning May 2016, we will begin recovering amounts under this cat bond if losses in the covered area for a single occurrence reach an additional attachment of that amount of $1.968 billion.

The full $300 million coverage amount is available on a proportional basis until such covered losses reach a maximum $2.468 billion.

Finally, our Northeast gen cat treaty was renewed on July 1. This treaty provides up to $800 million, part of $850 million, of coverage subject to a $2.25 billion retention for certain losses arising from hurricanes, tornadoes, hailstorms, earthquakes and winter storm or freeze losses from Virginia to Maine.

A more complete description of our cat reinsurance coverage, including a description of earthquake and international coverage, is included in our second-quarter 10-Q, which we filed earlier today as well as in our 2015 10-K.

While the total cost of these treaties is small in relation to our operating income, we were able to save some money by not replacing the expiring cat bond.

With that, let me turn the microphone over to Brian.

Thanks, Jay.

In business and international insurance, we're pleased with the results this quarter.

In domestic business insurance, we achieved very strong retention along with a modest increase in renewal rate change versus the first quarter in higher levels of new business.

Our strategy has not changed as we strive to retain our best-performing accounts, get rate where needed and write new business when it meets our target returns.

This has been our strategy for some time.

As a result, we continue to be viewed as a stable and financially strong market by our distributors and customers.

Along with the meaningful competitive advantages that we deliver, this consistent leadership position is resulting in additional growth opportunities for us.

Turning to the financial results for the segment, operating income was $393 million, with a combined ratio of 97.5%.

Cat losses for the quarter were $143 million net after-tax, which included $61 million from the Canadian wildfires.

Excluding the Canadian fires, cat losses were slightly higher than the prior-year quarter.

The underlying combined ratio, which excludes the impact of cats and prior-year reserve development, was 95.5%, up 2.4 points compared to the second quarter of 2015.

The loss ratio increased by 1.7 points driven primarily by non-cat weather, which was higher than both the prior-year quarter and our expectation.

The expense ratio was up 0.7 of a point over last year, driven primarily by commissions as the second quarter of 2015 benefited from a one-time favorable adjustment.

Net written premiums for the quarter were in line with the prior-year quarter, with domestic business insurance premiums up about 2%.

Given the returns that we are generating in this business, our focus continues to be on retention.

So, we were very pleased that retention remained at 85% for the quarter.

Renewal premium change was just over 2 points, including renewal rate change that was slightly positive.

New business of $525 million was up 10% versus the second quarter of 2015.

Turning to the individual businesses within business insurance.

Beginning with Select, we achieved retention of 82% and renewal premium change of nearly 6%.

Rate change was down slightly versus the first quarter of 2016, while exposure was in line with recent periods and new business was up slightly year-over-year.

In middle market, retention remained very strong at 87%, with renewal premium change of about 1.5 points, in line with recent quarters.

In terms of new business opportunities, given our leadership position in the market, submissions were up year-over-year.

Accordingly, we had more opportunities to quote on the types and classes of business we find attractive and to grow our new business while maintaining our return thresholds.

In other business insurance, retention of 82% was up 1 point year-over-year, while renewal premium change was 0.5 point and new business of $154 million was consistent with the second quarter of 2015.

As we've mentioned in the past, other business insurance includes our national property business.

While there continues to be some rate pressure in this business, returns remain attractive.

As a result, retention is our priority.

As you can see on Page 13 of the webcast, excluding national property, renewal rate change for other business insurance remains positive and was relatively consistent with recent quarters.

In international, net written premiums for the second quarter were down about 12% quarter-over-quarter.

Excluding the impact of foreign exchange, net written premiums were down about 9%, driven by disciplined underwriting actions in our UK business along with lower economic activity impacting the marine and energy lines of our Lloyd's business.

New business of $107 million was up 67% year-over-year, driven primarily by Optima, our new strategic personal lines auto product in Canada, which was modeled after our US-based Quantum Auto 2.0 product.

So all-in, a good quarter for the segment with strong production and profitability.

I'll turn now to bond and specialty insurance where we had another terrific quarter.

Operating income of $202 million was up significantly from the prior-year quarter, driven by a higher level of net favorable prior-year reserve development.

The underlying combined ratio of 80.9% remains exceptionally strong.

As for top line, net written premiums for the quarter were in line with the prior year, as growth in management liability was offset by lower surety premium.

The lower surety volume was due almost entirely to favorable one-time reinsurance impacts in the prior-year quarter.

Across our management liability business, retention remained at historically high levels and new business volumes were up, as we continue to execute our strategy of retaining our best accounts and writing more business in our return adequate product segments.

We continue to feel great about this segment's performance.

In personal insurance, results were again strong.

We remain particularly pleased with our ability to add top line growth at appropriate financial returns.

Net written premiums for the quarter of $2.1 billion were an all-time high, with double-digit growth in agency auto, while growth also accelerated in agency homeowners and other.

In both products, we continue to see strong retention and new business levels.

Agent and consumer receptivity to Quantum Auto 2.0 remains exceptional and importantly, a significant amount of Quantum Auto 2.0 business are now coming through their renewal cycles.

We're pleased with the retention rates we're seeing on those policies.

In homeowners, our growth momentum is building due to cross-over benefits from Quantum Auto 2.0 as well as pricing changes and improvements to agent and customer experience.

Turning to profitability, operating income for the segment was down compared to the prior-year quarter, due primarily to lower net favorable prior-year reserve development.

The underlying combined ratio of 89.5% was in line with our expectations, but up about 1 point from the prior-year, driven by the impact of the significant new business volume in recent years and normal quarterly volatility in weather and other loss activity.

In agency auto, the 101.3% combined ratio included 2.7 points in cat, primarily resulting from hailstorms.

Weather also impacted the underlying combined ratio with 1.7 points of non-cat weather-related losses, which is approximately 1 point higher than both our expectations and the second quarter of 2015.

The remaining increase in the underlying combined ratio was driven by the high levels of new business that I just mentioned.

As we've noted before, the Quantum Auto 2.0 business is priced to our long-term target returns and loss experiences performing in-line with our expectations.

Turning to agency homeowners and other, underlying financial results remained strong.

The second-quarter underlying combined ratio of 78.2% is well within our target return levels as we continue to execute our disciplined underwriting and pricing strategies.

So, overall for personal insurance, we feel great about the growth and the financial returns that the business is generating.

Before I turn it back to Gabi, I would like to take a moment to recognize Doreen Spadorcia, who announced her retirement from Travelers effective September 1 of this year.

Over the past 30 years, Doreen has played an important role on our management team.

We wish her well as she enjoys a new chapter of her life.

With that, let me turn it over to Gabi.

Thank you, Brian.

Tina, we are now ready to start the Q&A portion of the call.

If I could ask the participants to limit yourselves to one question and one follow-up, please.

Thank you.

(Operator Instructions)

Randy Binner, FBR & Co.

I wanted to talk about the underlying combined ratio in business and international.

I think in the comments and in the press release, you attributed the increase of 2.4% to non-cat weather, but as well as loss cost not being covered by earned pricing.

Can we get more granular on that?

Because we're at a point in the cycle where not covering loss cost with earned trend is potentially an issue.

Just wanted to get a better detail on that.

Sure, Randy, it's Alan, I'll take that.

So the underlying combined ratio in that segment, you're right, was down 2.4 points.

Part of that was commission and that was a year-over-year comparison thing, nothing going on there.

The other piece of that was in the loss ratio.

The majority of that does come from non-cat weather.

There is a relatively small piece of earned rate versus loss trend coming through.

One of the reasons we showed you Page 4 in the webcast was to try to put that in a little bit of context.

So, really, what's impacting year-over-year is the weather.

You get -- as we've said before, we've tried to make clear that rate and loss trend is one small piece of what goes on, but there's also taxes, expenses, base year, mix, the impact of changes in claims handling, things like that.

Everything else as we would've expected was a wash.

So, there's nothing in that earned rate versus loss trend that was surprising to us or that we think shouldn't have been evident from the written information we've provided over the last year or so.

Is that -- is the piece of it though that's the loss trend versus earned pricing, is that -- I think you're saying that it's not surprising you, but it does seem to be changing.

Is it a bigger piece of the equation than it has been in the last few quarters?

It's the earned impact of what we've been disclosing as written over the past year.

So it's earning in as we would have expected.

We called it out just in the interest of transparency.

But, again, nothing that surprises us.

Given where our written and earned returns are, nothing that we're not comfortable with.

Then, just finally on this.

So then back -- sorry, go ahead.

I was just going to say, as I said in my prepared remarks, Randy, we're very comfortable with the modeled returns on the business we're writing and so just to give you a little bit more context.

Then with non-cat weather, this would be truly episodic, nothing trend wise you're seeing in that?

That's correct.

Nothing trend wise from that.

It's the normal volatility in non-cat weather.

All right.

Great.

Thank you.

Thank you.

Kai Pan, Morgan Stanley.

First question is on the pricing, slight uptick in the second quarter.

Could you give a little bit more granularity in term by lines or by account size, what do you see underlying trends there in terms of both by -- from a supply and demand side?

What do you see -- could potential be, the upside to pricing or as downside to that pricing?

Yes.

Kai, this is Brian, I'll start -- a couple of broad dynamics.

I think, not surprisingly auto is the line where we're seeing the most positive pricing movement.

Comp is I think aligned with the most pressure.

That's really consistent with -- in workers comp, if you see what's coming out NCCI and other competitors, that's not surprising.

From an account size perspective, larger accounts under more pressure.

So, we've talked a lot and others had talked about large property business.

We think that's somewhat consistent with the returns that have been in that business but some pressures there.

I think even larger casualty business is under a little more pressure than the middle and small type of business.

I don't think from an industry perspective there's anything appreciable.

But, that would be the broad strokes.

Okay.

Then a follow-up on the reserve side.

It looks like reserve continued to be strong.

I just wanted to focus on -- there's three areas.

One is that, the worker's comp release in 2015, is that sort of like -- is that too early for a long-tail line of business?

Secondly, the $82 million environmental charges, it looks like we're having those amounts for at least for the last five to six years.

Just wondering any -- why wouldn't you book, if you know the pattern is there, why wouldn't you book a big charge upfront?

Then the third piece is on the pretty strong, bond and specialty continued release.

Any underlying loss cost trends there would be appreciated?

Thank you so much.

Thanks, Kai, this is Jay Benet.

Let's try to take them in order.

So as far as the worker's comp, you're absolutely right, we have a very long-tail line of business.

You look very, very carefully at things that happen in the short-term and don't necessarily react to them.

In the case of the workers comp, what you're seeing here is favorable development in 2015 that related to a relatively short component of it, relating to medical, where medical bills come in quickly, they get resolved quickly.

The assumption that we had put in there for medical inflation was just too high.

So, while the line itself is a long-tail line, there are components of it that are more short-tail.

That's primarily what we reacted to there.

As far as the environmental piece is concerned, I think if you look at us as well as the industry, you'll see a similar pattern, if you will, of people trying to, at various points in time, estimate what the ultimate is going to be for environmental losses and recognizing that you're dealing with something that is a very difficult kind of reserve to truly evaluate.

So in doing so, you have to make various assumptions as to what the future is going to look like in terms of new claims, new policyholders, what the average frequency and severity might be associated with things that have been reported or not reported.

It's those assumptions that you're constantly updating as time goes on.

So what we disclosed -- I'll point you to the 10-Q, what we disclosed this year is actually very similar to what we've disclosed in prior periods and what you've seen other companies disclose, that there really hasn't been a change of environment, per se.

Things have gotten a little better.

But the rate of them getting better has just been less than what was previously assumed or hoped for.

That's a chunk of the reserve addition.

In addition to that, we're always subject to judicial rulings or other things that take place episodically in various jurisdictions.

There was one that came out of the Northwest that increased modestly some of our claim cost there, that we also recognized.

Then finally, on the bond and specialty, this was a quarter for a fairly large prior-year reserve development benefit.

I wouldn't say there was anything unusual about it.

This is, as you know, it's a high severity, low frequency business.

It's one where information emerges slowly.

As it emerges, you're always evaluating whether or not it's blossomed to a level where you really feel action is appropriate.

So there are things that have been taking place, whether it's in items related to the financial crisis or whatever that we've been watching.

But things got to a point this quarter where we recognized that it was appropriate to take some reserve action.

It wasn't any one product line.

It wasn't any one particular claim or things like that.

It was pretty much spread across the whole management liability and surety book of business.

Thank you so much.

Michael Nannizzi, Goldman Sachs.

Can you talk a little bit about -- we talked about non-cat weather and BI and personal auto.

Can you talk about non-cat weather in the homeowners' book?

How that has looked recently?

Yes, Michael, this is Brian again.

In this quarter, similarly there was some but not to the same degree.

So, just to quantify it, because I said about 1 point on the auto side.

It was probably about 0.5 point on the personal property side of some impact there.

So, some but not as dramatic of that, again, that, specifically the PCS event.

Got it.

I guess, I'm just curious, because it looks like in the last few quarters, especially in BI, we've talked about non-cat weather impacting the underlying and a driver for the year-over-year comps being negative.

Then on auto, it looks like it was a factor this quarter as well.

But when I look at my estimates and your results over the last several quarters, homeowners has been consistently better.

I think better even relative to your outlook as you've disclosed in your Qs.

So, I'm just trying to understand how wide the impact is potentially disproportionate in these areas, is it because the nature of the risk is different?

Or reinsurance?

Or something else that's causing non-cat weather to be -- it looks like a tailwind in one potentially and a headwind in the others?

Yes, Michael, this is Michael Klein.

I think what you're seeing is a combination of factors in the home product and weather is one of them.

In the case of this quarter, I think what you see is, to Brian's point on the net impact, it's a couple moving parts underneath.

You've got the weather impact, but we also have sort of underlying experience on fire, non-weather water losses, that again, also fluctuate quarter-to-quarter.

So when Brian talks about weather and other normal quarterly fluctuations, that's really what he's talking about.

This quarter was an example of that, where we had some other non-cat weather, again, non-weather water, fire experience that came in better than expected.

Importantly, for the year-over-year comparison this quarter, there actually was a significant individual risk fire loss in the second quarter of last year that helped the quarter-over-quarter comparison, just to give you a little bit of color on that.

Got it.

That's very helpful.

Thanks.

Then, maybe Brian, on BI.

So, it sounded like from Alan's comments that the lift in the expense ratio was commission related.

I'm guessing that means that should continue?

Because, I'm also trying to square that.

Because at 33 --

No.

No, it's not?

Okay.

No.

The commission -- last year in the second quarter, we had a takedown in our contingent commission accrual that reduced the number.

So, that created a year-over-year difference.

The delta is not really in the base commission.

The delta's in these other things that cause quarter-to-quarter variances.

Yes.

Got it.

But I guess, my point is so the 33.3% expense ratio in this quarter, forget the comps, but just the absolute number this quarter, is that how we should be thinking about it?

Because it sounded like there were some one-offs last year, but this quarter, is that how we should think about the expense ratio from here?

It's Alan, Michael.

I guess I would say that we try not to forecast on these individual pieces, but what I'd suggest is maybe to go back over the last, I don't know, somewhere between two and four quarters and average the number and maybe use that as a proxy for a go-forward number.

But there's always going to be some volatility quarter-to-quarter.

Okay.

I got it.

Just because it screen relatively high when I go back.

So, it sounds like maybe there's some volatility.

I was just trying to understand what drove just some episode of a higher expense ratio that should cause that to normalize as you're suggesting it should.

But maybe I can follow-up afterwards.

Thanks.

Ryan Tunis, Credit Suisse.

My first question is for Brian.

I was just hoping for a little bit more color on what's been driving the solid new business growth in middle markets.

I think you mentioned submissions are up.

But I guess over the past couple quarters, it has been pretty strong.

Just to know what you're seeing on the new business front?

I think it's a couple of factors.

We believe, as I tried to say in my comments, driven by the fact that I think we're in a pretty strong position in the marketplace.

We've been executing in a very stable and effective way, I think, for a long period of time.

We've got some real competitive advantages.

The overall market is pretty stable but there has been some pretty well publicized dislocation.

At least a couple of very large carriers have been very open about taking profit improvement actions.

So that's created some specific opportunities.

But I think we think more broadly has also created a little bit of a mood in the marketplace of flight to quality.

So like I said, we've seen our submission activity up a decent amount.

We've also, consistent with that, been driving a message to our organization not strategically to do anything different with returns but to be more active and to be out there quoting more.

So, just to be blunt about it, we're pushing our place pretty hard right now to be real active in the marketplace and get out the quotes and not compromise on returns but try to meet the opportunity of the submission activity.

That's been paying off in, specifically, our core middle market commercial accounts business and the construction areas.

Ryan, I would add to that, that part of our motivation behind pushing a little bit more activity is just where returns are -- product returns in the marketplace after five or so years of price increases.

So there's a lot out there that's potentially attractive to us, consistent -- this is really important, consistent with our return objectives.

So, we're not making any compromises on that.

Okay.

Yes.

That makes a lot of sense.

I guess my other question was just on -- just looking at the margin trajectory in agency auto and how to think about that on a go-forward basis?

I guess year-to-date, there have been a couple things that have been -- that seem to have been weighing on the loss -- on the accident year loss ratio some.

There's the new business drag, then there's the elevated non-cat weather.

But something that we were a little bit -- that we thought was interesting was looking in 10-Q, it looks like the language looking into 2017 was that margins would remain kind of where they've been.

I'm just curious, as you lap the elevated new business and you get rid of the lower non-cat weather, why would margins not be improving as you look out into 2017?

I think, just a couple quick comments, then I'll toss it to Michael.

Certainly, we view the non-cat weather as episodic.

So we think that should absolutely return to normal levels.

On the new business dynamic, the good news is that we continue to write some significant levels of new business through that product.

So, we feel great that we are building real embedded value in that portfolio.

But it is going to take longer.

The stronger the production top line is, the longer it is going to take for that to really work through the loss ratio.

The other point I'd make is, the Quantum 2.0 product also was designed to have a lower expense ratio and a little bit higher on the loss side.

So you have to look at the combined in total to get that product.

Ryan, if I could just clarify something.

One of the things we actually added to the Q this quarter was, in the outlook section, we had previously spoken to underlying underwriting margins.

We thought it was important to add combined ratio because people often use those terms interchangeably.

As you know, the combined ratio is a function of the premium volume, whereas the margins are just dollars.

Just to clarify, what we're saying here in the Q is, for auto, which I think is what we're talking about, the early part of 2017, the Company expects underlying underwriting margins, the dollars, to be slightly higher because of what's taking place with writing higher volumes of business that's profitable.

But writing a higher percentage of new business relative to the base is also causing a slight uptick in the underlying combined ratio.

So you have this feature here where the dollars are going up at the same time that the combined ratio is going up.

So, I just wanted to clarify that.

Yes, that's helpful.

I think I was a little bit confused by that.

Thanks, guys.

Appreciate it.

Ryan, the other factor just to note, it's not all this new business impact from what we're writing.

As Michael explained a second ago, we haven't had this in the last quarter or so, this favorable non-weather loss activity.

As we also highlight in the 10-Q, we expect that to return to a more normal level.

So, that is also having, on a go-forward basis, a slight adverse impact.

Thanks.

Underneath it -- this is Michael Klein.

The only thing I would add is underneath it, just to reiterate, right, Quantum Auto 2.0 is priced at the same return level as the book of business has been price to.

So it really is the dynamic of preponderance of newer, call it -- think of it, as younger business, right?

The portfolio is getting younger because of the outsized growth that we've seen over the last couple of years.

But the long-term return targets for the product remain consistent with where they've been.

Amit Kumar, Macquarie.

Two quick questions, probably deserve follow-ups.

I apologize if I am still not clear.

Just going back to the discussion on BI and what Randy was asking.

If you ex out all of the moving parts and just focus on the underlying discussion on the negative loss cost trends versus on pricing, is it episodic, is what you're saying?

Or is this the usual trend line, i.e., you've talked about this in the past that over time this equation will get flipped and now we're at the point where it has flipped.

I'm just trying to understand that.

Yes, it's not episodic.

There is a trend there.

But I think what we've told you in the past is, there is some danger to looking at that very narrow definition of rate versus loss trend.

Obviously, anybody could look at rate, which is sort of around zero and loss trend, which is something north of zero on a written basis.

It has been moving in that direction for some number of quarters.

So as anybody -- as we would've expected, as that earns in over time, there will be some negative pressure on margin.

But what we've said in the past is, you've got to look at that in the context of all the other moving pieces that impact underlying underwriting margin.

Things like, we had a tax impact year-over-year this quarter.

There's expenses.

There's base year.

There's mix.

All those other things go into that number.

What we've shown you this quarter is all of that comes out to about a wash.

I guess I'd also point you to the outlook, where we give you a sense of, over the next 4 quarters or so, that we would expect the underlying underwriting margins, obviously, subject to the volatility of weather and whatnot, to continue to be broadly consistent.

The other thing I would emphasize -- this is Brian -- is that there is a lot of things in there that we have to react to.

There are other things in there that we are actively managing.

So you shouldn't have the impression that we're just sitting by.

As Alan just said, pricing is at roughly zero this quarter rate and loss trend is something north of that, that we're just conceding that margins will receive.

We're managing our mix.

We are taking actions in underwriting selection.

We are taking actions in claim execution, which are all impacting our loss trend.

Then we're reacting to weather volatility, et cetera, that might be moving through the business.

So, there's no denying that the sheer arithmetic of that narrow piece, the loss trend and the rate, right now is a negative.

We've said in the past, if it's relatively modest and it continues to be relatively modest, then there are all the other impacts that are going through.

Got it.

That's a fair comment.

The second question I had was just going back to the discussion on Quantum 2.0.

In the past, when you have talked about loss cost trends, frequency was pegged at 0.5.

Severity was at 2.5.

Has there been any evolution in that number?

Or is it still the same?

Nope.

We didn't see anything different in the core underlying trends as what was going on in the auto book.

We would be at exactly those numbers you just said, about 0.5 point of frequency, 2.5 of severity.

As we said in the Quantum Auto 2.0, specifically, still running consistent with our loss expectation.

Got it.

That's all I have.

Thanks for the answers.

Charles Sebaski, BMO Capital Markets.

I guess I'd like to follow-up again on -- following Amit's on Quantum 2.0 and the agency auto book.

The clarity -- I realize there's an additional disclosure in the Q, but how you guys expect the auto book to evolve out?

I guess the commentary of early 2017 being plus or minus 2016, which is plus or minus 2015 and you say that your underwriting to the same return dynamic.

But I guess this has been a book in flux that coming out of 2011 and 2012, where it's been underperforming and the introduction of Quantum 2.0 -- just, where is a run rate on what you should -- what we should expect?

There's so many moving parts, but it seemed like this should be a story that was improving.

I guess the commentary seems to be that it's maybe gotten as good as it's going to get.

Is that right?

So, the message absolutely is not, this is as good as it's going to get.

The core, from a profitability perspective -- let me start with a top line perspective, what's the story on Quantum 2.0.

We have been pleasantly surprised with the traction the product has gotten in the business and in the marketplace.

We clearly went in with an expectation that this would enhance growth and we've done better than our original expectations there.

You can see the numbers.

So we feel good about that.

We're not sure where that's going to go exactly in the future, but we think that we're going to be able to sustain some pretty decent amount of growth going forward.

From the profitability perspective, just, as is true of all business that we've ever written in this product, it matures over the years.

So there is a 10-year impact.

As Michael just said, the younger tenured business is typically not as profitable.

As we write the business, we're targeting a tenured longer-term return.

We're on track to do that.

So as this portfolio matures and as Quantum 2.0 -- as we get more and more of the aging of the book, we believe the trend should absolutely be improving profitability on this business.

Okay.

I guess, on that, could you give us some additional insight on the tenure or the curing of the auto book?

On, what is the average policy life for you guys on this?

I guess when I think of some commercial accounts that might be longer, it seems that auto still is a quicker turn product.

So, this introduction of new growth, which has really been strong, what's the timeline for curing on an auto book?

How long do you, on average, keep auto policies?

Charles, I don't know that I'll get into the specifics of the policy life expectancy, et cetera, but to give you a little bit more color -- this is Michael -- on the aging of the auto book.

You have to really think about both components of what's going on here, right?

So, first, you need to think about a vintage and think about a book of business we write in a given policy year and that new business that we write in that term, that comes on at a higher loss ratio early.

Then as we renew it, it improves.

If you look at the production statistics that we've been disclosing and the high levels of new business, what's happening inside the portfolio then, which is the second piece, is on average the portfolio is getting younger, right?

If younger business carries a higher loss ratio, that's what we describe as the increase in the combined.

So when you look at the outlook for the first half -- I'm sorry, the second half of 2016, we say the combined ratio will be higher due to the impact of the higher new business levels we've been writing.

That is the underlying dynamic that's flowing through the book.

That does continue into the first half of 2017, it just gets offset by an expectation of return to normalize non-cat weather levels.

So the underlying dynamic of the book getting younger, meaning less tenured, right?

Because we've had the accounts not as long, continues into 2017.

Recognize also though, that as we write more Quantum Auto 2.0 business, there is a benefit to the combined ratio from an offsetting improvement in the expense ratio, because it carries a commission.

That underlying aging of the book or reducing average tenure of the book, does continue into 2017.

It just gets offset by, again, an expectation of return to more normalized weather.

That's why the outlook reads the way does.

Because it's an assumption, not an expectation.

Sorry, it's an assumption.

Yes, sorry, Jay.

So, one other point on that.

Again, as Michael said, we're not going to give you explicitly the target we have for life expectancy on policies, we think that's part of our competitive conversation.

We clearly accept that hitting the expectations of retaining the business is important, which is why, in my comments, I emphasized that as we're going into year two and in some cases year three for the Quantum Auto 2.0 product, we feel really good that we are hitting the retention rates we expected.

So whatever our expectation was for life expectancy, along with our statement about losses are progressing, so is the retention of the business.

You can look at our long-term retention and begin to deduce some kind of policy life expectancy by just doing the arithmetic.

Thank you very much for the answers.

Jay Gelb, Barclays.

I was hoping to get a bit more insight on the pace of share buybacks.

It's slowed over the past few quarters on a linked-quarter basis.

The slowdown in share buybacks was actually a little greater than the reduction in earnings for the first half of 2015 to the first half of 2016.

Can you give us a bit more insight on that?

Jay, this is Jay.

I think you're looking back at a period of time where, as we've said in the past, over time, share buybacks are going to be enabled by earnings.

While there's pension contributions and whatever, it's the earnings of the place over time that will drive what the total share buybacks and dividends are.

You saw, I think you even asked the question in the past, can we continue to buyback at levels that are higher than our earnings?

Well, this is the flip side of it.

This is just an indication of the timing element of when we look and see what the future quarter projections are.

They obviously come out to be whatever they will be.

We're just picking numbers in the particular quarter that we're doing the share buybacks over a longer term view to say in this particular quarter $550 million sounds like the right number.

So I wouldn't read anything into it in terms of a change in philosophy or a change in execution.

It's just going to be what we refer to as the number quarterly variation in the policy and a disciplined that's been set for a long period of time.

I see.

Okay.

Thank you for that.

Second one, what was the overall impact in the combined ratio in 2Q from non-cat weather?

On a combined ratio basis?

We are looking.

Okay.

While you're -- if you have that number.

I also wanted to ask about -- I know the UK leave vote is not significant relative to Travelers in its overall business, but if you have any thoughts on that, I think that would be real helpful for us.

Sure.

In terms of Brexit, the most significant impact to us will be whatever it means for overall economic activity.

The amount of premium that we write out of the UK that relies on passporting is very, very small.

It's something like 5% of our overall international premium and about half of that is through Lloyd's.

We assume that one way or another, Lloyd's will address the passporting issue, as will we by the way.

A long way to go to see how that's going to work out and what the ultimate arrangements are going to be.

But other than the overall macro impact, we don't expect any significant impact to us.

Appreciate that, Alan.

Thank you.

Paul Newsome, Sandler O'Neill.

Congratulations on the quarter.

I wanted to ask about the competitive landscape in the personal lines business.

We've had a few management changes of recent note, like places at Progressive and State Farm.

I'm wondering if the components of who is being competitive out there have changed in the last year?

Paul, this is Michael Klein.

I would say, broadly speaking, the answer to that question is no.

I mean there -- competitive dynamics are very local in the business as well.

So while there have been changes in management at some national competitors and frankly, some local regionals as well, the broad competitive dynamics in the business remain consistent.

When we look at the rate filing monitoring that we do in the business to assess the rates that carriers are taking in particular states in auto or home, we don't see a lot of shift in behavior there.

So I would say, the competitive dynamics are broadly consistent.

So, the general idea that the agency folks keep losing share and the regionals and the direct people are the most competitive?

Is that still kind of really the case?

I think if you look at the latest A.M. Best data, you continue to see some incremental shift in share by channel.

Interestingly, I think consistent with the last couple of years, the majority of the gains you're seeing in the online direct channel are actually coming from captive and less and less from independent agency.

But broadly, I think those trends continue.

Great.

Thanks.

Appreciate it.

This is Jay Benet.

Somebody asked the question about what was the impact on the combined ratio of non-cat weather?

I will give you the number, but I also have to give you a little bit of a preface to it.

We are parsing things very thin here in talking strictly about that.

So, if you asked the question, what's the impact of non-cat weather on the combined ratio on a consolidated basis, we would say it's 150 basis points.

But I'd also point out that, if you asked the question about, well, what about other losses?

That goes in a different direction.

What about mix?

That goes in a different direction.

So we will answer your question, but there's lots and lots of things that are taking place here that make the one number not necessarily indicative of the trends or of what's going to -- or what the expectation is.

So I would just urge you to go back to the Q and the outlook to see what all of this, to us, actually translates into.

Jay Cohen, Bank of America, Merrill Lynch.

Just a follow-up.

The language in the Q on the commercial, on the business insurance segment, where you extended your commentary on the margin into 2017, again, the broadly consistent language.

I assume that's largely because the first half of 2016 had a high level of loss activity, non-cat weather?

Is that fair?

That's right.

There's some large losses in non-cat weather that, in the outlook, we would expect would return to normal levels.

Assume.

We would assume.

Got it.

Thanks.

Then the second question, maybe for Bill Heyman, because he hasn't had a chance to speak yet.

(laughter) Obviously, as we would expect, the non-fixed income returns are always going to be variable.

Can you give us a sense, just given how the portfolio is constructed today, what is an expected return for that portfolio at this point?

Thanks, Jay.

I think an expected return for non-fixed income ought to be in the high single-digits pretax.

If you look at this quarter, we made a bit more in real estate than we figured.

Real estate has two components: rentals from properties we own and NII from real estate funds.

Hedge funds were below what we expected, but positive.

The big variable was private equity, where the portfolio was almost flat, maybe a few million dollars.

We would've expected about $50 million pretax.

Now, what's interesting is, the cash flow from that portfolio has remained at historic levels.

It simply hasn't contained as much NII in the past.

So we still -- we would aim for a double-digit pretax return on these assets.

If we didn't feel we could do that, we probably would reduce our allocation.

But I think a fair expectation is high single-digits, pretax.

That's great.

Thanks so much.

Larry Greenberg, Janney.

I'll apologize for going back to this, but just listening to some of the chatter that's out there this morning, there's something of a view that you guys elaborating on the pricing loss trend relationship in business insurance suggests that you're seeing or thinking something new.

If I hear what you're saying, it really doesn't appear to be the case.

I just want to be sure that I've got that right.

Yes.

Larry, it's Alan.

Let me be really clear about it.

There's nothing that's changed.

There's nothing that's surprising us.

Everything is consistent with what we've reported to you on a written basis over the last four quarters.

The real reason we put it in there is we wanted to be completely transparent about it.

It's a very small piece.

But we thought, gee, if we didn't put in there, you or somebody would have said to us, gee, how can this not be happening because we're looking at your written numbers over the last three or four quarters.

So, I'll just reiterate.

It's small.

It's within our expectations.

It's wrapped up with all the other things that are typically moving around in a quarter.

It's about a wash in terms of the year-over-year impact.

Nothing -- no message intended in terms of anything out of the ordinary, completely expected, completely consistent with what we've seen on a written basis.

We continue to be very pleased with the modeled returns on the business that we are writing.

Great.

Then just my follow-up.

Jay, in your discussion on non-renewing that cat bond, I think you said that the reason was due to your current risk profile.

I'm just curious about that comment.

If you could elaborate a little bit on that?

Primarily our cat program is one that protects capital not one that really deals with income protection.

If you look at where the limits are and things of that sort and where the attachment points are, it's pretty high up there.

So, in any given period, we're looking at the entirety of it, where our property exposures are.

Just given the way the book has evolved over the last few years, we just felt that this wasn't really adding much on a risk versus cost basis.

So we decided not to renew it.

Okay.

Thank you.

Great.

That's it.

That concludes our call for today.

As always, we, in Investor Relations are available for follow-up questions.

Have a great day.

Thank you.

Thank you.

Ladies and gentlemen, that does conclude the conference call for today.

We thank you for your participation and ask that you please disconnect all lines.

Thank you.

Have a good day.