Good day, welcome to the Globe Life Inc.'s third quarter 2020 earnings release conference call. Today's conference is being recorded. At this time, I would like to turn the conference over to Mr. Mike Majors, Executive Vice President, Administration and Investor Relations. Please go ahead, sir.
Thank you. Good morning, everyone. Joining the call today are Gary Coleman and Larry Hutchison, our Co-Chief Executive Officers, Frank Svoboda, our Chief Financial Officer, and Brian Mitchell, our General Counsel. Some of our comments or answers to your questions may contain forward-looking statements that are provided for general guidance purposes only. Accordingly, please refer to the third quarter earnings release we issued yesterday, along with our 2019 10-K and any subsequent forms 10-Q on file with the SEC. Some of our comments may also contain non-GAAP measures. Please see our earnings release and website for discussion of these terms and reconciliations to GAAP measures. I will now turn the call over to Gary Coleman.
Thank you, Mike, and good morning, everyone. First, I'll point out that the company continues to effectively conduct business and our operations are running smooth. In the third quarter, net income was $189 million, or $1.76 per share, compared to $202 million or $1.82 per share a year ago. Net operating income for the quarter was $188 million or $1.75 per share, a per share increase of 1% from a year ago. On a GAAP reported basis, return on equity was 9.4% and book value per share was $77.60. Excluding unrealized gains and losses on fixed securities, return on equity was 13.6% and book value per share grew 10% to $52.39. In our life insurance operations, premium revenue increased 7% to $674 million, while life underwriting margin was $171 million, down 6% from a year ago.
With respect to premium revenue, we have been pleased to see persistency and premium collections improve since the onset of the crisis. However, the decline in margin is due primarily to approximately $18 million in incurred claims related to COVID-19. For the year, we expect life premium revenue to grow approximately 6%, while life underwriting margin is expected to decline 2%-3%, primarily due to the impact of COVID-19 claims. At the midpoint of our guidance, we anticipate approximately $56 million in COVID-19 claims for the full year. In health insurance, premium revenue grew 7% to $288 million, and health underwriting margin was up 20% to $73 million. The increase in underwriting margin primarily due to lower acquisition costs. For the year, we expect health premium revenue to grow approximately 6% and health underwriting margin to grow 11%-12%.
Administrative expenses were $63 million for the quarter, up 4% from a year ago. As a percentage of premium, administrative expenses were 6.6% compared to 6.7% a year ago. For the full year, we expect administrative expenses to grow around 5%. I will now turn the call over to Larry for his comments on the third quarter marketing operations.
Thank you, Gary. We are pleased with the third quarter sales. Direct consumer sales grew across all channels, and the agencies have adapted to virtual sales appointments and recruiting. They are thriving in this environment. Additionally, agent licensing centers have opened, and we're conducting some in-person sales in certain situations. I will now discuss current trends at each distribution channel. At American Income, life premiums were up 9% to $319 million. While life underwriting margin was flat at $100 million. Net life sales were $68 million, up 14%. The increase in net life sales is primarily due to increased agent count. The average producing agent count for the third quarter was 9,288, up 23% from the year-ago quarter and up 11% from the second quarter. The producing agent count at the end of the third quarter was 9,583.
We continue to see a significant pool of candidates, in part due to current unemployment levels. At Liberty National, life premiums were up 3% to $74 million, while underwriting margin was down 21% to $15 million. The lower underwriting margin is primarily due to higher claims related to COVID-19. Net life sales increased 2% to $14 million, while net health sales were $6 million, down 2% from the year ago quarter. The average producing agent count for the third quarter was 2,551, up 6% from the year ago quarter and up 7% from the second quarter. The producing agent count at Liberty National ended the quarter at 2,574. We have seen continued adoption of virtual recruiting and selling practices.
Also, the relaxation of certain local restrictions has allowed agents to be able to return to some in-person presentations in addition to virtual methods. This environment has also provided abundant recruiting opportunities, supporting continued agent growth for the future. At Family Heritage, health premiums increased 8% to $80 million, and health underwriting margin increased 19% to $22 million. The increase in underwriting margin is primarily due to a decrease in claims related to COVID-19. Net health sales were up 11% to $19 million. The increase in net health sales is primarily due to increased agent count. The average producing agent count for the third quarter was 1,371, up 21% from the year-ago quarter and up 10% from the second quarter. The producing agent count at the end of the quarter was 1,469.
We are pleased with the results from Family Heritage as this agent continues to successfully adapt to this environment.
At our direct-to-consumer division at Globe Life, life premiums were up 8% to $228 million, while life underwriting margin declined 17% to $34 million. Frank will further discuss the third quarter decline in underwriting margin in his comments. Net life sales were $44 million, up 50% from the year-ago quarter. As we said on the last call, times of crisis highlight the need for basic life insurance protection, and this has proven true with the pandemic. Application activity and sales were up across all direct-to-consumer channels. At United American General Agency, health premiums increased 11% to $114 million, while health underwriting margin increased 27% to $18 million. The increase in underwriting margin is primarily due to lower acquisition costs. Net health sales were $13 million, down 19% compared to the year-ago quarter. It is always difficult to predict sales in this highly competitive marketplace.
Group Medicare sales are even more volatile and are generally heavily weighted towards the end of the year. Although it is still difficult to predict sales activity in this uncertain environment, I will now provide projections based on knowledge of our business and current trends. We expect the producing agent count for each agency at the end of 2020 to be in the following ranges: American Income, 9,100-9,400; Liberty National, 2,700-2,900; Family Heritage, 1,330-1,530. Net life sales are expected to be as follows: American Income, for the full year 2020, an increase of 3%-7%. For the full year 2021, an increase of 4%-12%. Liberty National, for the full year 2020, a decrease of 2% to an increase of 2%. For the full year 2021, an increase of 3%-9%.
Direct-to-consumer, for the full year 2020, an increase of 32%-36%. For the full year 2021, a decrease of 6% to an increase of 10%. Net health sales are expected to be as follows: Liberty National, for the full year 2020, a decrease of 2% to an increase of 2%. For the full year 2021, an increase of 3%-9%. Family Heritage, for the full year 2020, an increase of 3%-9%. For the full year 2021, an increase of 2%-10%. United American Individual Medicare Supplement, for the full year 2020, a decrease of 25% to flat. For the full year 2021, a decrease of 1% to an increase of 7%. I'll now turn the call back to Gary.
Thanks, Larry. Excess investment income, which we define as net investment income less required interest on net policy liabilities and debt, was $59 million, an 8% decrease over the year-ago quarter. On a per share basis, reflecting the impact of our share repurchase program, excess investment income declined 5%. For the full year, we expect excess investment income in dollars to be down about 5% and down about 1% on a per share basis. As to our investment yield, in the third quarter, we invested $343 million in investment-grade fixed maturities, primarily in the municipal, industrial, and financial sectors. We invested at an average yield of 3.24%, an average rating of A plus, and an average life of 29 years. For the entire portfolio, the third quarter yield was 5.31%, down 16 basis points from the yield in the third quarter of 2019.
As of September 30, the portfolio yield was approximately 5.32%. Invested assets were $18.2 billion, including $16.9 billion of fixed maturities at amortized cost. Of those six maturities, $16 billion are investment grade with an average rating of A-minus. Below investment grade bonds are $840 million compared to $772 million at June 30. Percentage of below investment grade bonds to fixed maturities is 5.0% compared to 4.6% at June 30. Excluding the net unrealized gains in the fixed maturity portfolio, be low investment grade bonds as a percentage of equity is 15%. Overall, the total portfolio is rated BBB plus compared to A-minus a year ago. We had net unrealized gains in the fixed maturity portfolio of about $3.4 billion. Bonds rated BBB are 55% of the fixed maturity portfolio, same as at the end of 2019.
While this ratio is in line with the overall bond market, it is high relative to our peers. However, we have little or no exposure to higher risk assets such as derivatives, equities, residential mortgages, CLOs, and other asset-backed securities. We believe that the BBB securities we acquire provide the best risk-adjusted, capital-adjusted returns, due in large part to our ability to hold securities to maturity regardless of fluctuations in interest rates or equity markets. Because we invest long, key criterion used in our investment process is that an issuer must have the ability to survive multiple cycles. This is particularly true in the energy sector. Our energy portfolio is well-diversified across sub-sectors and issuers, and it is heavily weighted toward issuers that are less vulnerable to depressed commodity prices.
As we have discussed previously, approximately 57% of our portfolio is in the midstream sector, and 34% is in the exploration and production sector. The remaining 9% of our holdings are in the oil field service and refiner sectors. We have no exposure to the drilling sector. The composition of our energy portfolio was essentially unchanged during the third quarter, and the fair value increased approximately $53 million. While we have no intent to increase our holdings in this sector, we are comfortable with our current energy holdings. Finally, lower interest rates continue to pressure investment income. At the midpoint of our guidance, we're assuming an average new money rate of around 3.4% in the fourth quarter and a weighted average of around 3.5% in 2021.
At these new money rates, we expect annual yield on the portfolio to be around 5.33% for the full year 2020, and 5.22% in 2021. While we would like to see higher interest rates going forward, Globe Life can thrive in a lower-for-longer interest rate environment. Extended low interest rates will not impact the GAAP or statutory balance sheets under the current accounting rule since we sell non-interest sensitive protection products. Unfortunately, the impact of lower new money rates on our investment income is somewhat limited as we expect to have an average turnover of less than 2% per year in our investment portfolio over the next five years. I'll turn the call over to Frank for his comments on capital and liquidity.
Thanks, Gary. First, I want to spend a few minutes discussing our share repurchase program, available liquidity, and capital position. In August, the company resumed its share repurchase program. In the third quarter, we spent $118 million to buy 1.4 million Globe Life shares at an average price of $81.79. For the full year through the end of the third quarter, we have spent $257 million of parent company cash to acquire more than 3 million shares at an average price of $83.74. The parent ended the third quarter with liquid assets of approximately $435 million. This amount is higher than normal due to share repurchases through September of $257 million, being less than the $360 million of excess cash flows available to the parent through September, and a $300 million net increase in our borrowed funds since December 31st.
In addition to these liquid assets, the parent company will still generate additional excess cash flow during the remainder of 2020. The parent company's excess cash flow, as we define it, results primarily from the dividends received by the parent from its subsidiaries, less the interest paid on debt and the dividends paid to Globe Life shareholders. Keeping our common dividend rate at its current level for the remainder of this year, we anticipate the parent company's excess cash flow for the fourth quarter to be approximately $20 million. Thus, including the $435 million of liquid assets available at the end of the third quarter, we expect the parent company to have around $455 million available for the remainder of the year.
As I'll discuss in more detail in just a few moments, we believe the $455 million in liquid assets is more than necessary to support the targeted capital levels within our insurance operations and maintain the share repurchase program. As previously noted, during the quarter, the company issued a 10-year, $400 million senior note with a yield of 2.17%. The proceeds of this long-term debt offering, along with other cash at the holding company, were used during the quarter to reduce our short-term indebtedness by over $550 million and to more normal levels. In addition, we successfully negotiated a new $750 million credit facility with our banks that lasts through August of 2023. Now regarding liquidity and capital levels at our insurance subsidiaries. As we continue to navigate this current environment, we are keenly focused on liquidity and capital with our insurance operations.
With respect to liquidity, our insurance company operating cash flows continue to be very strong. In general, while we do expect higher COVID-19-related life claim payments over the course of the year, these higher claims are expected to be largely offset by higher premium collections and lower health claim payments. We do not see any issues with the ability of the insurance companies to fund all remaining dividends payable to the parent during the remainder of 2020. Now with respect to capital. As previously discussed on our earlier calls, Globe Life targets a consolidated company action level RBC ratio in the range of 300%-320%. At December 31st, 2019, our consolidated RBC ratio was 318%, near the high point of our range.
Taking into account only the downgrades and credit losses that have occurred through the end of the third quarter, we estimate this ratio would have declined to approximately 310%. At an RBC ratio of 310%, our insurance subsidiaries have approximately $50 million of capital over the amount required at the low end of our consolidated target of 300%. This excess capital, along with the $455 million of liquid assets we expect to be available at the parent, provide over $500 million of assets available to fund future capital needs. As we discussed on the last call, the primary drivers of additional capital needs from the parent are lower statutory income due to COVID-19-related factors, lower statutory income due to investment portfolio defaults or other credit losses, and investment downgrades that increase required capital.
At this time, we anticipate that our 2020 statutory income, before any realized gains and losses, will be approximately $20 million-$40 million lower than 2019. To estimate the potential impact on our capital losses and downgrades within our investment portfolio, we have modeled several scenarios that take into account consensus views on the economic impact of the recession, the strength and timing of the eventual recovery, and a bottoms-up application of such views on the particular holdings in our investment portfolio. We have also analyzed transition and default rates as published by Moody's and evaluated the impact to our RBC ratios should we experience the same transition and default rates as were experienced in 2001 and 2002, as well as from 2008 - 2010.
Taking into account these various models, we now estimate our RBC ratios would be reduced from year-end 2019 levels in the range of 30-55 points, requiring an additional $75 million-$200 million of capital to maintain a 300% RBC ratio. It should be noted that not all of this additional capital will be required by the end of 2020, as a portion of these defaults and downgrades are expected to occur after the end of this year. Even if all this capital was needed currently, the amount needed is well below the amount of liquidity available at the parent company. Our base case assumes $60 million in total after-tax credit losses, plus approximately $2.1 billion of downgrades to our fixed maturity portfolio.
Through the third quarter, we have experienced approximately $40 million in losses for statutory reporting purposes and $960 million of downgrades, mostly from category NAIC-1 to NAIC-2. It is important to note that Globe Life statutory reserves are not negatively impacted by the low interest rates or the equity markets given our basic fixed protection products. Given the strong underwriting margin in our products, our statutory reserves are more than adequate under all cash flow testing scenarios. At this time, I'd like to provide a few comments relating to the impact of COVID-19 on our third quarter results. As noted by Larry, life underwriting margins declined at both our direct-to-consumer and Liberty National distributions during the quarter. These declines were primarily due to higher COVID-19 policy obligations.
During the quarter, we estimate that direct-to-consumer incurred an additional $10 million related to COVID-19 claims and that Liberty National incurred an additional $4 million. Absent these additional losses, direct-to-consumer's underwriting margin would have been 19.5% of premium for the quarter and would have grown by approximately 8%. In the Liberty National distribution, absent the estimated policy obligations due to COVID-19, their underwriting margin would have been 25% of premium for the quarter and flat versus the year ago quarter. In total for our life operations, we estimate that our total incurred losses from COVID-19 deaths were approximately $18 million in the third quarter and $40 million year to date. Absent these additional losses, our total life underwriting margin would have been approximately 28% of premium and up 4% over the year ago quarter.
Finally, with respect to our earnings guidance for 2020 and 2021, we are projecting net operating income per share will be in the range of $6.84-$7 for the year ending December 31st, 2020. The $6.92 midpoint is consistent with prior quarter's guidance. As I'll discuss in a moment, we do expect higher life policy obligations in 2020 than previously anticipated due to higher projected COVID-related deaths in the U.S. However, at the midpoint of our guidance, we expect the higher life claims to be offset by higher premiums, lower expenses, and higher share repurchases than previously anticipated. On our last call, we indicated the midpoint of our guidance assumed approximately $45 million of claims related to COVID-19 on an assumption of around 225,000 deaths. We continue to estimate that we will incur COVID-related life claims of approximately $2 million for every 10,000 U.S. deaths.
However, at the midpoint of our guidance, we now estimate approximately $56 million of COVID life claims for the full year 2020, reflecting an expectation of approximately 280,000 COVID related deaths in the U.S., higher than previously anticipated. With respect to our health claims, we estimate that our supplemental health benefits for all of 2020 will be approximately $7 million lower than what we expected at the beginning of the year due to COVID, similar to our estimate on the last call. Taking into account the higher COVID life obligations, we expect the life underwriting margin for 2020 as a percentage of premium to be approximately 25.6% at our midpoint. Absent the higher COVID related policy obligations, the life underwriting margin percentage would be similar to the percentage for the full year 2019.
The health underwriting margin as a percentage of premium for the full year 2020 should increase to approximately 23.8%. For 2021, we are projecting net operating income per share will be in the range of $7.30-$7.80. The $7.55 midpoint is a 9% increase from the 2020 midpoint. We are anticipating COVID related life claims in 2021 of approximately $32 million at the midpoint of our guidance, with no significant benefit expected from lower health claims. Obviously, the amount of COVID related claims in 2021 will depend on many factors, including the development of effective therapies and vaccines. The larger than normal range for our guidance reflects this additional uncertainty. Those are my comments. I will now turn the call back to Larry.
Thank you, Frank. Those are our comments. We will now open the call up for questions.
Thank you. If you would like to ask a question, please signal by pressing star one on your telephone keypad. If you're using a speakerphone, please make sure your mute function is turned off to allow your signal to reach our equipment. Again, press star one to ask a question, and we will pause for just a moment to allow everyone an opportunity to signal. Our first question comes from Andrew Kligerman. Please go ahead, sir.
Hey, good morning. I wanted to start with a question on your sales outlook. I'm just kind of looking at the 2021 expectations, 4%-12% growth in sales and American Income , Liberty National [has been in alignment]. That's kind of down six or up 10 after a year where you're up roughly 35%. Maybe a little more color on why 2021 should actually be quite strong based on these guided numbers you've given. You said they've adapted to virtual. Maybe it's the recruiting that's been up so much, but maybe a little color around those numbers. They're really terrific.
First of all, I want to apologize. Your audio wasn't the clearest. I'll try and answer the question. I think you asked why are we predicting maybe sales aren't quite as strong in 2021 as 2020. I think that was the essence of your question.
More around the lines of just that they're very strong in 2021, in my view, both in the agencies and direct-to-consumer. What are the qualities that are enabling that? It looks like recruiting was very strong this year. That probably plays in well. You talked a little on the call about virtual, and how they've adapted. I just wanted a little more clarity on that.
I think your question as you followed up, and thank you, is why would the sales be so strong in 2021?
Yeah.
Let's talk about direct-to-consumer first. Direct-to-consumer, I think it's not likely we're going to have the 50% rate expansion in the third quarter going forward. We do expect this level of increased sales at least through the end of 2020, and likely into the first quarter of 2021. That's really based on the increased demand we're seeing for basic life insurance protection. In the last three quarters of 2021, I think sales growth is going to be more challenging given the large sales increases in 2020.
With respect to the three agencies, again, we see that the demand for both life and health insurance has been very strong. We think as we have the pandemic continue through 2021, whether it's mid-year or through the full year, it's likely to have a positive impact on sales. I think the uncertainty with the agency is that in-person sales and recruiting can be a challenge during the pandemic if the restrictions come back in place. However, we can offset some of those challenges through our use of virtual recruiting and sales.
Let's see. Okay. In terms of adverse selection in this environment, you saw a pickup in pressure on the underwriting margin, and naturally from COVID-19 in both direct-to-consumer Liberty American Income. Could you talk a little bit about the business you wrote, say, from April or March to present, what you've done from the vantage point of putting controls in place to prevent adverse selection? Of those COVID claims that you mentioned, I think you said $10 million of COVID in direct-to-consumer, $4 million in claims from Liberty. What portion of those claims might have come from business written from April on in that?
I'll answer the first part of your question, which is the underwriting process. I'll have Frank address the second part of the questions, which is the actual experience. For the time being, starting in really March, we've limited the maximum face amounts we'll issue to older ages. We stopped issuing additional coverage to existing policyholders of older ages. We also temporarily stopped issuing policy to applicants with certain health conditions. At the same time, our underwriting and our actuarial departments have studied the business on a weekly basis. What we haven't seen is any shift in that business, either by geography, the demographics, or by, when I say demographics, age groups. We think it's consistent in terms of product mix. We don't think there's adverse selection that's occurring. Those are additional steps we've taken, and we'd take additional steps if we saw some development.
Frank, do you want to answer the rest of this question?
Yeah, I'd probably add, the one thing is we've actually seen an increase in the amount of applications with respect to the juvenile block that we have in those older ages. As we know, the most susceptible to claims for COVID-19 are at the older ages. In fact, about 85% of our claims are actually in age of 60 and above. When we look at our in-force as a whole, we only have about 4% of our in-force is over age 70, and around 12% is age 60 or above. Right now, as we look at the claims that we've incurred, about 98% of those have been issued before 2019. With respect to policies issued since March 1st, we have paid eight claims through October 17th, totaling about $42,000.
We have not seen any kind of significant claims on any policy that we've been writing since, really since the first of the year. I will say that the distribution of claims is really pretty well throughout our entire blocks. Probably about 2/3, roughly 2/3 of our claims are coming from policies that were issued in 2010 or earlier. A lot of them are obviously in our older policies, older issues.
Very helpful. Thank you.
Thank you. Our next question comes from Jimmy Bhullar. Please go ahead.
Hi, good morning. First, I had a question on your expense ratios in both the life and health businesses. They were lower than in the past, and I wanted to get an idea on whether it's persistency or something else that's driving that and what your outlook is for expense ratios in the next few quarters.
Jimmy, the primary reason for the reduced expenses is the increased persistency. That's certainly true on the life side. On the health side, it's true, but on the UAGA, we also have implemented a rate increase this year, which also helps drive the expense for premium down. I think for the year on the life side, we're looking at the amortization of being just slightly lower than what we had last year. It'll be more pronounced on the health side, where we'll be more at 18% of premium versus 19% of premium in terms of amortization last year.
On persistency, there were concerns earlier this year that with the weaker economy, you might see a little bit of a drop-off. In reality, it's actually gotten slightly better. What's your view on sort of the reason for that, and are you still concerned about a drop-off in persistency if the economy gets weaker entering, say, this year or next year?
Well, I think that's a possibility that if the economy worsens, that we could see that. We haven't seen it yet, and we've actually seen, as we've talked about, an improvement in persistency. While we're also seeing higher sales. People recognize, in this pandemic, the need for life insurance. That's why more are buying, and then the people that have the policies more are making sure they keep that policy enforced. We've seen improvement in our premium collections. We've seen a reduction in delinquent premiums. It's been positive to this point, and we expect it to continue into the next year because we think the pandemic still will be at the forefront of people's thoughts.
Okay. Just lastly on, how are you thinking in terms of taking advantage of the lower stock price and potentially front-ending some of the buybacks versus the need to sort of preserve capital given the risk of a deterioration in credit?
Yeah, Jimmy, I would say, for the remainder of this year, we're comfortable in being able to utilize all of our excess cash flows for the buybacks for the remainder of the year, which would kind of really point to somewhere in that $120 million-$125 million, to get us up to $380 for the year. That would again be approximately what we have for excess cash flows. We'll take a look to see, as we get close to the end of the year, what happens with the stock price, what happens with the economy, how comfortable we feel with our investment portfolio. We'll consider that if we accelerate some from 2021 perhaps. Right now, I would say that we'd anticipate just really continuing on to utilize our excess cash flows through the remainder of the year.
Okay. Thank you.
Thank you. Our next question comes from John Barnidge. Please go ahead.
Thank you. How many deaths does the $32 million in life claims assume in 2020 guidance? I imagine, there's probably an assumption for improved therapeutics embedded in that.
Yeah. We're using kind of that same rule of thumb for that 2 million, for about every 10,000 U.S. deaths. That kind of have a range. We're kind of estimating 100,000 - 220,000 deaths, and kind of at that midpoint around 160. That 32 would kind of relate to around 160,000 deaths in the year. Really what that kind of supposes is that, we continue to have that the average daily deaths continue to decline and that trend continues over time, but it does continue on into the second and even into the third quarter of the year.
My follow-up. Curious why there's no assumed health benefit in 2021 since there's an assumed COVID-19 life impact. I ask this because I can see how there could be a secular decline in Medicare Supplement claims utilization, given general concern over infectious disease that wasn't present in the U.S. previously.
Yeah, I think from what we see at this point in time is that we anticipate the utilization, especially around the non-MedSup claims, getting really back to normal. We're not expecting any kind of a catch-up, if you will, for missed procedures, but I think without the substantive closures of clinics and such, that we would anticipate just kind of really getting back to more normal levels of both MedSup type claims and appointments as well as traditional medical services.
Thank you very much.
Thank you. Our next question comes from Erik Bass. Please go ahead.
Hi. Thank you. Maybe just to follow up on John's question on the health business. What are you assuming for an underwriting margin in 2021? You had mentioned some lower acquisition costs, so is that something that you would expect to continue into next year?
Yeah. Are you talking about just on the MedSup business or the health business as a whole?
The health business as a whole, just kind of what level of underwriting margin you're assuming percentage-wise?
It'd actually be relatively close and kind of in the same range, in that 23%-24% range for all of 2021.
Okay. Thank you. Apologies if I missed it, did you give the outlook for premiums that you're assuming for both life and health in terms of the year-over-year growth in your 2021 guidance?
At the midpoint of the guidance, we're looking at about a 6% increase in life premiums and a little over 7% increase in health premiums.
Thank you. If I could just squeeze in one more, just on recruiting. I know historically you've talked about seeing sort of a stairstep pattern when you kind of bring in a lot of new agents and then kind of the agent count tends to flatten out a little bit. Is that what you would expect going into 2021 at this point, or how should we think about that?
I'd expect it to be a stairstep process, so I think we'll have an increased agent count. While high unemployment's helped recruiting, we've also had a real addition of middle management American Income. It's grown 22% year to date, and middle managers are really responsible for much of the recruiting that takes place. The 22% increase in middle management, I think we'll see stronger recruiting into 2021. Also, virtual recruiting has allowed us to reach a greater number of possible recruits. Finally, in 2018 and 2019
American Income added approximately 15 new agency owners, and these additional offices have contributed to the increase in agents. While it's typically to be a stair-step process, I think we'll still have increases in 2021.
Got it. Thank you.
Thank you. Our next question comes from Ryan Krueger. Please go ahead, sir.
Hi. Good morning. For 2021, could you provide your margin outlook in percentage terms for the life insurance business? If you have it, what it would be if you excluded your assumption for COVID claims next year?
Yeah, Ryan, for the total life margin, we expect it to be around 26%, and it would be around 27%, 27.1% is what we'd anticipate without the COVID benefits in there.
Got it. In the midpoint of your EPS guidance, the 755, that includes the $32 million of COVID claims. It would be kind of almost $0.25 higher if you did not project those COVID claims.
That is correct. The midpoint includes the $32 million.
Thanks. Just one last one. I think you provided the yield assumption, what are your expectations for excess investment income growth in dollars for 2021?
Ryan, excuse me. Ryan, at the midpoint of our guidance, for 2021, we're expecting excess investment income to be flat. We'll have an improvement in investment income that's going to be offset by the additional interest on the policy liabilities. Virtually, from a dollar standpoint, it'll be flat. From a per share standpoint, it'll be up somewhere around 3%-4%.
Got it. Thank you very much.
Thank you. Our next question comes from Tom Gallagher. Please go ahead.
Hi. Just a follow-up on health persistency, the favorable persistency and the lower DAC amortization this quarter. Are you assuming that benefit will fully continue into 2021, or should we assume some fade of that benefit?
I think we assume it's going to be through 2021. Over time, we'll probably see it revert back more to a normal trend. That's been taken into consideration during our guidance.
Got you. Any particular views as to what's driving that improved persistency? Is it awareness over need for health insurance or any views as to what's been driving that improved persistency?
Yeah, I think you hit on it. I think it's similar to what we're saying on the life insurance side as well. It's the need for the insurance.
Got it. Then just a question on the new FASB LTDI accounting changes. Any sense for when you would expect to disclose expected impacts? If you're able to provide any kind of broader ranges on GAAP earnings or book value that you would expect to be impacted from it?
Yeah, I would guess that we'd look at either toward the end of 2021 or as we get about this time next year, that I would hope that we would start to be able to get maybe some preliminary indications of it. Obviously, we're still working through putting the systems in place and getting our estimates and looking at the impacts of what the new accounting guidance would ultimately be. With COVID, some of the activities that we had been done in that area got put to the side a little bit, so it's not progressing maybe as quickly as it might've been otherwise. The FASB did extend that out a year. I would say, again, whether it be towards the end of next year or the beginning of 2022, we should be able to get some guidance on that.
Okay, thanks.
Thank you. As a reminder to our audience, you may ask a question by pressing star one now. It appears that we have no additional questions at this time.
Okay, thank you for joining us this morning. Those are our comments. We'll talk to you again next quarter.
This concludes today's call. Thank you all for your participation. You may now disconnect.