Hi, everyone. I'm Ryan Krueger from KBW, pleased to have with me today Max Brodén, who is the CFO of Aflac. Thanks for joining us again this year, Max.
Thank you.
I thought I'd kick it off, start discussing claims experience, first in the U.S. You've seen quite favorable claims throughout most of the pandemic in the U.S. Can you give us a little color on what you've seen and also how claims have trended in more recent months?
Yeah. Our claims utilization tends to have a negative correlation to infection rates overall. Every time we see spikes in infection rates, we tend to see a drop in claims utilization. This makes a lot of sense when you think about it. People are staying home more. They're not out and about. You get less car accidents. You get less sports accidents. People are not going for elective surgeries, less physical exams, et cetera. All the things that drive claims for us tends to go down as people stay at home more. If you look at this year, it's a great example of that. We had infection levels spike up dramatically over the Christmas, New Year time period. Then we saw, especially in the month of January, a significant drop-off in claims utilization.
That started to gradually increase somewhat in the month of February and March. In the second quarter, it kind of stalled. It stalled at a somewhat low level. This is something that we did experience in 2020, and we see it in 2021 as well. We are at a lower level than what you would generally expect, and that's why when you look at our more normalized underlying benefit ratio, it was running at 49.1% in the second quarter compared to our reported including IBNR releases at 43.5%. That gives you a better feel for what is the sort of core underlying rate.
By the way, I would mention that this correlation, we have an element of that in Japan as well, not to the same extent, less simply because infection levels are not at the same level in Japan as they are in the U.S. They have been rising more recently, but they haven't been as pronounced as they've been in the U.S.
Got it. I guess the follow-up would be, do you see much risk that at some point in time when we get past the pandemic, that there could be some reversal of the favorable experience, and you could actually see higher than normal claims due to either increased health issues from not going to the doctor or other factors like that?
First of all, I think it's important to understand that we are generally a frequency company, i.e., what's driving our benefit ratio is frequency, not necessarily severity. There are components of severity, but in the scheme of things, it's really frequency that drives the benefit ratio. It differs by product. If you start with, for example, Accident Insurance, our largest line of business in the U.S., I doubt that you and I are going to go out and have an extra accident, as we sort of normalize our behavior post-pandemic. Just because we didn't have one last year doesn't mean we're going to have two the next year. For that line of business, I do not expect that we will have any sort of over and above frequency, but rather that we just go back to where we were.
The one line of business to really keep an eye on and think about is Cancer Insurance. We have seen a decline in terms of claims in our cancer line of business. At the same time, we do not believe that cancer incidence rates are down. Just because people are getting COVID and we are in a pandemic does not mean that we do not develop cancer at the same rate as we did before. It's simply that we believe that screenings are not taking place, we are not detecting it. When we do eventually detect it's more likely to be at a later stage. Instead of cancers being detected at stage 0 or 1, we start detecting at stage 3 or 4. We could have a little bit of an element in terms of severity there.
The way this will manifest itself is likely to be that we could have a period of time where we would have above-trend claims coming through. That's something that we monitor closely, we are well aware of this. We quite frankly expect this, we try to run scenarios of this as well. That is coming through in the guidance that we give and outlook that we give. We try to take this into consideration.
Can you give us just any rough sense of, in the U.S., how big the Cancer block is relative to the overall block?
You can look at, think about policies in force being a little bit lower than what the new sales contribution from Cancer is.
Got it. Okay, thanks. Maybe sticking with claims experience in Japan, though. I guess maybe less focused on the pandemic, can you discuss some of the short-term and long-term benefit ratio trends that you're experiencing in Japan at this point and some of the key drivers?
On a short-term basis, it is somewhat pandemic-related. We see some of the same dynamics in Japan as we do in the U.S., not to the same extent. As I said earlier, I think it's because infection rates have been lower in Japan than in the U.S., their whole hospital system have been operating on a more normal basis and have seen less disruptions compared to what we see in the U.S. Therefore, it's been more normal there. The long-term trends are, we talked about this for many years, and we sound like a broken record, it's still what's going on is that hospitalization trends and stay in hospital for certain procedures continue to be on the decline.
The reason for this is simply that there's a significant shortage of beds in Japan, the need for healthcare services are increasing as the population is increasing in Japan. These trends we don't really see changing. We continue to see, obviously, the population aging. We don't see an increase in supply of hospital beds, therefore, this will continue to put pressure on hospitals to shorten the stays. There's also a greater use of outpatient services as well. You take all of this together and it leads to favorable outcomes in terms of actual to expected on our policies, especially our Cancer block of business.
If I think back several years ago, I think there was some expectation by Aflac that this was going to slow down at some point in time. Is it fair to say you haven't really seen much of a slowdown in this trend and the shorter hospital stays is kind of continuing at the same pace?
The trend is obviously it's not going to be as dramatic going forward as it's been in the past, That's just simple math. You can't have that continuous. The curve is going to start to flatten out eventually. We've seen some of that flattening out to happen, but the trend is still on the decline. One thing I would like to point out, though, that when we price new business, we do not price for this continuous improvement to continue, as we price for current experience when pricing new policies.
Got it. Moving to persistency. Probably when we spoke at our conference a year ago, there was definitely the thought that the U.S. persistency was probably going to deteriorate due to pandemic-related conditions, but it seems like you just haven't really seen that. Can you talk about what has occurred? Also, I know you have some other actions you've been specifically taking to try to improve persistency in the U.S.
You didn't ask about Japan, but I do want to mention that in Japan, we see very stable persistency. It has remained in the 94%-95% range. We haven't really seen any impact there from the pandemic conditions. You're right that in the U.S., as we entered the pandemic, there were clearly a lot of uncertainties how this was going to play out, economically, behavior, ultimately how claims would come in, et cetera. As we thought about it and when we were running internal stress tests, we were definitely thinking that persistency could become an issue because of economic pressures, increased unemployment rates, et cetera. That has not played out. Clearly, that's very fortunate. Unemployment rate is at a very low level, and we have actually experienced an increase in persistency rate.
In the U.S., we're currently reporting at the end of the second quarter persistency over 80%. Now, I do want to caution because there are two components that have sort of pushed it up from pre-pandemic levels. We were around the 78%, and we've now moved up to slightly above 80%. There are two components that I would argue are one time in nature and not necessarily something that I would expect long term. The first is emergency orders by states. States have put in place a no lapse emergency order, which means that if a policyholder is not paying the premium, we are still not allowed to lapse the policy. You still have a number of states with emergency orders in place. We estimate that that increased our June, sorry, the second quarter persistency rate by about 60 basis points.
Lapse rates on first-year policies is about twice the level of lapse rates of policies that survive the first 13 months. That means that in years or time periods when you have significant sales increases, you tend to actually see your persistency go down. In times with lower sales that we've been in right now, persistency naturally goes up simply because of this mechanic. We estimate that in the second quarter that improved the persistency rate by about 80 basis points. There's still some natural improvements in play going on, and there were some deliberate initiatives that we took last year, and that was in order to drive agent persistency, account persistency, and ultimately policyholder persistency.
We introduced a number of new products, value-added services to make it easier and better for agents. We also added agency loans to make sure that they could stay in business even as production levels were quite low. We also ran wellness campaigns and notified policyholders of the wellness component of their policy so they can get some extra cash as well. We believe that all of this altogether have helped to push up our persistency. That component of the improved persistency rate, I do view as real and hopefully that should be permanent.
Got it. If you were to normalize for the two factors that are less permanent, do you think you can do more to improve persistency over time, or are we kind of at a pretty good level at this point?
No, I do think we can do more. A lot of it is coming down to driving all these components. The better our agents are doing, the longer they stay in business, that means that they're going to serve their accounts better. The better they serve their accounts, the more likely persistency is to go up within those accounts. When accounts get served well, the policies that are associated with that worksite account stay in place for longer as well. We're trying to improve persistency in all these three levels, and they're also connected. If we do a good job of this, and these are efforts that we have a lot of initiatives in flight in order to drive this long term, because at the end of the day, acquiring a new policy costs a lot of money. There's significant customer acquisition expenses for us.
If we can rather spend a little bit of money in order to keep existing customers half a year, one year, two years longer and collect those premiums, that is economically very attractive to us and long-term should improve our premium rates.
Got it. Thanks. Let's move to expenses. You've had some pressure on expense ratios, I think from two things, top line pressure, but also investments into new business initiatives. You've also taken some offsetting actions to reduce expenses. Can you talk about where you see expense ratios headed in both the U.S. and in Japan over the intermediate term when you take these things into account?
Yeah. Unfortunately, a lot of things have moved against us. Lower revenues, and then we also have had to invest significantly in order to stand up a couple of new businesses in the U.S., like Group Life and Disability, Network Dental and Vision, and Direct to Consumer. All of these are not fully contributing in terms of revenue generation, but we got the expenses for it. We get hit at both the numerator and the denominator, and that has obviously led to expense ratios, especially in the U.S., that are too high. Therefore, we are attacking that both by cutting expenses and then obviously, we strongly believe that as the new businesses come on stream and start generating revenues, we are going to get an uplift there, and that certainly should help.
In the U.S., we hope that this should start to play out already next year with a slightly lower expense ratio than what we are running at for the full year this year. In Japan, it's more of a slow grind, where I would expect our expense ratios on a GAAP basis to be fairly stable. Part of this is that on a GAAP basis, we do continue to have some headwinds in terms of revenue generation because of the paid-up impact on primarily our first sector savings block. We were through the time period of the five-year paid-up impact from the WAYS block, but we do have a block of business of 10-year paid-up product, and we're going through the paid-up impact of that with associated revenue headwinds here in 2022, 2023, and 2024.
That's not going to have as significant impact as the five-year paid-up impacts were, but it's still going to be a headwind for us. The starting point is that we need to cut in order to stand still and be at the same ratio, so to speak. It's going to be a little bit challenging, but I do believe that our expense ratios should be roughly flat.
Thanks. That's helpful. I'm going to shift gears a little bit onto sales. In Japan, there's quite a number of different dynamics going on between the Japan Post, new products, pandemic. Can you talk through some of the dynamics that are impacting Japan sales at this point, and also help us think about the potential outlook for the second half of the year?
Yeah. Thinking about the second half, obviously, if you start with Japan Post, they began actively selling again in the second quarter. As I'm sure you understand, this is a significant organization. They went through a lot of change, and a lot of that change is moving to a new sales model, a new compliance model, et cetera. To get training up and running on the sales and distribution people, and also get the confidence among them plus the confidence among consumers, it's going to take a while. I think that the recovery there is really going to be slow and gradual. At the same time, we're very confident that over time, Japan Post will continue to be a very significant distributor of ours, and their standing in Japanese society will continue to remain strong.
If you turn to the other distribution channels, to some extent, we are dependent on the virus environment. You have seen case counts in Japan rise, that has led to a number of emergency orders, a state of emergency being declared in a number of prefectures. You may have seen as even yesterday, Japan extended for a number of important prefectures for us, i.e., for example, both Tokyo and Osaka extended the state of emergency until the end of the month, i.e., September 30th. All of this is making it a little bit more difficult on the margin for us to interact with consumers because the fact of the matter is that the more confident the consumer is to interact with us on a face-to-face basis, the more meetings and sales we're going to generate.
We're also more efficient in terms of close rates when we do it face-to-face rather than virtual. Even though we do have the virtual capability in place, it is not as efficient as face-to-face selling. You take all that together, and I think it still stands that we do expect a recovery in the second half, but it's going to be slow and gradual.
Thanks. You also, on the last quarter call, I think for the first time, discussed plans to enter the elderly supplemental care market in Japan, which would be a new market for you. Can you give a little more detail on the product characteristics that exist in this market and also how you see the potential growth opportunity over time?
Yeah. Clearly the aging population in Japan, it leads to significant market headwinds for certain products. For something like nursing care, we see significant underlying demand. In fact, the percentage of Japanese individual aged 75 and over is about 15% right now, but in just 10 years' time, it's expected to rise to close to 20%. This is a market that will experience most likely significant underlying demand growth. That's where our product is coming in, where we do think that we can help filling that gap. Now, I do want to make it very clear that nursing care in Japan is different from how products were priced in terms of the long-term care market here in the U.S., especially back in the pre-2000 sales of long-term care policies.
First of all, you have a much more extensive government long-term care program, and our product is sitting alongside that program. We carry the same definitions that triggers care generally as the government program, which makes it sort of supplemental in nature. We also made sure that as we price this product, we make sure that we price it for very low lapse assumptions, because this is a product that inherently have been lapse supported in not just the U.S., but in many jurisdictions around the world. We're clearly in a low-interest rate environment, and we should probably, when we price a product like this that tends to be long-term in nature, you need to obviously price for the low interest rates to stay as well.
We feel quite good about the pricing that we can achieve on this product, also long-term, the clear underlying demand for it. The third angle of it obviously will be that this needs to be risk managed, and we risk manage it by capping payouts, make sure there's no unlimited payouts. We sit alongside the government and make sure that the interests between government policyholders and the insurance company are aligned.
Got it. In the U.S., can you talk a little bit about how the sales recovery has differed so far between your traditional career agent channel and in the group or broker channel? Also, how sales are tracking so far this year relative to your expectations?
The broker channel clearly have done much better than the career agency channel, and you would expect that to be the case. The career channel, this is what you should think about as going after small businesses. Generally, the target is the three to 99 marketplace. These sales are being conducted at the worksite and face-to-face. Well, we were chatting just before this fireside chat that both Ryan and me were actually at our worksites today for the first time in a long time. That's just one example. People are generally not at the worksite, and people's confidence to transact face-to-face is not really there. This has been the case throughout the pandemic, and therefore, obviously, something like the career channel will be heavily impacted by that. The other thing I wanted to mention as well when it comes to the career channel is recruiting.
That's also been challenged. We have a model where we generally do not recruit licensed agents. We tend to go after non-insurance salespeople, and we bring them in and we train them and we license them and bring them into our company. That's generally how people come into the Aflac sales and distribution network. We have a very strong culture around this. When the pandemic hit, a lot of states did the licensing, not online, but face-to-face. They actually shut it down. We have had, in a number of places, we could not even license new recruits. It made it really difficult for us to recruit and license and grow.
That is important in our model, which is a fairly high recruitment of agents that occur all year, the survival rate is a little bit lower than what you will find than if you were to, for example, recruit already licensed agents. That means that our career model stands out a little bit compared to many other models that you see in the insurance space. That also means that we are heavily impacted by the pandemic and probably more impacted than many other distribution channels because of that. When you move to the broker channel, it's a different story. Our broker channel have continued to chug along pretty good. Here we're targeting large corporates that to a greater extent already pre-pandemic, were doing enrollment virtually. As an example, I work for a large corporate. I enroll virtually. Done that ever since I arrived at Aflac.
That's how it works. You don't necessarily sit down with an enroller face-to-face. The processes and procedures in the broker channel are much easier to sort of overcome. That's why you have gradually continued to see growth rates for the broker channel and most of the pressure, we're really seeing it in the career agency channel.
Can you give an update on the progress you've made so far in some of your newer growth initiatives in the U.S. in regards to group dental, vision, and Direct to Consumer?
Yeah. Let's take them in opposite order. Direct to Consumer, this is where we were standing up a brand-new digital platform where consumers, regardless if their employer are offering Aflac or not, can gain access to Aflac policies. This was launched in the beginning of January of this year, this is a full end-to-end digital platform that is completely new. In the beginning, as we continued to sort of test it out, we restricted the flow to the Direct to Consumer website, that has gradually been lifted. We're now seeing more and more flow coming through, the sales are gradually increasing as well as a function of that. At this point in time, in the scheme of Aflac, it's still relatively small, but it's certainly building, we expect this to be meaningful over time.
This is a channel we believe that more and more consumers would like to use. One thing I want to mention, though, is that obviously from a P&L standpoint, growth in this channel is negative. First of all, we have incurred expenses to build up the platform. Now as we start selling, the acquisition expenses are not deferred, i.e., we cannot DAC it the same way we can do for a traditional agency-sold policy where commissions are DAC'd because we cannot track any specific acquisition expense to a specific policy, therefore, we can't DAC it. This business is likely to generate losses for a couple of years, even as it gains scale.
If you turn to Network Dental and Vision, this is coming out of the Argus acquisition which was really a TPA that we acquired that generally served major medical companies and did the administration for dental and vision programs. In there, we also saw a great opportunity to use some of the capabilities that they had to grow out a national Network Dental and Vision product. We are doing that right now. We are now being licensed in more and more states, the network is building. We are slowly starting to be in the market selling that product as well. I expect that to be relatively low this year, but starting next year is really when we should really see the dental product contribute more to our sales.
It will take probably about two to three years for it to gain enough scale where it's throwing off profitability as well. If you lastly take the Group Life and Disability business that we acquired through a reinsurance transaction from Zurich, we're still in the building phase of that. There is an existing business, it generates some sales, it generates some premiums, but we're still sort of growing it. We're still building connections with brokers and companies, we're sort of still the new kid on the block. I would describe that we bought a building with a lot of the infrastructure there, it wasn't fully decorated on the inside, that's really what we're doing now.
We're decorating it in order to make sure that it fits with Aflac, it fits with the rest of our product portfolio, we do want it to do well on itself, it's very important that it strengthen the sales of the other Aflac core product set as well. That's really how we get this sort of one plus one to become two, hopefully more than that over time.
Thanks. I'm going to shift gears a little bit to the balance sheet. You stepped up share repurchase in recent quarters, but you still have a pretty strong excess capital position at the company. How are you thinking about that balance at this point between redeploying capital above and beyond your free cash flow versus retaining a cushion going forward?
Yeah. Part of the reason why you have this very strong cash and capital position at the holding company is that in the very beginning of the pandemic, we raised $1.5 billion of capital through debt securities, and we have since then kept that at the holding company. That was out of abundance of caution, because when you go into what could have been an economic, could have been a financial crisis you need more capital is very important. This capital has essentially been sitting there, and fortunately, we haven't had the need to use it. Over time, I would expect that we were not going to sit with this kind of cushion at the holding company. It's a significant capital cushion. It's also a drag on our return on equity.
It's important that this gets invested wisely, either through projects that can grow our business or it will be returned to shareholders through dividends and share repurchase. You said that correctly, that we have stepped up our share repurchase a little bit this year. Your question, would we philosophically be willing to repurchase or deploy capital more than our sort of free cash flow generation to the holding company? I would say that philosophically, there's no reason why we couldn't do that. Our capital position is and should always be a function of the risks and opportunities that we face.
We want to make sure that we have strong capital ratios in our operating subsidiaries, which I do think that we have today, both today and given the growth opportunities that we're seeing in the U.S., we have a high RBC ratio that can help us fund any new business strain coming from the different new initiatives that we talked about earlier. Our capital at the holding company and the cash flow that we do receive is a very strong foundation for continue to increase the dividend and being a strong deployer of capital back into our own stock or through other projects that can grow the franchise long term. Philosophically, does that mean that we temporarily could have deployment that is greater than the free cash flow? Absolutely. The other way around as well.
There will be times when we run lower than our free cash flow and times when we run higher than our free cash flow. Basically, we should make sure that we, over time, optimize the level of capital to the risk and the opportunity set that we see.
I guess I'll move to a topic we'll all be discussing more unfortunately, which is LDTI. I guess it seems like Aflac has directionally discussed the GAAP LDTI impacts more than most of your peers so far. I'm curious when you might consider disclosing the quantitative impacts to the market.
Yeah. We still haven't decided when to exactly come out with that. I do anticipate to advance the ball somewhat at our FAB later on this year. In terms of the exact impacts and sensitivities, we still haven't fully decided about that. The one thing I do want to make clear, though, is that the LDTI implementation, this is a very significant overhaul of our accounting framework, and it's probably the biggest one in a long, long time for U.S. life insurance companies. This has no impact on the way we manage and run the business and has no impact on the way we manage capital as well. We deem any sort of movements to generally be non-economic in nature.
Now, we've also been quite clear saying that the day one implementation could have a negative impact on our AOCI, to make a complicated issue quite clear, I would describe it as the following. The main reason for that is our long duration liabilities in Japan. That means that we sold the business a long time ago. At that time, we locked in the discount rates at the interest rate levels at that point in time. Clearly, interest rates levels are now much, much lower. When you then move to a current discount rate for discounting those liabilities, clearly, that's going to be significantly lower, which means it leads to higher reserves, and that's going to flow through AOCI, and decrease shareholders' equity.
At the same time, the offsetting impact is morbidity that's been very favorable for us over time, but we can't fully have that as an offsetting factor in the LDTI implementation on day one.
We're almost at time. I just wanted to ask one more on this, which is, given your comment about obviously it doesn't change the economics or how you run the business, are there alternative metrics that you're considering providing, like an embedded value metric or more on free cash flow or those types of things to augment the GAAP disclosures?
It's something that is up for consideration. We have developed those metrics. We use those, especially EV metrics, we use that in how we look at our new business, both in Japan and the U.S., and we evaluate distribution channels based on this product lines, based on this, and blocks of business based on this. It's something that we potentially would consider. I would argue that this is probably something that we as an industry are going to get together and really discuss because I think it's important that we find consistency among the different players. That's going to make it easier in the conversations that we have with you, and I think it's going to make your job a lot easier as well. It's still to be determined, but it's definitely something that we are using internally more and more.
Great. I think we are out of time, we'll wrap it up there. Thanks a lot, Max, and to Aflac for participating and to the audience for listening, we will end it there.
Thank you very much. Appreciate it