Ladies and gentlemen, thank you for standing by, and welcome to the MetLife second quarter 2016 earnings release conference call. At this time, all participants are in a listen-only mode. Later, we will conduct a question-and-answer session. Instructions will be given at that time. As a reminder, this conference is being recorded. Before we get started, I would like to read the following statement on behalf of MetLife. Except with respect to historical information, statements made in this conference call constitute forward-looking statements within the meaning of the federal securities laws, including statements relating to trends in the company's operations and financial results in the business and the products of the company and its subsidiaries.
MetLife's actual results may differ materially from the results anticipated in the forward-looking statements as a result of risks and uncertainties, including those described from time to time in MetLife's filings with the U.S. Securities and Exchange Commission, including in the Risk Factors section of those filings. MetLife specifically disclaims any obligation to update or revise any forward-looking statement, whether as a result of new information, future developments, or otherwise. I would like to turn the call over to John Hall, Head of Investor Relations.
Thank you, Greg. Good morning, everyone, and welcome to MetLife's second quarter 2016 earnings call. On this call, we will be discussing certain financial measures not based on generally accepted accounting principles, so-called non-GAAP measures. Reconciliations of these non-GAAP measures and related definitions to the most directly comparable GAAP measures may be found on the investor relations portion of metlife.com in our earnings release and on our quarterly financial supplements. A reconciliation of forward-looking financial information to the most directly comparable GAAP measures is not accessible because MetLife believes it's not possible to provide a reliable forecast of net investment and net derivative gains and losses, which can fluctuate from period to period and may have a significant impact on GAAP net income. Joining me this morning on the call are Steve Kandarian, Chairman, President, and Chief Executive Officer, and John Hele, Chief Financial Officer.
Here with us today to participate in the discussions are other members of senior management. After prepared remarks, we will have a Q&A session. In fairness to all participants, please limit yourself to one question and one follow-up. I'd like to turn the call over to Steve.
Thank you, John, and good morning, everyone. Last night, we reported second quarter operating earnings per share of $0.83. Adjusting for notable items, operating earnings were $1.27 per share, which compares to $1.51 per share on the same basis in the prior year period. Two reserve actions account for much of the operating earnings pressure: $161 million relating to our variable annuity actuarial assumption review, and $257 million for modeling improvements in the reserving process, mostly for our book of universal life policies with secondary guarantees. Compared to a year ago, foreign currency, equity markets, and interest rates were all headwinds in the quarter. As previously announced, we accelerated the VA policyholder behavior assumption review to the second quarter in light of our plan to separate a substantial portion of our U.S. retail business. In addition, we completed our VA economic assumption review.
The balance of our annual assumption reviews will take place in the third quarter as usual. Following these reviews, we have strengthened our VA reserves on a GAAP basis to reflect changes in our lapse in benefit utilization assumptions, resulting in an after-tax non-cash charge of $2 billion. This action clearly had a large impact on our reported GAAP net income of $64 million in the quarter. The reserve strengthening consists of several elements and is associated with our GMIB product. Adverse behavior changes include lower lapses, lower elective annuitization, and higher systematic withdrawals, resulting in increased annuitization under no lapse guarantees and increased utilization of enhanced death benefits. With a 10-year deferral period required to exercise many of the riders in our GMIB product, it was only in the past year that we accumulated sufficient credible data to update our underlying behavior assumptions.
This process was further supported by an independent industry study released in the second quarter of this year. Following these changes, a greater portion of our VA liabilities follow a fair value base GAAP reserve methodology. Prior to the assumption review, more of our VA liabilities were subject to an insurance accounting model, which follows an accrual base GAAP reserve methodology. In contrast to GAAP, in his comments, John Hele will discuss the assumption review process results in greater detail. As I mentioned, we also adjusted reserves in the quarter to reflect modeling improvements and other refinements. The largest impact, $201 million of the $257 million of reserve strengthening, was for our book of universal life policies with secondary guarantees, where we refined the calculation of excess death benefit reserves based on a more granular approach to projecting policyholder premium payments.
Turning to investment performance, variable investment income totaled $285 million in the quarter, which is below our run rate quarterly guidance. Private equity and hedge fund returns were both below plan. Given the performance year to date, it is likely that full year VII will fall below the $1.2 billion low end of our guidance range. One of the most significant geopolitical developments in the quarter was the decision by British voters to exit the European Union. Operationally, MetLife's hub in Dublin largely insulates us from passporting issues. From an investment perspective, our U.K. exposure is limited to $17 billion, with nearly all of those assets currency matched with liabilities. Strategically, MetLife remains committed to the U.K. as an attractive long-term market.
More troubling is the referendum's residual impact, namely, a global flight to safety that's put further downward pressure on interest rates, with the 10-year Treasury yield near record lows. In our view, lower for longer is not going away anytime soon, and life insurance companies will need to adapt to the low rate environment. Turning to strategy, MetLife's Accelerating Value Initiative, combined with macroeconomic factors, led us to pursue a plan to separate a substantial portion of U.S. retail business. We continue to believe the separation will allow U.S. retail to compete more effectively while enabling MetLife to generate a higher free cash flow ratio with less volatility and lower capital intensity. We are working diligently on the separation plan with full engagement from our advisors and internal team. To that end, we've achieved several important milestones on the path to separation during the quarter.
We announced the rebranding of the U.S. retail business as Brighthouse Financial. We made key management appointments. We sharpened Brighthouse's business strategy. We completed the variable annuity assumption review ahead of schedule, which provides key information to help establish Brighthouse's financial foundation. We expect to file an S1 or Form 10 shortly after our September 27th board of directors meeting. We believe a properly executed separation will create value for our shareholders, whether through an IPO, a spin, or some combination thereof. This involves putting together the right strategy, structure, capital, and people. Once we have certainty regarding the capital requirements and final execution plan for the separation, we will return our focus to deploying excess capital. The ultimate timing and form of the separation will be determined by a number of factors, including the capital markets environment.
Turning to regulatory matters, I would like to provide a brief update on the government's appeal of the United States District Court decision rescinding our designation as a systemically important financial institution. On June 16th, the government filed its opening brief with the United States Court of Appeals for the District of Columbia Circuit. Amicus briefs in support of the government's position were filed on June 23rd. MetLife's brief will be filed with the court on August 15th, and amicus briefs in support of our position are due by August 22nd. The government's deadline for filing its final reply brief is September 9th. The clerk of the circuit court has discretion to set the time for oral argument for any date after September 9th.
In the ordinary course, oral argument will be scheduled to take place two to four months after the filing of the final brief, and a three-judge panel will be announced approximately 30 days prior to the date for oral argument. However, the government has requested expedited treatment of the appeal. At this time, we have no indication as to when the oral argument will be scheduled. A final decision in the case is likely to follow within a few months of oral argument. Throughout this process, MetLife will vigorously defend the United States District Court carefully reasoned decision. I would like to conclude this morning with an update on our ongoing focus on expenses, which I discussed on our Q1 earnings call.
As a reminder, I said our goal was to ensure that even on a separated basis, MetLife would have a lower cost structure than it has in a combined basis today. As a result, we are announcing that MetLife is targeting $1 billion in pre-tax run rate expense savings by the end of 2019. This amount is well in excess of approximately $200 million of stranded overhead resulting from our planned separation. It will improve the company's overall unit costs. An example of the kind of initiative that will drive these savings is our recent agreement with CSC, a leading provider of managed services to Fortune 500 companies, to administer nearly seven million MetLife closed block life and annuity contracts. Our new expense target comes on top of the $1 billion in gross pre-tax expense saves that we announced as part of our 2012 strategy and achieved in 2015.
In light of the significant headwinds our industry is facing, MetLife must do even more to avoid simply running in place. On a net basis, our 2012 initiative generated $600 million in pre-tax savings, which has helped bring down our operating expense ratio over the past four years. With these actions, we expect to reduce MetLife's expenses by 11%. We know this will require us to reduce headcount, which is never an easy step for an organization to take. Our overall goal is to be more efficient so that we can better serve our customers and provide a fair return to shareholders. With that, I would like to turn the call over to John Hele to discuss our financial results in greater detail.
Thank you, Steve, and good morning. Today, I'll cover our second quarter results, including a discussion of our annual variable annuity assumption review, insurance margins, investment spreads, expenses, and business highlights. I will then conclude with some comments on cash and capital. Operating earnings in the second quarter were $924 million, or $0.83 per share. This quarter included five notable items, which were highlighted in our news release and disclosed by business segment in the appendix of our quarterly financial supplement, or QFS. Reserve adjustments, primarily resulting from modeling improvements in the reserving process for our book of universal life with secondary guarantees, decreased operating earnings by $257 million, or $0.23 per share after tax. The largest component of this adjustment was in retail life and other.
The new universal life model does a policy-by-policy projection of premiums and death claims, resulting in an immediate increase in future policy benefit reserves, and a decrease in projected earnings. The immediate impact was a reserve increase of $201 million after tax, and an ongoing higher projected reserve increases of approximately $10 million per quarter after tax. The annual retail variable annuity assumption review decreased operating earnings by $161 million, or $0.15 per share after tax. I will discuss the results of the assumption review in more detail shortly. An adjustment to reinsurance receivables in Australia decreased operating earnings by $44 million, or $0.04 per share after tax. We had unfavorable catastrophe experience, which decreased operating earnings by $15 million, or $0.01 per share after tax.
Finally, variable investment income was below the 2016 quarterly plan range by $9 million, or $0.01 per share after tax, and the impact of deferred acquisition costs. Turning to bottom-line results. Second quarter net income was $64 million, or $0.06 per share. Net income was $860 million lower than operating earnings, primarily because of derivative net losses of $1.5 billion, driven by the non-cash charge resulting from the annual variable annuity assumption review. This was partially offset by derivative gains of $725 million, primarily related to changes in interest rates and foreign currencies. In addition, there was $330 million in after-tax derivative gains resulting from other variable annuity derivatives and hedging. The difference between net income and operating earnings in the quarter included a favorable impact of $1.8 billion after tax related to asymmetrical and noneconomic accounting.
Book value per share, excluding AOCI other than FCTA, was $53.20 as of June 30th, up 5% year-over-year. Tangible book value per share was $43.98 as of June 30th, also up 5% year-over-year. Now let me discuss the results of the retail variable annuity assumption review, which we accelerated in connection with the plan separation. As a result of this review, we made changes to policyholder behavior and economic assumptions, as well as the risk margin for the retail variable annuity block of business. Given a 10-year waiting period for the exercise of certain options within our GMIB products, we have only recently begun to observe sufficient and credible evidence to support a shift in our underlying future behavior assumptions for key blocks of this business. Our study consisted of our own emerging experience, combined with recently available relevant industry data.
As a result, we have strengthened our VA reserves, resulting in an after-tax GAAP charge of $2 billion to net income, including an after-tax GAAP charge to operating earnings of $161 million. This non-cash charge consists of three significant components. First, changes in four policyholder behavior assumptions for the guaranteed riders of GMIBs to reflect current company and industry experience. One, we lowered the percentage of policyholders who elect to receive a fixed income annuity under their GMIB rider. Two, we lowered the percentage of policyholders who elect reimbursement of the initial premiums paid when that amount exceeds their current account balance. Three, we increased the percentage of policyholders who elect dollar-for-dollar withdrawals, particularly those in qualified plans at higher ages. Four, we lowered the ultimate lapse rate on certain contracts. Where material, we have differentiated these adjustments between Metropolitan Life Insurance Company and our Delaware company, MLUSA.
The combination of these changes resulted in a partial shift in accounting for these guarantees from mainly an insurance accrual-based model to a more embedded derivatives under a fair value model, resulting in an after-tax charge of $1.5 billion to net income. Second, changes in economic assumptions resulted in an after-tax charge of $279 million. These changes included reducing the long-term separate account return assumptions for variable contracts with traditional mutual funds from 7.25% to 7.0% and managed volatility funds from 7.0% to 6.75%. We also reduced the projected ultimate 10-year treasury rate from 4.5% to 4.25%, which is reached in 2027. Third, the updates resulted in an after-tax charge of $222 million related to the risk margin required on reserves subject to fair value accounting. Let me explain the accounting a little more. GAAP accounting for living benefits is split between accrual and fair value accounting rules.
If policyholders elect to annuitize or they die, this risk is measured under an insurance accounting model, which uses an accrual method reflecting best estimate assumptions. At MetLife, we assume 10-year U.S. Treasury interest rates start at June 30, 2016, levels and increase to 4.25% by 2027, and that long-term separate account returns are 6.75%-7.0%. When policyholders elect dollar-for-dollar withdrawals and exhaust their account values, they receive fixed income annuities. The guaranteed payout portions of those annuities, usually for 5 to 10 years, are considered embedded derivatives and are measured at fair value. The portion beyond the guarantee period is measured using the insurance accounting model. Under fair value accounting, the interest rates used for discounting and separate account returns, including equity fund returns, are calculated actuarially, but on average, are set to current market rates.
For example, at June 30, 2016, the 10-year swap rate was 1.36%. That means that all liabilities are discounted at 1.36%, and all separate account assets are assumed to grow at 1.36% before fees. Contrast this to the modest impact on statutory reserves. Using the same change to policyholder behavior assumptions, the statutory impact is a reduction in reserves of $266 million after tax. The majority of our statutory reserves are calculated using a series of actuarial methods and also assume that the yield curve as of June 30, 2016, slowly rises based on a mean reversion to a level approximately consistent with the long-term 10-year Treasury interest rate of 4.25% in 2027 and an equity return assumption of 7.5%. As you can see, under either accounting system, the ultimate cost of these guarantees will be based not only on policyholder behavior but on future economic conditions.
With respect to second quarter margins, total company underwriting was unfavorable by approximately $0.08 per share versus the prior year quarter, after adjusting for notable items in both periods. Retail life and property and casualty were the primary drivers of the less favorable experience this quarter. Retail life's interest adjusted benefit ratio was 56.7% after adjusting 21.3 points for the notable reserve adjustment discussed earlier. This was unfavorable to the prior year quarter of 53.0% and modestly above the top end of the annual target range of 51%-56%. The year-over-year shortfall was driven by higher frequency and severity of claims. In property and casualty, the combined ratio, including catastrophes, was 108.9% in retail and 103.1% in group. The combined ratio excluding catastrophes was 86.9% in retail and 92.5% in group.
The combined ratio excluding catastrophes was above the prior year quarter of 80.2% in retail and 85.5% in group. We experienced higher catastrophe losses in our homeowners business due to hail and windstorms in the Midwest. Elevated non-cat losses in auto were due to higher frequency and severity. We continue to take pricing and underwriting actions in auto. We believe these actions will yield improved results in the third and fourth quarters of 2016 and will bring results in line with our target combined ratios over the course of 2017. Underwriting results in group life and non-medical health were generally in line with the prior year quarter. The group life mortality ratio was 85.7%, favorable to the prior year quarter of 86.1%, and the low end of the annual target range of 85%-90%. We experienced lower claim severity versus the prior year quarter.
The non-medical health interest adjusted loss ratio was 80.7%, modestly unfavorable to the prior year quarter of 80.5%, and within the annual target range of 77%-82%. Turning to investment margins, the simple average of the four U.S. product spreads in our QFS was 190 basis points in the quarter, down 37 basis points year-over-year. Lower variable investment income accounted for 16 basis points of this decline. Pre-tax variable investment income or VII, was $285 million, down $142 million versus the prior year quarter, mostly due to lower returns on alternative investments. Product spreads excluding VII were 166 basis points this quarter, down 21 basis points year-over-year. Lower core yields accounted for most of this decline. With regard to expenses, the operating expense ratio was 22.7%, favorable to the prior year quarter of 24.3%.
The lower operating expense ratio in the quarter was primarily due to lower employee benefits, expense efficiencies, and the timing of certain projects. I will now discuss the business highlights in the quarter. Retail operating earnings were $184 million, down 73% versus the prior year quarter, primarily due to the reserve adjustments discussed earlier. Excluding these and all notable items in both periods, operating earnings were down 17% due to unfavorable underwriting and lower investment margins. Retail's operating premium fees and other revenues, or PFOs, were $3.1 billion, down 6% year-over-year, primarily due to annuities. Annuities PFOs were $1.1 billion, down 12% year-over-year due to lower fees as a result of negative fund flows and lower premiums due to a decline in SPIA sales. Retail sales were down 21% year-over-year, primarily due to lower variable annuity sales.
VA sales were down 39% year-over-year, primarily due to the sales suspension by a major distributor this year. Conversely, we continue to see strong growth from Shield Level Selector, which more than doubled its sales year-over-year. Group voluntary and work site benefits, or GVWB, reported operating earnings of $221 million, down 4% versus the prior year quarter but essentially flat adjusting for notable items in both periods. The primary drivers were higher auto claims and lower investment margins, mostly offset by volume growth. GVWB PFOs were $4.6 billion, up 4% year-over-year. Sales were up 25% year-over-year, with growth in most group and voluntary products, as well as across most markets. Corporate benefit funding, or CBF, reported operating earnings of $302 million, down 26% versus the prior year quarter, and down 18% after adjusting for notable items. The key driver was lower investment margins.
CBF PFOs were $650 million, up 43% year-over-year due to higher pension risk transfers or PRT. As we have noted before, PRT sales can be lumpy, but we continue to see a good pipeline and remain optimistic about future growth opportunities. Latin America reported operating earnings of $128 million, up 10% from the prior year quarter and up 36% on a constant currency basis. After adjusting for notable items in both periods, Latin America operating earnings were up 40% on a constant currency basis. The key drivers were volume growth, lower expenses, and improved underwriting margins. U.S. Direct, which is included in Latin America's results, had an operating loss of $9 million versus a $20 million loss in the prior quarter, reflecting volume growth and lower expenses. Latin America PFOs were $994 million, down 9%, but up 4% on a constant currency basis.
Excluding U.S. Direct and adjusting for SPIA sales in Chile, which tend to be uneven, Latin America PFOs grew 7% on a constant currency basis. Total Latin America sales were up 29% year-over-year on a constant currency basis. Excluding U.S. Direct, Latin America sales were up 47% on a constant currency basis. This increase was primarily driven by a large group employee benefit sale in Mexico, as well as growth in most markets. Turning to Asia. Operating earnings were $259 million, down 39% from the prior year quarter and 17% on a constant currency basis after adjusting for notable items in both periods. The key drivers were higher taxes and lower fixed annuity surrenders in Japan, partially offset by high investment margins and volume growth. Asia PFOs were $2.1 billion, down 8% from the prior year quarter and 13% on a constant currency basis.
I would highlight two items that negatively impacted Asia PFOs this quarter. First, the deconsolidation of our India operations beginning in 2016 dampened growth by two points year-over-year. Second, the withdrawal in Japan of single premium A&H JPY products in 2015 reduced growth by six points year-over-year. As noted previously, these products do not meet our hurdle rates in the current interest rate environment. Excluding the impact of the deconsolidation of India operations and the withdrawal in Japan of single premium A&H JPY products, PFOs were down 5% on a constant currency basis. The key driver for this decline is a shift from premium-based products to higher value fee-based products. Asia sales were down 3% year-over-year on a constant currency basis, reflecting the impact of management actions to improve value in targeted markets. In Japan, sales were down 13% year-over-year.
We have seen a successful shift in sales to higher return foreign currency denominated life products, which were up 65% year-over-year, and away from low return JPY life products, which were down 54% year-over-year. Japan third sector sales were down 31% versus the prior year as a result of exiting single premium A&H and the negative impact on package sales from a reduction in JPY-denominated whole life product sales. Sales in Asia, excluding Japan, were up 19% versus the prior year. EMEA operating earnings were $64 million, up 28% year-over-year and 36% on a constant currency basis. The key drivers were lower expenses and volume growth. EMEA PFOs were $633 million, down 4% from the prior year period, but at 1% on a constant currency basis. Excluding certain one-time items, PFOs were up 3% on a constant currency basis.
Top line performance was in line with expectations. We continue to see a favorable shift towards high margin products. Total EMEA sales increased 10% on a constant currency basis, mainly driven by growth in employee benefits and Accident & Health policies. In connection with separation starting in 3Q, we will reflect the U.S. retail business, now known as Brighthouse Financial, in a standalone operating segment. This will occur in concert with the resegmentation of the business units that will form the remaining business of MetLife. Among other changes, we will reorganize and report our auto and home insurance business as a single separate segment known as property and casualty. We will have full details on the resegmentation and reformatted financials to you prior to reporting on the third quarter earnings.
A financial reporting impact of the resegmentation will be a GAAP charge to operating earnings in the third quarter of less than $300 million after tax to reflect the loss of the aggregation benefit for GAAP reserve testing associated with the variable and universal life business. Brighthouse Financial will be reported in a standalone operating segment that will no longer receive credit for the broader diversification of the consolidated U.S. retail universal life business of MetLife. As a result, projected earnings for variable and universal life business within the new Brighthouse Financial segment will be adversely affected. Further details will be provided with the third quarter results. I will now discuss our cash and capital position. Cash and liquid assets of the holding companies were approximately $4.9 billion at June 30th, which is down from $5.3 billion at March 31st.
This amount reflects a senior debt repayment upon maturity of $1.25 billion, as well as regular cash flows, including subsidiary dividends, payment of our quarterly common dividend, and other holding company expenses. We did not refinance the maturing debt in the second quarter as we prepare for the ultimate separation of Brighthouse Financial. Turning to our capital position, we report U.S. RBC ratios annually, so we do not have an update for the second quarter. For Japan, our solvency margin ratio was 870% as of the first quarter of 2016, which is the latest public data. For our U.S. companies, preliminary year-to-date second quarter statutory operating earnings were approximately $1.3 billion, and net earnings were approximately $500 million. Statutory operating earnings were down 39% from the prior year, primarily due to the impact of lower interest rates on reserves, largely variable annuities and lower net investment income.
In addition, lower fees earned on separate account balances contributed to the decline in operating earnings. Under statutory accounting, most of the impact of hedging is not reported in operating earnings or net income, rather the gain is recorded in statutory capital. We estimate that our total U.S. statutory adjusted capital was approximately $31 billion as of June 30th, up 6% from December 31st, primarily from an increase in unrealized gains on derivatives of $3.5 billion. In conclusion, we recognize financial results were well below expectations this quarter. Challenging market factors as well as updates to our reserves all contributed to this performance. However, we remain confident the steps we are taking to implement our strategy will drive improvement in free cash flow and create long-term sustainable value to our shareholders. With that, I will turn it back to the operator for your questions.
Thank you. Ladies and gentlemen, if you'd like to ask a question, please press *1 on your touchtone phone. You will hear a tone indicating you have been placed in queue. You may remove yourself from queue at any time by pressing the # key. If you're using a speakerphone, please pick up the handset before pressing the numbers. Once again, if you have a question, please press *1 at this time. One moment, please, for your first question. Your first question comes from the line of Jimmy Bhullar from JP Morgan. Please go ahead.
Hi, good morning. First, how does the charge affect your view of capital? I realize it's a GAAP charge, not a stat charge, and the stat basis you had a positive, how does it affect your view of your capital? Secondly, how does the charge affect just your view of the earning power of the annuity business? If you think about earnings in the last 3 years, I think they've been pretty consistent on an annual basis between $1.5 billion-$1.6 billion, and the size of the charge relative to earnings is pretty large. How do you think about the earnings power of the annuity business given this charge?
Hi, Jimmy. This is John. In terms of the statutory reserves, you saw there was a small release.
Statutory capital, which is at CTE90, has a slightly smaller change. The companies have good RBCs right now for that. I'd like to just caution that changes to statutory reserves are not the true determinant of overall capital. There's a variety of factors and that, looking forward, when Brighthouse is separated, it'll be a standalone company, not part of a broader diversified MetLife. Those will have to be as we file our S1 or Form 10, you'll see more details as to the capitalization of Brighthouse as a standalone company. In terms of the earnings power, the earnings power is operating earnings, and those are based on the fees of the base benefits. We see fees sort of leveling off a bit here.
You saw in this report on the statutory that the VA fees are sort of leveling off because the total balances, we're seeing some negative flows. I think you'll see that staying more steady going forward.
Your next question comes from the line of Seth Weiss from Bank of America. Please go ahead.
Hi, good morning. Thanks for taking the question. John, want to dig into the charge a little bit more. It sounds to me like at a high level that the charge could be split between partial what's an assumption review, but also moving from this accounting standard of the insurance SOP 03-1 to the embedded derivative convention. If we try to bifurcate that $1.5 billion of policyholder charges between what's assumption changes and what's a change in the accounting standard, can you help us think how we would split that?
It's a great question, but it's very difficult to do because we have changed several assumptions. If you lower annuitizations, which we did, and increase dollar for dollar, that's an assumption of what people are doing, but then the accounting changes it as well. It's interrelations between all these. The actuaries call it the cross effect, and it's a pretty big number. If I give you one piece, but there's a big cross effect in how they all add up, and we can't reattribute the cross effect to it. It's a little misleading to just look at one piece of it. They're all interrelated as you do all these, and so in the end, it's really the total number that we have to communicate.
Okay, maybe just-
You can see the increase to the FAS 133 reserves is a significant increase, and it's most of that $1.5 billion.
Okay. Just maybe backing to the economics of it all, you gave four points as to what the policyholder behavior assumption changes were. It just seemed, I guess, taking a simplified view, that if you increase the dollar for dollar utilization, that tends to be bad news for a charge. Conversely, if you increase the annuitization update, that tends to be positive for the business. I'm not sure if I'm reading that correctly, but I guess my question is twofold. Is that correct? Do I have the right read on that? Why is that the case?
We lowered the number that take annuitizations based on our experience, and increased, particularly at older ages, we're seeing more people taking dollar for dollar and in particular in qualified plans. How these things interact with each other depend upon the future fund performance and where interest rates are. You could see under GAAP, which for dollar for dollar, we use the fair value accounting, which assumes much lower returns for separate accounts and interest rates. There's a cost to that. Whereas in statutory, which assumes a mean reversion of interest rates and separate account returns that are reflecting equity markets that are, say, call it more normal, there's actually a decrease in reserves. There's a switchover point here between what happens to economic assumptions between the ultimate cost of these. Why is that?
The dollar for dollar becomes a cost if you run out of money in your account, and then you trigger the guarantee. If you have okay investment performance, you may not trigger the guarantee. You'll have enough money in your account for a long period of time. If your separate accounts are earning 1.36%, you're going to run out of money, and you have more people triggering the guarantees. It's a combination of these two, and where all this ends up will likely be someplace in between these. It'll depend upon what policyholders do as well as what equity markets are and what interest rates are.
Okay, great. Thanks a lot.
Your next question comes from the line of Jay Gelb from Barclays. Please go ahead.
Thank you, and good morning. On the third quarter actuarial assumption review, do you have any perspective in terms of the potential magnitude of that impact?
We did accelerate not only the policyholder behavior review, but also the economic assumption review and the risk margin review for the VA business, which is a very large book of business relative to the total risk profile of MetLife. The other reviews that we do every year on mortality and morbidity and the other blocks of business, we'll just have to wait Till we get there. If you look back over MetLife's history, those tend to be more modest.
That's helpful. Thanks, John. Also, based on the timing of the filing of the S1 or Form 10 for Brighthouse Financial, should we be expecting any share buybacks in 2016?
As I mentioned in my remarks, it's Steve here. Our first order of business is to work on the separation and make sure we have a clear understanding about capitalization of Brighthouse Financial to form the separation. Once we get through that, we will focus on the issue of how we'll use any remaining excess capital. We're not at that point yet where we can really speak to when share repurchases may begin.
I appreciate that, Steve. Post the transaction, would you still anticipate that Met can meet its targeted goals of return of capital as a percentage of earnings?
Our philosophy hasn't changed. Our Accelerating Value Initiative is based upon driving up, among other things, the ratio of free cash flow to total operating earnings. We think the separation plan, as well as expense initiative that we just announced, goes to that as well. We've also said philosophically many times that excess capital belongs to our shareholders, that it's in the form of dividends, share repurchases, and any acquisitions that make strategic sense that are accretive to shareholders over time.
Thank you.
Your next question comes from the line of Thomas Gallagher from Evercore ISI. Please go ahead.
Good morning. My first question is just thinking about the third quarter review. John, you had already mentioned the $300 million separation related charge due to lack of a diversification benefit. As you think about anything else or areas to watch out for, can you highlight any particular areas, whether that's goodwill or DAC, that we should be focused on where the review will come under more scrutiny in 3Q?
Well, there'll be two things going on in the third quarter, just to clarify. There'll be our annual assumption review for other than variable annuities, there's the resegmentation. Both can have an impact, as I've said historically MetLife, other than for variable annuities, the assumption review has been more modest. The resegmentation we've highlighted for you what we've seen already in terms of the resegmentation of the diversification for universal life and variable life not having the diversification credit across all of MetLife. We will be checking our goodwill in the third quarter, but I can't give you any insight into that until we get all that work done.
Okay, then the follow-up is, you had mentioned after the resegmentation occurs, there's going to be an adverse effect. You mentioned the charge, but is there also going to be ongoing adverse effect on the operating results? Is it going to be a lower level of operating earnings from the variable annuity business and the life insurance business? How should we think about that?
Yes, I did say that we expect there could be some ongoing impacts, but we won't be able to clarify that for you until we get to the third quarter.
Okay, thanks.
Your next question comes from the line of Ryan Krueger from KBW. Please go ahead.
Hi. Thanks. Good morning. First on the cost saves, the $1 billion growth amount, are there any reinvestment assumptions that would net to a lower amount other than the $200 million of stranded overhead?
Yes, Ryan, there will be investments. We're still working out the details of those. As we develop our plans in more detail, we'll share with you really the net impacts. By the time we get to 2019, we expect almost all that to be falling through to the bottom line, but there will be investments. We did the $1 billion saves in 2012. There was a lot of reorganizing. This next set of will be really quite a structural and an investment in processing and efficiencies such as you saw the CSC announcement that we did. This next $1 billion will take some investment to get the long-term cost saves, but we do plan to have this flow through to the bottom line end of 2019 run rate, which will be 2020 going forward.
Okay. Got it. On the corporate segment, I guess you're at $411 million of year-to-date losses, which is higher than the $500 million-$700 million full year expectation you had. Can you give us any update on how to think about the corporate segment going into the second half of the year?
Yeah. This quarter was higher. We had lower VII, which impacted it. Just happened to be the assets that were there. The alternative investments were lower. Taxes had an impact in the quarter there. Corporate is used to level out taxes for the year. We had to take a charge to level out the tax rate for the year. Those are more one time.
Okay. Thank you.
Your next question comes from the line of Suneet Kamath from UBS. Please go ahead.
Thanks. Just wanted to follow up on the expense savings of $1 billion.
How should we think about that splitting between the separated company and old MetLife?
This will all be for RemainCo.
Okay. Got it. As we think, I guess, for Steve, about the ROE of the company, I know there's lots of moving pieces, but by our calculation, you're somewhere in the 9%, maybe 9.5% range on a normalized basis. I think in the past, you've talked about 11% as a target. I guess I'm just trying to understand what gets us from nine and change up to 11 in this low interest rate environment.
Suneet, the things you've heard about over the last couple of quarters, including the separation of the U.S. retail business, as well as the expense initiative that we just announced, are all related to driving that ROE number back up to a range that we'd like to see it at.
Is that range still around 11, or?
Well, we'll talk to you more about ranges at Investor Day in December.
Okay, thanks.
Your next question comes from the line of Eric Berg from RBC Capital Markets. Please go ahead.
Thanks very much. John, would you just remind us briefly all the choices that customers have at the end of the waiting period for the GMIB? You mentioned, of course, that they can annuitize at the end of the waiting period. Under your most popular contract forms that you've written over the years, what are the other paths that a customer could take besides annuitization?
Okay. Although it does vary by contract, the GMIB has, after a 10-year waiting period, the ability to take your balance and get an annuity to be paid for life, based on your benefit base, not on your account balance. You get the greater of current annuity rates or a guaranteed set of annuity rates that do have a setback in. That's with the benefit base or your current account balance at current annuity rates. You can also then, for many of the contracts, not take that option and instead elect dollar for dollar, which is almost like a WB benefit for life. If you run out of your dollar for dollar money by taking up to your percentage allowed, if your account balance goes to zero, then you get an annuity for life.
That almost gives you the same thing, and we're seeing less people take annuitization options and more people elect dollar for dollar, particularly as they age and more in qualified plans. There's another feature that I spoke about that allows sort of a refund of your initial premiums paid in the first, I think, 120 days. We call it the multi-principal option. You can surrender your rider and get this top-up of money to be your premiums paid, but you lose then your other benefits of the other rider features for it. Then, you can always surrender your contract for cash at any time. There's no waiting period, but you have a surrender charge for the first typically five to seven years, depending upon the contract.
I have one second and final question regarding universal life. I'm trying to understand better than I do why this matter of looking at things on a policy-by-policy basis would lead to a reserve increase in universal life. The reason I ask the question is, life insurance companies are always dealing with groups of policies, cohorts, averages. There's no reason to think that using an averaging method rather than a policy-by-policy approach would lead to a bias of overstatement or understatement. Why is, if you follow my question, I understand that going with a policy-by-policy approach is more precise than using an average, but why didn't the average work, and why did it lead, in other words, to a reserve deficiency, this averaging approach?
Within the cohorts that were being modeled before, there was a much wider variation than the average within the cohort. Although there was an average amount of premiums paid within a group, there was a much bigger percentage of people paying the minimum versus people paying a higher amount. We were underestimating the number of policyholders that will trigger the guarantees, and they trigger them sooner. There's a wider dispersion. By triggering policy by policy, it came out to be a more precise calculation than the average way that was being done previously.
I got it. Thanks very much.
In terms of, to put the charge into perspective, the universal life piece of the charge of the 257 was about $200 million. That's about 1.6% of the total reserves for universal life with secondary guarantees. It's about $12.5 billion. It is a true-up of reserves. It's a better method to calculate it, which we do from time to time. In terms of perspective, it's not that material of a change.
That's helpful. Thank you.
Your next question comes from the line of Randy Binner from FBR. Please go ahead.
Thanks. Actually, just two questions. One follow-up there with Eric's question. Are there any other major blocks that need to be migrated to these more granular systems? This is in regard to the SGUL. Just wondering if we might see similar reserve charges if you continue this, what I think is a technology migration.
This is one of the biggest and most complex areas. Of course, lower interest rates has made this a larger reserve needed as universal life with lifetime secondary guarantees. Lower rates have made these contracts to be more valuable to their customers. Therefore, we've had to increase the reserves on them.
Great. Then back to the main VA charge. I'm just trying to think at a higher level. It seems like once people exhaust their separate account, you kind of have a cliff effect where the swap rate becomes your discount rate rather than the mean reversion assumptions under stat. I think I'm saying that right. If that is the case, can you size what percentage of people you think in that overall VA block we're going to continue to monitor at Brighthouse are expected to exhaust their separate account and therefore be exposed to the 10-year swap as a discount rate?
We're seeing less people annuitize. If you annuitize, then you don't pay dollar for dollar. Less people annuitize, more people pay dollar for dollar. The actual calculations are highly complex. We run tens of thousands of runs over a wide range of scenarios. When we talk about these interest rates, these are the average interest rates, both even in stat and GAAP. It's just the mean of those interest rates in stat reverts to roughly this 4.25% in 2027, whereas the GAAP piece, if you pay dollar for dollar, a piece of your risk, the guaranteed piece, the certain period of your annuity, has to be reserved under GAAP, using a mean reversion interest rate of 1.36%.
Right.
There's a very wide difference, including your separate account returns. I want to stress, that's the separate account return before fees. You generally have a negative return in your separate accounts. You can see how that can have a high cost to this. If interest rates migrate higher, you have to mark this every quarter to the current yield curve. As interest rates go up, this charge becomes less as you go through time. That's why I wanted to give you the wide range of the changes because what the ultimate cost of these things will be will depend on interest rates and what policyholders do. This is over a long period of time. Our average age of these contracts is in the 60s, and people will elect these options well into their 80s and higher.
These are long-term contracts, and how this works out over time will be the ultimate cost. We do match this to our experience, and we do it on a regular basis as we get credible experience. We now have a new industry study that we just got that had some good credible experience for GMIBs, and that's why we set our reserves appropriately.
Thank you.
Okay. We're at the top of the hour. We're going to bring the call to a close. Thank you to everyone for joining us today. Have a good day.
Ladies and gentlemen, that does conclude your conference for today. Thank you for your participation and for using AT&T TeleConference Services. You may now disconnect.