美國家庭壽險 (AFL) 2018 Q3 法說會逐字稿

Welcome to the Aflac Third Quarter Earnings Conference Call. (Operator Instructions) Please be advised today's conference is being recorded. I would now like to turn the call over to Mr. David Young, Vice President of Aflac Investor and Rating Agency Relations.

Thank you, and good morning. Welcome to our third quarter call. This morning, we will be hearing remarks from Dan Amos, Chairman and CEO of Aflac Incorporated, about the quarter as well as our operations in Japan and the United States. Then Fred Crawford, Executive Vice President and CFO of Aflac Incorporated, will follow with more details about our financial results, outlook and capital management. We will then open our call to questions.

Joining us this morning during the Q&A portion are members of our executive management team. In the U.S., Teresa White, President of Aflac U.S.; Eric Kirsch, Global Chief Investment Officer; Rich Williams, Chief Distribution Officer; Al Riggieri, Global Chief Risk Officer and Chief Actuary; and Max Broden, Treasurer.

We are also joined by members of our executive management team in Tokyo at Aflac Life Insurance Japan. Charles Lake, President of Aflac International and Chairman, Representative Director; Masatoshi Koide, President and Representative Director; Todd Daniels, Director and Principal Financial Officer; and Koji Ariyoshi, Director and Head of Sales and Marketing.

Before we start, let me remind you that some statements in this teleconference are forward-looking within the meaning of federal securities laws. Although we believe these statements are reasonable, we can give no assurance that they will prove to be accurate because they are prospective in nature. Actual results could differ materially from those we discuss today. We encourage you to look at our annual report on Form 10-K for some of the various risk factors that can materially impact our results. The earnings release is available on the Investors page of Aflac's website at investors.aflac.com and includes reconciliations of certain non-GAAP measures.

I'll now hand the call over to Dan.

Thank you, David, and good morning, and thank you for joining us. Let me begin by saying that the third quarter of 2018 concluded a great 9 months for Aflac and well positioned us to achieve the goals we set for the year. As you saw from the earnings release yesterday, I am pleased that we expect to come in at the high end of the upwardly revised 2018 adjusted EPS outlook. Fred will provide more details in his comments shortly.

Aflac Japan, our largest earnings contributor, generated strong financial results. In yen terms, Aflac Japan's pretax profit margin was ahead of expectations, both for the quarter and for the first 9 months.

Aflac Japan's third sector sales results of a 2.6% decrease was consistent with our expectations. This reflects sales growth in our new cancer insurance and a decline in our medical insurance sales. As medical sales came off a strong year bolstered by refresh of core products, our distribution turned its focus this year on our new cancer product as they tend to do when a new core product is introduced. We expect similar results in the fourth quarter and continue to anticipate third sector new sales growth for the year to be in the low-single digits.

As I've said many times, our focus is on defending and growing our leading third sector franchise. We are indifferent as to the mix of medical and cancer sales as long as we're satisfying the needs of the consumer and our distribution partners.

Regarding distribution. We had meaningful production across all channels. Traditional agencies have been and remain vital to our success.

Our alliance partners has also had significant contribution to our sales results, with such an extensive distribution network, including Japan Post, 20,000-plus postal outlets selling our cancer insurance. We are solidifying our goal to be where people want to buy insurance. Our focus remains on remaining -- on maintaining our leadership position in the sale of third sector products that are less interest rate-sensitive and have strong and stable margins. We will continue to refine our existing product portfolio and introduce innovative new third sector products to maintain our market leadership.

Turning to Aflac U.S. We are pleased with our financial performance. The pretax profit margins exceeded our expectations, both for the quarter and for the first 9 months. Our third quarter new annualized premium sales, together with our sales outlook, keep us on track to achieve the lower end of our anticipated 2018 new annualized premium sales growth of the 3% to 5% increase. As you think about U.S. sales, keep in mind that Aflac is different from our peers and that the majority of our sales come from independent sales agents. We are fortunate to have such a strong independent field force, which is truly unique within our industry. These career sales agents are best positioned within the industry to reach and, therefore, succeed with smaller employers and groups with fewer than 100 employees.

Aflac's independent career agents have been the driving force behind Aflac's ability to dominate the smaller case market. And I continue to believe this market is ours to grow.

We continue to expect higher growth in broker sales. Our team of broker sales professionals has made great strides in successfully strengthening Aflac's relationship within the large broker community. We continue to make investments in our group administrative platforms to support this growth. These investments are taking hold. While broker business is a smaller percentage of our overall business, it has been driving most of the growth for the year.

Equally important is our extensive network of independent sales agents who work with local and regional brokers. It is very encouraging that as brokers looked for solutions for their clients, they found that Aflac's product portfolio helps fill those needs. Brokers are looking to partner with a strong brand like Aflac and leverage our outstanding track record of experience and extensive fulfillment capabilities.

Aflac's expert agents in our independent field force had demonstrated their ability to accelerate growth by working with brokers and broker sales professionals.

Across the company, we continue to invest in digital initiatives designed to address pain points in the development, sales, administration and customer experience related to our products. These initiatives take many forms, and you see it coming through the segment results in the form of elevated, near-term expense ratios and venture investings. I am pleased with the progress both in Japan and in the United States and our ability to continue these investments without losing focus on driving strong pretax profit margins.

Turning to capital deployment. We remain committed to maintaining strong capital ratios on behalf of the bondholders, the shareholders and the policyholders. At the same time, we're balancing our financial strength with increasing the dividend, repurchasing shares and reinvesting in our business. We continue to anticipate that we'll repurchase in the range of $1.1 billion to $1.4 billion of our shares in 2018, assuming stable capital conditions and the absence of any compelling alternatives.

Of course, it goes without saying that we treasure our record of dividend growth. With this quarterly declaration, 2018 will mark the 36th consecutive year of dividend increases. As we move through a period of market volatility, our dividend track record is a nice reminder of the relative stability of our business model and earnings. As we communicated earlier this year, the Board reserves the right to examine the dividend on a quarterly basis. But we have reset our review cycle for the dividend increase to the first quarter.

Looking ahead, we believe our strong earnings growth will continue to reflect the underlying earnings power of our business in Japan and in the United States as well as our prudent approach to deploying excess capital in a way that balances the interest of all stakeholders. At the same time, it will reinforce our dedication to delivering on the promise that we make to our policyholders.

I'll conclude by reiterating how proud I am of our management team, the employees, our sales organization in Japan and in the United States, as they worked incredibly hard to generate results that we've shared. Quarters like this only fuel my excitement for Aflac's future and what we can accomplish.

Now I'll turn the program over to Fred who will cover the financial results. Fred?

Thank you, Dan. Our earnings results for the third quarter performed as expected and consistent with the recent trends and guidance we provided at this year's Financial Analyst Briefing.

For the quarter, adjusted earnings of $1.03 were driven by strong pretax margins, both in Japan and the U.S. We recorded a favorable tax item in the quarter of $8 million or $0.01 per share related to the filing of our 2017 tax return and associated true-ups. The yen-dollar exchange rate had very little impact on the quarter's results as compared to the 2017 period. The strength of our performance year-to-date and stability in our margins gives us confidence we will come in at the high end of our revised guidance range of $3.90 to $4.06 per diluted share for 2018.

Underlying our outlook is continued strength in investment income and benefit ratios, offset somewhat by a planned increase in expenses as we continue to invest in future growth and efficiencies across the enterprise.

Turning to our Japan segment. We reported a pretax profit margin of 20.1%. Our total benefit ratio was in line with our guidance range at 70.7%. And we expect to end the full year in the 70% range. Our benefit ratio in the quarter reflects the continued shift in business mix, positive claims trends in our cancer business and associated reserve adjustments. Our expense ratio in Japan ticked up to 20.8%. And we expect the ratio to be just under 22% for the fourth quarter and 20% to 21% range for the full year.

Turning to the U.S. results. Our overall pretax profit margin in the quarter was 20.7%. Our total benefit ratio came in modestly below our guidance range at 50.6%. And we expect to end the full year in the 51% range. Much like our experience in Japan, along with the generally favorable claims trends, we are seeing the effects of business mix with a shift towards naturally lower benefit ratio and higher expense ratio product lines.

Our expense ratio in the U.S. increased to just below 35%. We expect accelerated spend in the fourth quarter with a projected expense ratio in the 38% range. The increase is primarily driven by the post-tax reform investments earmarked for the fourth quarter and timing related to advertising spend. We continue to see full-year 2018 coming in around the 35% range.

Investment income performance both in Japan and the U.S. continues to deliver strong results. The year-to-date outperformance has been driven primarily by the accelerated growth of our floating-rate portfolio, further benefiting from higher LIBOR rates. As we have previously commented, we elected to lock in the majority of our U.S. dollar investment hedge costs in Japan for 2018 and are likely to take a similar approach in 2019 as our outlook is for continued upward pressure. We will discuss our tactical approach to the Japan-U. S. dollar portfolio and associated hedging in more detail on our December outlook call.

Commenting briefly on our corporate segment. Investment income is influenced by the movement of capital and the building liquidity position at the holding company. As Max discussed at our Financial Analyst Briefing, we continue to make progress on managing our economic exposure to the yen, while lowering enterprise-wide costs associated with Japan's U.S. dollar investment hedging. We accomplished this by entering into offsetting hedge position at the holding company with the financial impact recognized through investment income line of our corporate segment. We have built this offsetting position to approximately $1.7 billion in the quarter, contributing $9 million on a pre-tax basis to the quarter's earnings. We continue to refine the program with the appropriate limits, controls and holding company liquidity buffers.

We ended the quarter in a strong capital position. Japan's solvency margin ratio is estimated in at the approximately 975% level. Our estimated U.S.-only risk-based capital ratio at quarter end stands at roughly 825% and includes an estimate of the full adoption of U.S. tax reform. We are projecting RBC in the mid-600% range for the year-end 2018. Our estimate includes an extraordinary dividend to the parent and moving $500 million in excess capital from the U.S. insurance entities.

We ended the quarter with approximately $2 billion of capital and liquidity at the holding company. We've set aside $1 billion as a capital buffer and $500 million for liquidity in support of holding company derivative positions. As we work to internalize Japan-U. S. dollar portfolio hedge costs through offsetting hedge positions, our minimum liquidity standards will increase accordingly.

Overall credit conditions and asset quality remain very strong with little in the way of impairments in the quarter. Including dividends and share repurchase, we returned $521 million to our shareholders in the quarter. We repurchased 7 million shares of our stock for $322 million and remain tactical in our approach. As Dan mentioned, we are maintaining our current outlook for a range of repurchase of $1.1 billion to $1.4 billion in 2018.

Finally, in October we issued JPY 53.4 billion in long-dated senior notes. We are one of the few foreign corporate issuers in the recent years to raise yen debt with all tenures being a minimum of 12 years. This is a tribute to our commitment of maintaining a strong credit profile. Among other things, proceeds from the issuance will help bolster holding company liquidity and support of corporate hedging activities, while not materially impacting leverage.

I'll now hand the call back to David to begin our Q&A session. David?

Thank you, Fred. Now we are ready to take your questions. (Operator Instructions) Operator, we will now take that first question.

(Operator Instructions) The first question is coming from the line of Nigel Dally of Morgan Stanley.

Got a question for either Fred or Max just on internalizing the hedge. Moving to the -- and how quickly you're looking to ramp up that program? And on the back of that, whether we should expect further declines in corporate expenses?

Max mentioned this at the FAB, and I'll let him add any color he has. But we talked about the potential of managing down roughly 10% to 25% of our overall hedge costs as we move through 2019. And that's the goal and how we look at it. We haven't necessarily sized the limits on the program. Right now, as I mentioned, we're at $1.7 billion. We would expect to build from there but then stop and make sure we can monitor and have all of the appropriate controls and so forth in place and liquidity in place. And we may build from there. We'll provide more detail on our December outlook call as to how we see 2019 rollout, Nigel. And so right now, just know there are $1.7 billion currently. We would expect to build from here. But we'll give more detail on the outlook call.

Okay. And just for the follow-up. Question on Japan sales. This year's cancer products seem to lead to a very quick but short spike in sales. In the past, it seemed like a positive impact of new product introductions and it lasted several quarters. So interested in why the pattern this year seems to be somewhat different to what we've seen in the past.

Koji, do you want to answer that? Or do you want me to?

(foreign language) First of all, at the beginning of the year, we were expecting that our target will be slightly negative to flat. (foreign language) And in the second quarter, we have focused on the sale of cancer insurance and have conducted TV commercials as well as sending out direct mails. (foreign language) And as a result, our sales performed beyond our expectation, and we ended up at almost equivalent level as the record high sales. (foreign language) So as a result, in the first half, we have changed our third sector outlook to low single-digit. (foreign language) And in the third quarter, because there has been acceleration of sales to the second quarter among agency channel, that has impacted. (foreign language) And as a result, there has been some decline in sales. (foreign language) However, throughout the entire year -- for the full year, we are expecting to achieve the target that we said we will achieve at the beginning of the year.

Nigel, what I would say in addition to that, that really simplifies it is that we have never done as big a mail ad as we did, and we have never had TV ads that talked about the mail-out. They all hit in the second quarter. And because of the TV ads, it skewed the business to come in more in the second quarter, whereas they would have been slower to respond to the mail-outs, but the TV ads increased their ability to perform and go ahead to submit the applications. So we saw that big surge, which we warned everybody about in the second quarter that it wouldn't carry over, and there's a third. And sure enough, the mail-outs and the business that took place with that with the TV ads came in, in that second quarter and was a much shorter tail than it normally would've been, had we not been doing the TV ads.

The next question is coming from the line of Jimmy Bhullar of JP Morgan.

I had a couple of questions. First, just on U.S. margins. They've been pretty strong this year. I think for the first 3 quarters, the benefit ratio was down 100 basis points. How much of this is a mix shift towards lower benefit ratio and higher expense ratio products versus just favorable claims experienced that might not continue?

I would say that there is definitely an issue related to the -- a good issue related to the mix of business. And -- but it's a bit more subtle than you're finding, for example, in Japan, where you have a more dramatic shift in earned premium between first sector and third sector. In the U.S., the shift that's going on essentially in terms of our in-force book is that the cancer in-force book of business, which naturally carries a higher benefit ratio and lower expense ratio is starting to become a lower percentage of the overall in-force as we saw more accident in hospital. But particularly, the growth in the group side and that group business tends to be a lower benefit ratio-, higher expense ratio-type of business. And that mix has been taking place fairly steady over the last 5 years or so. And that is contributing to a generally better benefit ratio and a little bit of upward pressure on the expense ratio. But I would say, in this particular quarter, you have dynamics really related to natural claims experience. The claims levels coming in this quarter were lower than expected, and that makes its way through our numbers, including modest IBNR adjustments associated with it, et cetera. So there is really both going on, and -- but I would say, what is of a more permanent nature is that, that slow steady shift in in-force business towards lower benefit ratio product and higher expense ratio.

Okay. And then on Japan sales, like we expected cancer sales to come down from 2Q, just given the new product introduction. But what's surprising is the medical business sales continue to drift lower, and I think this year's going to end up being one of the lowest years that you've ever had in terms of medical sales. And I understand that some of it is that people are pushing the cancer product this year, but sales have been sort of declining even prior to that. So just -- when are you planning on introducing a newer medical product? And do you feel that you're sort of been marginalized in that business versus where you used to be, just given maybe a more proactive push by competitors into that market?

(foreign language) Well, first of all, in terms of our medical insurance market, there are many companies that are entering this market because of the low-interest rate. What I mean by that is that these companies are shifting from savings-type product to medical insurance. (foreign language) And also, another background is that some of the domestic insurers are converting their medical riders to base policies. As a result, the share is being distributed to more companies. (foreign language) And in our case, as we say, our medical insurance has gone a cycle -- a year cycle ever since we launched a new product. On top of that, from the second quarter and on, our agencies really focused their sales on cancer insurance sales. As a result, medical insurance sales declined. That's exactly what you mentioned. (foreign language) And these cancer insurance and medical insurance are our key products. We are revising these product in a very well-planned manner. (foreign language) And in terms of medical, because the competition is very severe, there is no change to our policy to the renewing our product in a shorter cycle. (foreign language) So we do have a plan to revise our medical insurance.

And I would make one other comment about that is that regarding medical insurance, one specific area that we're taking a lot of focus on is the nonstandard medical. We've been very successful in that market. Everyone is offering new products now. And so what we're looking at is the nonstandard medical. We used to call -- we call it our general lever, but we're looking at now to see how we think we can have an even better product in all of our medical area including that.

And we should assume that you'd introduce a new product sometime early next year, given sort of the typical timing of new product launches?

Japan, y'all want to answer that?

(foreign language) Yes, we cannot explicitly because of the FSA filing issue, but then yes, that is somewhat what we are thinking of.

The next question is coming from the line of Mr. Tom Gallagher of Evercore.

Dan, just a question about your Japan sales also. Can you talk a bit about the dynamic of where you see sales going over the next couple of years considering you have this bigger concentration of cancer sales now and presumably, a bigger concentration of sales through Japan Post? I think in the past, when you've pivoted to the new product, you've been able to sell it through effectively all of your distribution, but obviously, that won't be the case with Japan Post. Does that pivot make it more of a challenge for '19 and '20? Or something you think you can manage through and still be able to grow?

Well, I think the question is a good one. And what I would argue is Japan Post sales going to be down, or is it going to continue to grow. And right now, we're in negotiations to see where they'll end up next year. But they certainly have not penetrated that market, and it has great potential. So I'm encouraged about still future growth, continued sales with Japan Post. It's a little too early to tell yet, but I hope to get to those numbers with Japan later on. I also think that the gap product that we are introducing -- the gap product is really not just a new product. It's really an expanded distribution. It's early on, remember it will be more like internet sales directly to consumers and predominantly, younger people. So the premium will be lower, but what it can do is set up a base of policyholders that we can then add new riders and new policies to as they age and as they have more need for insurance coverage. So we're watching that carefully and hoping that, that will grow certainly with the brand and the high name recognition that we've got. It will play into our hand as well.

The other thing that I'll touch on is first sector protection policies. It's too early to tell in terms of what it will look like in '19. But I can tell you that we're very happy with the profitability that it's showing because it's driven by the mortality underwriting, and it is very good for us. And so that's another area that even though when we talk about sales of the third sector, this has the profitability along the same lines as third sector. Not that big yet but has the potential to grow our market as well. So you add those together. And we'll give you a lot better look at the outlook call on exactly what's going to happen. But I did want to give you some insight into it.

That's helpful. Thanks, Dan. And Fred, just my follow-up. This shift that you've seen into cancer and weaker medical. Will that change the -- obviously, that's changing the mix in terms of earned premium for third sector to some degree. Can you talk about the margin differentials if there is one on both benefit ratio and expense ratio between those 2 products? And whether that's going to -- we're going to see a change in third sector at all in terms of margins?

Yes, yes, there is really not. The margin -- the pricing dynamics on cancer and medical are substantially similar in nature. There is no significant difference in the margin targets that we price from the product end. The products also behave very similarly in terms of benefit ratio expectations and expense ratio dynamics. What I would note, not necessarily your question, but what I would note is that we are starting to talk more about first sector protection products, and while small, we had a nice tick-up in our first sector protection sales in the quarter as we introduced a new Whole Life policy. Because the type of first sector product we're selling now is not a savings-oriented product, it's also been priced to the very similar margin dynamic as we experience on third sector. And so we continue to kind of broaden our product mix and offer up more product selection to our -- particularly our core agencies, which includes first sector protection, and I just want to note that, that also has a very similar margin dynamic. And you'll see that incorporated into our projections as we go forward. We're starting to see first sector, that runoff of first sector and the paid-up first sector product start to slow now. And we're hoping that here over time, we turn a corner and start seeing some premium growth as we move forward. But again, remember, this has all been in the spirit of a very good economic value-build going forward.

The next question is coming from the line of Mr. John Barnidge of Sandler O'Neill.

Do you have any ability to quantify by how much Hurricane Florence impacted sales volumes in the U.S. in 3Q '18? And then, maybe, by how much do you anticipate Hurricane Michael to impact sales in 4Q '18?

I think -- this is Rich Williams. We do not quantify any impact to our sales results for the third quarter due to the hurricane. Obviously, some of our markets did feel the impact, but overall, we feel good about the results. And also, as we look to the activity for the fourth quarter, we expect to be within our 3% to 5% guidance and do not attribute the impacts of the hurricane to change that.

Okay. And then you have a new digital medical insurance product. It's essentially a new channel of delivery. Can you talk about how it's being received, expected expense ratios for this relative to a normal medical insurance product? And should we see this as a driver of lower expense ratios company-wide? Because presumably, it's got less expenses.

In terms of direct to consumer, I would say, first of all, you should not expect that particular initiative to have a material impact on our overall ratios for obvious reasons. But typically, in the early days, particularly the early days of building out a direct-to-consumer channel, you're going to naturally have a disproportionate amount of investment and expenses relative to premium generation for a period of time. So clearly, there is going to be a level of investment. But the products are priced in the margin as such that it would normalize into very typical benefit ratio, expense ratio dynamics going forward. But from a just pure business-build perspective, you're obviously, going to have some investment before you get the premium -- premium levels to where they can absorb the cost structure.

John, one other thing I might say about hurricanes, which is not in your question, but it is important to note when looking at our ratios quarter-over-quarter. Hurricane season last year, of course, was particularly more severe because it wasn't just a big hurricane season, but it hit highly populated areas. And one thing you'll note, looking at quarter-over-quarter benefit ratio and expense ratio, is last year's third quarter had a higher benefit ratio and had a lower level of DAC amortization. And the reason for that is when hurricanes hit, what often happens is the states will come in and either by county, in some cases, or the entire state. And the last year's hurricane season was effectively the entire state of Florida and parts of Texas. They will actually require insurance companies to basically freeze the policies in place and not allow them to lapse. When you do that, you actually -- you have obviously, a spike-up in persistency during that period, which causes a natural climb in your benefit reserves to reserve for claims activities. And then you'll have much lower DAC amortization because the policies aren't lapsing. So one thing I would note is we think this quarter is, while a very good benefit ratio in the quarter, it's a somewhat normal quarter, but when you look quarter-over-quarter, you'll see the benefit ratio dropped quite a bit, but also DAC amortization picked up significantly. And that was really related to last year's hurricane season. So just a point of information.

The next question is coming from the line of Suneet Kamath of Citi.

I'm going to ask a question about Japan. It's something we touched on at the FAB Meeting. So when we think about cancer insurance sales over the past several years, and we noted that policy count was fairly stable. It didn't really grow that much. So when you roll out these new cancer products, what percentage represents sort of upsales to existing policyholders versus actually attracting new policyholders?

I'm -- what I'm going to do is ask Todd Daniels, who's on the line with us in Japan to add some color to this. So Todd, why don't you talk a little bit about your perspective on the question that came up in FAB but then -- related to policies and lapse and reissue?

Yes, thanks, Fred. First off, on the lapse and reissue topic. Over the course of the year, after we introduce a new product, you can expect anywhere from 20% to 25% of our sales to be a result of lapse and reissue activity. And typically, what we're doing with these new products is we're generating incremental economic value to the company. We're benefiting the policyholder because they have up-to-date coverage. And we feel like that's a win-win for the customer and the company at the same time. And regarding cancer sales and the growth rates and specifically the table that was at FAB, if you went all the way back to 2002 to today, I think that the numbers approximately show that we grew about 1.5 million policies. That gets us in the neighborhood of 15 million policies in-force today. The total market also grew. While the same phenomena, we spoke to medical earlier, where you have companies coming in, offering things like term coverage or, in some cases, one-year free coverage of cancer, their policies are actually getting counted in the total policy in-force count. So we believe that market share number is a bit skewed when you look at the publicly available information. If you were to -- if you had the information on a premium basis, we feel like we would be in a much higher position, but unfortunately, that information is not publicly available.

Got it. And then when you talk about Japan Post and further penetrating that distribution channel, is your sense that you're really accessing new policyholders? Or again, is this sort of reaching policyholders that you have with these upgraded products just through a different channel?

Koji, why don't you take that?

(foreign language) Well, in terms of JP, their customers are brand-new customers to us. (foreign language) And of course, as JP is selling our products to their existing policyholders, but then at the same time, they're also using our product as a door knocker to selling other insurance.

Okay, can I just follow up with one just quick one on this topic? So normally, when we talk about persistency in Japan, we talk about it with respect to premium persistency. But if we were to look at persistency on a product or a policy basis, what would those numbers look like? How'd they compare to what you're seeing on a premium persistency basis?

Todd?

Yes, I think that you would see similar trends. I don't think the numbers would be that different. A couple of things you have to keep in mind when you shift to the policy-based persistency is things like the first sector business that have different premium dynamics. So you may see slightly different persistency metrics across the paid-up-type products on a policy basis than you would for premium. But in general, the trends behave very similarly, especially on the third sector business, which has generally whole life premiums. You're not going to see that much difference. The average premium per policy for cancer, medical and most third sector is in a similar range. So you don't have an influence of any odd-sized premiums impacting your premium persistency number.

One thing I would note on this topic is in the quarter, you notice that we had elevated or a tick-up, if you will, in the expense ratio in Japan, and not all of that but a good portion of that driven by DAC amortization being up. The thing to remind you all of is that it makes a difference as to what cohort of policies are lapsing and reissuing. And for example, because our new cancer product included a very popular waiver of premium feature, and that, of course, being priced for and properly priced to the margins and so forth in a higher premium per policy-type dynamic. That type of a feature can attract lapse and replacement from newer, more freshly issued policies or policies issued in last 5 years. When you -- when you have lapse and reissue on those types of policies, that cohort, you'll obviously release more DAC or amortize more DAC than you do release reserves. So you'll have a more pronounced impact or a higher level of DAC amortization than you do benefit on the benefit ratio side. So keep that in mind. That has a lot to do with what cohorts are lapsing and reissuing as well.

The next question is coming from the line of Andrew Kligerman of Crédit Suisse.

You got a nice bump-up from an investment income from what I understand are floating-rate securities. Could you give a little color on the size of that portfolio both in Japan and the U.S.? And maybe what types of securities are floating-rate?

I'll ask Eric to jump in on that.

Thank you. Well, I'll start with Japan. The size of that portfolio today is about little over $6 billion. And it's comprised of bank loans, middle-market loans and transitional real estate. I don't have the exact breakdown between them, but the bulk is probably in transitional real estate followed by middle-market loans and bank loans. They're offloading off of LIBOR, transitional real estate is typically investment-grade. Middle-market loans are typically below investment-grade, B, same with bank loans. But all are highly negotiated to underwriting standards that we dictate with those providers. U.S. is a few hundred million. I don't have those numbers right at my fingertips. So it's a smaller percentage of their portfolio. And also keep in mind, for Aflac Japan, we like them particularly as part of our dollar program because they are easy to hedge. The 3 months hedging that we do typically, can go out to a year, matches the duration of those floating-rate instruments and has a high correlation to LIBOR, which drives the interest income on those instruments as well as the hedge costs. So for Aflac Japan, for our dollar program, we particularly like them, but of course, it has to start with how do we like the assets, the credit underwriting and things of those sorts.

Got it. And then shifting over to the U.S. sales with the outlook this year, 3% to 5%. And it looks like the career agents have been kind of flattish over the last few years, and the independent brokers are really giving you the upticks. As you look at this newer distribution, do you think it could get you into the high-single digits or double digits over the longer term? Or is this 3% to 5% kind of a good run rate to look out over the next few years?

It's, first of all, 3% to 5% for the fourth quarter is the expectation. Obviously, that puts us at the lower end of the range. We've definitely been -- with the fourth quarter being the opportunity for larger-case enrollments, we've been selective at manage profitability. And so that's why we're comfortable with that 3% to 5% range. As Dan alluded to in his comments, broker sales occur both with our broker distribution team and with select members of our agency field force. And so as the percentage of our total sales gets towards broker sales, I think you'll see us, over the next couple of years, have the opportunity to adjust our range. I wouldn't categorize it as high-single digits because I think it depends on where the market is.

I'll make another comment to that as well. This is Teresa. If you really think about how the market is changing and how we're seeing more brokers in the voluntary space, the need that they had and they've asked us to assist with is people who are educated on the voluntary product space. And so some of what you've seen in the career agency force is a shift of some of those agents to assist with helping brokerage firms to get into the voluntary space. We have specific programs and training that help them to do that. So there is a little noise in that career agent number in that any production written through that brokerage is really going to count as broker production. So -- although the career agent is paid a commission on some of the production as well. So I just -- I mentioned that because it's helpful to understand really the change in the market.

Yes, didn't know that. That makes a lot of sense. But is there a shot at getting double-digit sales longer term? Or is it just very competitive in this market?

Our career agents' sale is about $1 billion. And so it's a -- so to give that a shot, well, we've got to jump over a big number, which is that $1 billion number. On the broker side, that's where the market's growing. And so that's what we're seeing Aflac grow consistent with the market. And we are also, as Rich alluded to, we're also looking at other activities to help us to drive growth both in the career agent and the digital distribution line. So you'll hear a little bit more about that at our outlook call.

Let me give you a different perspective a little bit on this. Career sales for the 9 months are down about 1%, and they count roughly 2/3 of our sales. Broker sales and group, both are up. Broker's up 8.6%, and group is up 26%. So as that mix changes and it goes from 1/3, 2/3 to 50-50, if we're continuing to see broker sales grow at double-digit, which we think we can do, that number with time is going to shift towards the larger numbers if we do what we hope we can do. So that's not to say we're climbing that, we'll be talking about that at the outlook call. But just think of it in terms of when you look at us -- and we saw the similar thing with the corporate agencies in Japan, how they were the foundation of the company when we started, but the growth came from the new distribution channels. And so they're still very important, just like our individual agents that were selling for us. But it's a changing marketplace. And I think it's going to evolve with time, and that growth pattern that you're seeing and the job they're doing with our brokers is very positive when you break it out.

The next question is coming from the line of John Nadel of UBS.

Fred, a pretty open-ended question. I was just wondering if you could provide any early -- relatively early thoughts on potential impacts from the new FASB accounting, the long-duration focused accounting?

Yes, I would say that our comments remain fairly consistent with what June Howard had covered at FAB. And that is we're -- we're like everybody in the industry, we're working on the project. I think what I would say is that there are certain aspects of the new FASB adoption that I do think could be beneficial to just investors, if you will. And in terms of transparency, I think DAC amortization, for example, is going to be easier to follow and provide greater stability. I think some of the unlocking, if you will, of reserves can provide good, meaningful information to investors. And so I don't feel as if the new FASB guidelines are necessarily a negative.

What is probably a -- mostly the challenge is volatility. And that is you're going to likely have some additional volatility in results. That's going to have to be explained and understood as to what is economic volatility and what is not economic volatility to the company. So I don't have a heck of a lot more color to provide you on that because we're still working through the project, making sure we're interpreting and applying things properly. My editorial comment would be, you've got one fundamental application of the new FASB rules, but they're going to react differently depending on your business model. And Aflac has a very unique business model, unique in the U.S., with our supplemental and health-oriented products, and unique in Japan in that we are heavily weighted towards third sector. These are all really good economic issues for the company because they provide stability and good clarity of margin and reliability of margin at a relatively low capital required to support the business. So these are all the types of economic business lines that the rest of the industry is looking to grow over time. And so -- but the unique nature of our position in those markets means that when you adopt the guidelines, you may have a different type of reaction on certain line items, whether it be an OCI or reserve dynamics because it's just the mix of our business. So what's going to be important for us is not only explaining the impact of the adoption but, as important to me, why that adoption means something different because of the makeup and mix of our business. So we're working on it, John.

That's helpful, Fred. The economics of the economics, right? It's worth talking about accounting.

That's right.

The second I have is one for Eric. And I don't think the credit markets are necessarily behaving in such a way to indicate that we're at the precipice of a turn, but the equity market certainly seems to be acting in that way. Is there anything that you're doing? And I'm really talking about here in the last 30 days or so, as market conditions have really started to shift, is there anything as you're reviewing the portfolio that's sort of rising to a "watch list" sort of category, I put quotes around watch list, that you think maybe it's time to start altering allocations?

Yes. But not in a dramatic fashion because we too don't see an event vis-à-vis credit market, even despite equity markets that would terribly impact from a credit standpoint and the quality standpoint. We don't see anything probably for another 1.5 year. Having said that, what we are thinking about and acting on, as an example, is when we look at our traditional credit book, like investment-grade bonds, the high-yield bonds, the old yen private placements, which are much less reduced from years ago, but nevertheless, we still own them, we are saying to ourselves, "Well, to the extent we see something in 2020, and we're not calling for a major sell-off but a change of the credit cycle, are there credits that we'd rather not own? Because in that market environment they'll start to deteriorate around the edges." So we have been doing some de-risking, particularly some old yen private placements throughout the year. And year-to-date, we've de-risked about $1 billion of positions. Again, nothing that raises an alarm for us. It's just good hygiene. We want to be better positioned when the credit cycle turns.

The other part of the portfolio I would call out is in our middle-market loan portfolio. We have definitely seen a lot of money chasing those types of deals. It's well written about in the press. Our production this year has been a little slower than we expected because we are upholding our credit underwriting standards. There are a lot of players in the market that are loosening those because they've committed to their investors, they're going to buy those assets. Well, we've not taken that view. And we're in the middle now of planning for next year as we go through, what we call our silver and gold plan here internally. And we have to think about on the investment-side deployment, and particularly to private markets like the floaters, where we've made a commitment. So in the middle-market loan space, just from a standpoint of planning, we're planning to be more conservative with respect to how much we might actually be able to underwrite and deploy. Having said that, the other part of our floating-rate portfolio, transitional real estate, we still find very good relative value. Spreads have come in somewhat, but that market has not been as hot, if you will, as the middle-market loan space. So we see that as a place to potentially offset some of the lower deployment in middle-market loans. But to your question, John, around the credit environment, those are the 2 areas I would focus in on, where we're making adjustments that will have minor impacts to the net investment income. They're not major. But when that credit cycle turns, the actions we take today will serve us well at that point in time.

The next question is coming from the line of Alex Scott of Goldman Sachs.

First question I had is sort of along the same lines. I guess, just high level on spread compression. We sort of haven't seen any part of its -- the tactical moves that should have been made the last, call it, a year or 2. Could you give us idea of sort of where you are with the commitment to invest with NXT Capital? And sort of how close we are to winding down some of the investments that have improved new money yields? And ultimately, are we going to kind of drawback into a little bit of yield compression? Or should it be more stable and flat from here?

Well, that's a little difficult to answer because the new money yields will be not only impacted by spreads but the overall level of just interest rates. And remember, in our book, we've got Japan interest rate and dollar interest rate. And from that perspective, both are generally heading up now. So even to the extent we're buying yen assets, and we are, we do need good, old-fashioned yen assets whether they be JGBs, private placements, yen publics, for ALM purposes for Japan. Those new money yields have been trending higher around the edges. So that's good.

In the U.S., relative to our floating-rate portfolios, we still have capacity in both asset classes of transitional real estate middle-market loans. We have definitely seen some spread compression offset by higher LIBOR yields because those coupons are set based on LIBOR plus the spread. So net-net, we've seen a small decline but not large. When we put that money to work, I would keep in mind 2 things that those portfolios for us: the floating-rate portfolios, the loans, become mature portfolios. There is a natural speed of prepayment on those loans. They typically can be anywhere from 3 to 5 years. But you do start to get natural prepayments. So you will see in our new money maturities not only of old fixed-rate bonds but even prepayments of some of these floaters, which then, therefore, gives us ample capacity to reinvest back in the market depending on the credit underwriting standards.

So a shorter answer to your question is for Aflac Japan I'd expect the new money yield next year, if conditions were the same as now, to be hovering around that high 2%, low 3%, maybe 3.25% type area. I'll just keep in mind, from a tactical perspective, as we go through the year, some of our investment strategies could pivot depending on what happens in any of those markets. But that's what I would say if just market conditions stayed the same and our new money allocation stayed approximately the same, which at this point, we would expect.

And maybe just one follow-up. On the rolling of the FX hedges, is this -- I know that, that was sort of locked in, sort of more onetime. Is that a Q4 item? Or more of a 1Q item, where you have to roll more of those hedges?

Yes, in terms of the rolls, going back to what we told everybody for 2018 at the end of last year. We chose to lock in the majority of our group 1 hedges. Our group 2 hedges are locked in. If you remember the charts I showed back at the analyst meeting, think of those. And a lot of the termed-out hedges are actually starting to roll now, between now and the end of the year. But we are in the middle of our planning sessions now. And thinking about -- for 2019, we're thinking about hedge costs, interest rate, the net income of our floating-rate assets. So we right now are discussing what we want that strategy to be. And we will report out at the outlook call more specifics on our decision around that.

I would tell you although to keep in mind the following, and this is the beauty of the strategy, particularly with group 1, the floating-rate. Regardless of whether we lock in or not, if we don't -- if hedge costs continue to go up, and assuming that's driven by LIBOR, which is the biggest determinant of hedge costs, not the only though, well, hedge costs will go up, the income from the floaters will go up. And the nice thing is that net number, the growth income from the loans or the floaters less the hedge costs, if there's a trend up, we get an -- we get a boost on both sides of that statement, if you will, which means our net should be pretty consistent and solid. To the extent we choose to lock it in, there is somewhat of 2 elements of that decision. One part of that decision is we have a view that hedge costs will continue to increase during the year. And secondly, we might have a view that all things may equal. It provides a little bit more stability to net investment income and earnings. So those are some of the factors that go into our decision-making, which as I said, we're looking into now to be able to report out at the earnings (sic) [outlook] call.

The next question is coming from the line of Humphrey Lee of Dowling & Partners.

Just to follow up on Fred's early comments on the cohort of our products being kind of being up-sold or switching to the newer product. I did notice for this quarter a year ago, I guess, redundant reserve releases were lower, but DAC amortization being higher. I'm just wondering kind of based on kind of your expectation of these type of switching, are you largely over from those kind of more recently issued policies switching to the new product? Or should we still expect some of this dynamic to continue in the coming quarters as some of these activities continue?

One thing I would say just at the start of the answer to the question is that keep in mind that the concept of lapse and reissued policies, and these are policies that are being purchased because they represent an improvement for the consumer. And remember, we are re-underwriting when we reissue the policy. This has been going on for as long as anyone can remember in terms of being in the third sector business. So it's very natural that both on the medical side and on the cancer side, you will see a level of lapse and reissue. In fact, we've done some analysis here more recently and when looking really even after -- at the last 3 years worth of quarters, it's very much a consistent level of benefit ratio improvement that comes from lapse and reissue. It tends to range from 90 basis points to maybe as much as 130 basis points, depending on whether a new product has been issued or what have you. So you need to keep in mind that there is not an abnormal necessarily -- it's not abnormal to see a level of lapse and reissue. That's been normal. It's more that when you do launch a new product, you'll see a period of time where it pops up a bit from what you would expect. And I would expect that to calm down for the rest of -- as we go through the rest of the year. It's already starting to calm down. Todd, do you have any color on it from your perspective?

No, I think that's right, Fred. And just remember, it all depends on the cohort of policies that lapse, whether they're more recently issued, or if they're older in-force policies. And that will determine the amount of benefit reserve release versus DAC amortization that you get on the P&L.

Got it. So and then shifting gear. A question for Eric. You've talked about -- there's a little bit more allocation to the floating-rate assets, especially to the real estate transitional assets. I think that's kind of the reason why the new money here in the U.S. kind of jumped to kind of like 5-plus this quarter? Is there any kind of outsized allocations this quarter? Just kind of -- to help that kind of new money yield? Or is that how the new run rates we should be thinking about going forward?

Yes. Thank you. And also, I'll take the opportunity because Andrew had asked, I didn't have it at my fingertips, but I do now. For Aflac U.S., there is about $1 billion in total approximately on the balance sheet currently of those floating-rate assets. And about 2/3 is in transitional real estate. And about 1/3 in middle-market loans. Relative to your question, no, our asset allocation for the quarter was pretty even throughout. There was, at the beginning of the year, a higher amount in transitional real estate because we were successful at finding and deploying assets in that asset class. But now, it was pretty more normal and spread out. But we'd expect that new money yield to be traveling in the 4.5% to 5%, broadly speaking. Glad that you did ask.

The next question is coming from the line of Ryan Krueger of KBW.

I had a follow-up question on, I guess, from Japan on the internal hedging at the holding company. I guess, once you've completed that, how should we think about the balance sheet sensitivity to move in the yen? Will -- if there'll be -- will there be more capital sensitivity on a go-forward basis due to the reduced hedge ratio?

Actually, when thinking in terms of GAAP capital, if you will, in terms of equity, because you are -- have more of a mark-to-market through net income of both the positions, you would expect the offset to calm down some of that type of volatility. That's not the express design of the program. The design of the program is economic in that we are wanting to defend the economic value of our Japan franchise for our shareholders in dollar terms. And so we think of that hedge as being a combination of unhedged U.S. dollars in Japan, borrowing in yen at the holding company and now this strategy of back-to-back hedges at the holding company. These are 3 tools that we're using effectively to protect the economic value of Japan franchise from movements in the yen, particularly a weakening yen over a long period of time. And so that's really the basis of it. It is economically in its design. It's not meant to be some sort of hedge on earnings or a hedge that's specifically looking to calm down GAAP definition of equity tangible or AOCI. So that's the way I would describe it, Ryan, that we've dialed it in economically.

And I would just add that in terms of capital volatility, we do not expect any increased capital volatility from this. We hold significant capital, and we've put in place a number of processes and procedures in order to limit capital volatility. So at the holding company, we would not expect any significant additional capital volatility from this.

Operator, I believe that concludes our call. And I want to thank everyone for joining us this morning for our call, and hope that you'll join us on December 3, Monday, for our 2019 outlook call. Please feel free to contact our Investor and Rating Agency Relations department if you have any questions in the interim. And we look forward to speaking with you soon.

Thank you, speakers. This concludes today's conference call. Thank you all for joining. You may disconnect at this time.