The Lapse Problem

Policies and a Compensation System Designed for Policy Failure

As discussed in the Mission Statement, investment-oriented life insurance policies (“LIIS”) rarely fulfill the goals of the people who buy them because the policies do not “persist,” meaning that they do not remain in force until the death of the insured, and therefore the policy’s death benefit is never paid. Consumers buy the policies in the hope of being able to draw tax-free income from the policy to support them in retirement or for other uses. But a policy cannot deliver any tax benefits if it does not persist, and, in fact, there may be severe tax consequences to the policyholder if the policy terminates before the death of the insured and policy loans are outstanding. An LIIS policy that fails to persist should be considered a failed policy. By this standard, failure rates for LIIS (80-90%) indicate that the insurance marketplace is not serving consumers well.

The Lure of Tax-Free Retirement Income: Insurance companies and agents who sell LIIS heavily market the idea that LIIS policies can deliver tax benefits that make it a good investment for people who are saving for retirement. Popular books such as Patrick Kelly’s “Tax-Free Retirement” and the “Missed Fortune” books by Douglas Andrew extoll the supposed virtues of LIIS and claim that readers would be better off investing in LIIS than IRA’s or 401(k)’s. Agents often troll for customers by combining a retirement investment pitch or “seminar” with dinner at a nice restaurant. The lure of tax-free retirement income is a powerful one – and there is some truth in the lure because life insurance is indeed favored by the tax laws. Death benefits paid under an insurance policy are not generally subject to income tax, accumulated cash value built up inside a policy is not taxed while the policy is in force, and policy loans are not considered taxable as long as the policy remains in force.

The favorable tax treatment for life insurance that ultimately pays a death benefit sets up an apparently simple strategy for using an LIIS policy to draw tax-free retirement income: the policyholder can invest money in the policy, which can accumulate tax-deferred earnings over time. The policyholder can later borrow money from the policy to support himself or herself in retirement. No taxes are due on the loans when they are taken. Upon the death of the insured, the death benefit, which is tax-free, pays off the loans, and no one ever pays any taxes on the investment gains that accumulated in the policy over time. Sounds great, right?

The idea works in theory, but in practice very few policyholders ever obtain any of the assumed tax benefits because their policies do not persist until the death of the insured. Such policies cannot fulfill the consumer’s objectives because no tax benefits can be obtained if the policy lapses before the death of the insured. If the policy lapses, ordinary income taxes are due in the year of lapse on all gains in the policy regardless of when those gains were made. Worse yet for the consumer, the taxes that are due will not be based on the relatively favorable tax rates applicable to capital gains but will be based on ordinary income rates at the higher marginal rates that result from recognizing all the gains at the same time.

Example:  Let’s say that the consumer -- we’ll call her Alice -- buys an indexed universal life policy that credits earnings to the policy according to the performance of the S&P 500.  Alice wants to use the money to help support herself in retirement and has been told that she can use the policy for tax-free retirement income.  Alice invests $300,000 in the policy and over a long period of time the cash value of the policy increases to $550,000.   Alice retires and begins drawing money from the policy.  The first $300,000 that she draws is not subject to tax because that represents money that she paid into the policy -- what the IRS calls her “cost basis” or “tax basis.”  But beyond $300,000, the money represents investment gains, so in order to draw that money tax free, Alice borrows the money from her policy and uses the loan money to support herself in retirement.  The money she receives in loans is also not subject to tax under current tax law.  Eventually, when Alice dies, the tax-free death benefit will pay off the loans, and no taxes will ever be paid on Alice’s $250,000 in investment earnings.  But what happens if the fees charged on Alice’s policy exceed the interest credited to her policy so that the policy runs out of the money needed to pay policy expenses and lapses before she dies? 

A Tax Nightmare:  Let’s say that Alice’s policy lapses after she has withdrawn her original $300,000 cost basis and taken $250,000 in loans.  At that point taxes would be due at ordinary income rates on the entire $250,000 she had taken out in excess of her cost basis.

If Alice had invested in mutual funds consistent with the index underlying the policy instead of LIIS and had sold her mutual funds over time to support herself in retirement, then she would have recognized gains yearly and paid taxes on those gains at long-term capital gains rates, which are significantly more favorable than ordinary income rates.  (The federal 2020 capital gains rate for a single filer is 0% up to $40,000, 15% between $40,000 and $441,450, and 20% above $441,450.)     Making matters worse for Alice, the $250,000 gain she must recognize when her policy lapses is taxed at the high marginal rate applicable to a person with $250,000 in income in one year.  Recognizing the entire $250,000 all at once puts Alice into the 35% federal tax bracket and the 9.3% state tax bracket (if Alice lived in California), and she would be required to pay all that tax by April 15, 2021!   

Lapse Rates:  The nightmare scenario laid out above is not the only way that consumers can suffer the consequences of buying LIIS policies.  A far greater number of LIIS purchasers never make it to the point of taking loans because their policies lapsed, by surrender or abandonment, before the policyholders ever took a loan.  They suffer no adverse tax consequences, but they lost all (or most, if they surrendered their policies or took withdrawals) of the money that they invested in their policies, while chasing the dream of tax-free retirement income that was rarely realistic given that cumulative lapse rates for LIIS are in the range of 80-90% (or, conversely, given that persistency rates are in the range of 10-20%).

A 2016 study by Daniel Gottlieb and Kent Smetters found that “lapsing is the norm,” relying on annual lapse rate data published in 2011 by The Society of Actuaries and LIMRA (an insurance industry research organization) as well as data about the percentage of life insurance policies that resulted in payment of a death benefit.  They report that “nearly 88% of universal life policies do not terminate with a death benefit claim.”  Perhaps even more troubling is that “76% of universal life policies sold to seniors at age 65 never pay a claim.”  See p.5

A 2021 version of the Gottlieb and Smetters study was published in the American Economic Review and is available for a small charge at https://www.aeaweb.org/articles?id=10.1257/aer.20160868.  The 2021 study reiterates the prior findings and provides some additional data, including findings that 29% of permanent policies lapse within the first three years, and 57% of permanent policies lapse within the first ten years.

Computing cumulative lapse rates from published annual lapse rates requires some math and assumes that the published average annual lapse rates are consistent over the life of the policy.  We will get to the math, but let’s start with the average annual lapse rates, which vary by type of policy.  Perhaps because whole life policies have a known, fixed premium that, if paid, will keep the policy in force for as long as the policyholder wants to keep the policy, whole life policies have lower lapse rates than universal life policies, which are designed to have flexible premiums that give the policyholder more options but also less predictability about what it will cost to keep the policy in force.  Variable universal life policies, in which money put into the policy is invested directly in securities such as mutual funds, have higher lapse rates than other universal life policies, which include traditional universal life policies that pay a declared interest rate, as well as indexed universal life policies, under which the interest credited to the policy is based on the performance of a market index such as the S&P 500.

Data published by LIMRA in 2019 show annual lapse rates for universal life policies of 4.3% for traditional and indexed universal life and 6.0% for variable universal life, while lapse rates for whole life were substantially lower at 2.9%.  [https://www.soa.org/globalassets/assets/files/resources/research-report/2019/2009-13-us-ind-life-persistency-update-report.pdf]. LIMRA’s 2012 study, based on a larger sample of insurers (27 in 2012 versus 16 in 2019), found slightly higher annual lapse rates: 4.5% for traditional and indexed universal life, 6.2% for variable universal life, and 3.1% for whole life.  [https://www.soa.org/globalassets/assets/files/research/exp-study/research-2007-2009-us-ind-life-pers-report.pdf]

Data published by A.M. Best is comparable, showing average annual lapse rates of 6% or more for universal life and 4% or more for whole life; lapse rates for mutual companies were found to  average a full percentage point lower than for stock companies.  [ http://www3.ambest.com/bestweek/purchase.asp?record_code=251134.]  Data published by the International Monetary Fund in 2016 show annual lapse rates of 4% to 6.5% for North American insurers, which are far higher than  lapse rates found in Europe (2% to 3%).  [https://www.imf.org/External/Pubs/FT/GFSR/2016/01/pdf/c3.pdf]

If these lapse rates are applied annually and projected out over a number of years, then the formula for the cumulative persistency rate is specified by the following equation:

100% times [1minus the annual lapse rate]number of years 

This translates to the following persistency rates across a range of annual lapse rates, using assumed averages of 30 and 40 years for the length of time between policy issuance and the death of the insured (though the average time between policy issuance and the death of the insured may significantly exceed 40 years):

Average Annual Lapse Rate Persistencey Rate After 30 Years Persistencey Rate After 40 Years
3%40%30%
4%29%20%
5%21%13%
6%16%8%
7%11%5%

These data show that for the vast majority of consumers LIIS is an investment that is highly likely to fail because the policies will not persist until the death of the insured and will never generate the tax-free income that they are touted to generate. The reason these policies will not persist is that the fees charged are ultimately too high in relation to whatever interest credits are applied to the policy.

The challenge of keeping a policy in force increases over time because cost of insurance rates escalate rapidly as the insured ages. This happens because cost of insurance rates are based on the risk that the insured will die during any given year – a risk that obviously increases as the insured ages. Increases in cost of insurance rates of 1000%, or even far more, over thirty years are not unusual.

Policies sold to low- and middle-income consumers are especially likely to fail because the fees the insurer deducts from the cash value of the policy are usually larger in relation to the amount invested in the policy, which prevents the policy from accumulating enough value to persist until the death of the insured. Low- and middle-income consumers are also less likely to be able to continue funding the policy after adverse financial events or “shocks” (such as loss of employment, high medical expenses, or the needs of dependents), which can make it difficult to continue funding the policy.

Gottlieb and Smetters attribute the observed low persistency rates to an inability by consumers to adequately account for the strong impact that shocks have on policy lapse. They find that this results in the offering by insurance companies of policies with front-loaded fees that “encourage the policyholder to lapse after a background shock, increasing the insurer’s profits.”  Further, despite state regulations prohibiting such practices, many policies “produce cross-subsidies from consumers who lapse to those who do not. The policies are offered even if some of the consumers have correct expectations about all shocks. Moreover, no firm can profit from educating biased consumers about their failure to account for background shocks.”

Whatever the reason for the extremely low persistency rates of LIIS policies, the data is clear that the vast majority of LIIS policies will not persist. And the policies that do persist are especially likely to be held by people with greater economic resources, whose policies have been subsidized by people without such resources. Legislation is needed to remedy this failure of the life insurance market.

Potential Public Policy Approaches: The lapse problem might be solved by a combination of several very different approaches.

Imposing Higher Duties on Agents, and/or Restructuring Agent Commissions and Non-Cash Compensation

Our thoughts on the application of fiduciary or “best interest” obligations to agents selling LIIS are discussed in the Best Interest Section, the public policy arguments for applying such obligations to insurance agents are directed at protecting consumers from being sold policies that are unsuitable for any reason; they are not proposed specifically to prevent lapse. However, we expect that the application of fiduciary or “best interest” obligations to agents will reduce sales of LIIS policies that are most likely to lapse, and thus the application of such obligations is a reasonable and much needed partial solution to the lapse problem.

The same is true of approaches that restructure agent commissions so that agent compensation is tied more closely to the persistency of the policy. Currently, agents receive a highly disproportionate share of first year premiums (some insurers pay commissions even greater than 100% of first year premiums), while renewal commissions are much smaller. Indeed, some companies base all commissions on first year premiums. Under that structure, agents have no financial stake in whether the policy is renewed because all commissions were earned as soon as the policyholder paid the first year’s premium. Commission structures could be modified to give agents a stronger incentive to focus sales efforts on policies that are more likely to persist.

Many agents also receive substantial non-cash compensation (e.g., prizes and trips) based on how much of an insurer’s products the agent sells. These incentives may influence the agent’s judgment and conduct against the consumer in the same way as cash compensation. Indeed, non-cash compensation programs may influence the agents judgment or conduct even more strongly than cash compensation where the agent is close to qualifying for the prize or trip but is not quite there. We agree with the position of the California Department of Insurance (stated in the annuity context), that such compensation schemes should be eliminated. See DOI Letter of June 12, 2019 (“Producers, insurers, and intermediaries shall identify and eliminate any sales contests, sales quotas, bonuses, and non-cash compensation that are based on the sales of specific annuities or specific types of annuities made within a limited period of time.”).

Increasing Consumer Understanding of the Lapse Problem
Understanding how LIIS policies really work

At a minimum, consumers who are considering LIIS need to understand the fees that will be charged against the policy and the fact that no tax benefits can be obtained (and there are serious potential tax detriments) if the policy is not maintained in force until the death of the insured. However, providing this information to consumers, even if it is clearly conveyed and understood, is unlikely to solve the lapse problem. According to Gottlieb and Smetters, consumers have difficulty assessing how their ability to maintain the policy in force may be affected by the risk of future shocks (such as loss of employment or high medical or dependent needs), which is a risk they face regardless of whether they buy the policy but is nevertheless critically important to whether the policy will persist.  The Covid-19 pandemic is a good example of the kind of unanticipated shock that can lead to large numbers of policy lapses.

Understanding the High Risk of Policy Failure

Consumers would benefit from the knowledge that most policies do not persist and thus do not deliver any tax benefits. Disclosure of actual lapse rates applicable to the policy the consumer is considering buying would be very useful. Although it takes many years for the actual lapse history of a particular policy design to develop, insurers assume certain lapse rates in designing their policies. The assumed lapse rates are incorporated into elaborate analyses and actuarial tests (called “lapse-testing”) that insurers perform under regulations supervised (at least in general terms) by state insurance regulators. These lapse rates could be disclosed in a way that lets consumers who are interested in potential tax benefits assess the likelihood that the policy will fail to deliver those benefits. If such disclosure were required, competitive pressure would likely encourage insurers to design policies that can be expected to have lower lapse rates. Requiring disclosure of lapse rates may thus serve as a market-driven way to engineer policies with lower lapse rates.

If insurers do not believe that the lapse rates they assume for regulatory purposes are reasonably accurate projections of their policies’ actual lapse rates, then we invite them to make the relevant data available for regulators and the public to examine.

Requiring Redesign of Policies to Reduce Lapse Rates

Gottlieb and Smetters find that high lapse rates are caused at least in part by the prevalence of policy designs that front-load fees. Such policy designs, which include policies that offer persistency bonuses (fee reductions or interest bonuses given to policies that persist beyond a certain period of years, e.g. ten years) could readily be eliminated or at least discouraged by changes in regulations governing lapse testing.

Persistency bonuses sound like good things, and the few policies that do persist will derive a benefit from the award of any persistency bonuses. But any such benefits are paid for by the losses suffered by those whose policies lapsed. The policies that benefit are more likely to be well-funded and held by policyholders with greater resources, who are subsidized by policyholders with lesser resources and less well funded policies that cannot survive long enough to begin earning persistency bonuses. Moreover, nonguaranteed persistency bonuses increase uncertainty for consumers, who are shown illustrations based on the granting of such bonuses but who have no way to evaluate the likelihood that such bonuses will be provided. State insurance regulators should prioritize a thorough study of how existing regulations affecting policy design may affect lapse rates, including whether persistency bonuses should be eliminated or discouraged.