Executive Summary

Life insurance policies are primarily bought to provide a benefit when the insured dies. But the termination of a life insurance policy before death can occur for a number of reasons:

• The policy is no longer needed
• The policy can be exchanged for a new, more suitable policy
• The policy’s cash value is needed for another purpose outside of the life insurance plan

Statistically, every policy has some chance of being surrendered. Buyers should consider the likelihood of a surrender when evaluating the policy’s cash value. Using a reasonable probability of surrender, buyers can evaluate a policy’s three performance measurements: premium, death benefit, and surrender value.

Why is Cash Value Important?

The need for a death benefit is the primary motivator for a life insurance purchase. As a result, the buyer may only consider the possibility of death when selecting the optimal policy and may not contemplate the surrender or exchange of the policy for its cash value.

Table 1. Permanent U.S. Life Insurance Outcomes, 2020
Values in Millions

Statistically, it’s just about a coin toss whether a policy is surrendered or pays a death claim, based upon financial statement data life insurers filed with state regulators. A review of the statements (Table 1) filed by the top 50 life insurers ranked by in-force, permanent face amounts (representing 80% of the permanent face amount in the industry) shows that 1.4 million permanent policies were surrendered, while 1.2 million policies paid death benefits in 2020.

A total of 57% of the top 50 companies’ policy count are permanent policies. If term insurance is excluded,² the surrender ratio adjusts to 2.3%. Based on this data, we believe that a permanent policy is approximately twice as likely to be surrendered as to pay a death claim.

This demonstrates that surrender benefits should be a primary consideration when selecting an insurance policy—not just an afterthought. But putting a value to surrender benefits can be tricky.

An approach we suggest is to pull from actuarial methods and to think about policies in terms of probabilities. Specifically, we know that for every year that a life insurance policy is in force, there is a probability that the policy will be surrendered or will pay a death claim. The probability the policy will pay a death claim is based on the mortality rate for the individual.³ The probability the policy will be surrendered depends on several factors (e.g., client liquidity needs and market conditions), most of which cannot be predicted. But we can make a simple estimate based on historical data, and based on the preceding data, we’ll assume the probability of surrender is 3% for this analysis. Acknowledging the probability of the insured’s death, the probability a policy will have been surrendered will increase over time.⁴ Table 2 shows the effect of the surrender assumption over time.

Table 2. Probability of a policy remaining in force with an assumed annual surrender rate of 3%⁵

An assumed annual surrender rate of 3% results in more than a one-in-four chance that the policy will no longer be in force by the end of the 10th policy year. And the odds of the policy lasting 30 years are less than half. Even if you assume a lower surrender rate of 2%, the odds of it no longer being in force by year 20 are one-in-three.

It’s important to keep in mind that a policy surrender isn’t necessarily a bad thing—it simply means the policyholder’s needs changed. Higher cash surrender values help position buyers for that potential outcome.

A Tool to Evaluate Cash Value: Mortality and Surrender-Adjusted IRR (MSIRR)

With the material probability that the policy could be surrendered, it’s meaningful to have the ability to calculate a value for the cash value in a policy over time. To accomplish this, we can look at the IRR of the expected policy cash flows in each year and consider the probability of death and surrender for each year of the policy. (See sidebar “What is Mortality and Surrender-Adjusted Internal Rate of Return (MSIRR)?” for details about the calculation).

For example, let’s assume a policy with a $1 million death benefit has a cash value of $100,000. If there is a 3% probability that the policy will be surrendered over the next year, then the expected value of surrender is $3,000 ($100,000 * 3%). Likewise, if there is a 0.5% probability of death, then the expected value of the death benefit is $5,000 ($1,000,000 * 0.5%). This same calculation can be performed each year for an illustration by adjusting the probability the policy will still be in force.

To understand how different products perform under this metric, let’s compare illustrations for two different variable universal life products. The illustration design is a 10-pay premium to endow the policy at age 120 for a male, age 55.⁶ Product A has an annual premium solve of $41,418. Product B has an annual premium solve of $43,548, which is 5% higher than Product A. A rudimentary analysis of the two products might lead one to conclude that Product A is better than Product B because the illustration assumes a smaller premium is required to achieve the same result. But that does not account for each policy’s cash surrender value performance.

Below are the illustrated values for both products. Using a 3% assumption for the probability that the policy could be surrendered each year, we can now factor the difference in surrender values into the analysis by doing a MSIRR calculation.⁷

What is a Mortality and Surrender-Adjusted Internal Rate of Return (MSIRR)?

MSIRR is computed by taking the probabilities of both mortality and surrender into account. For each duration of the illustration, a portion of the illustrated surrender value and death benefit is attributed as cash flow, based on the probabilities of surrender and mortality. The remaining portion of the policy values are reduced by the rate of surrender and mortality [Expected survival rate of policy _t+1 = Expected survival rate of policy_t * (1–r _surrender) * (1-r _mortality)]. This process is repeated each year of the illustration until the expected survival rate of the policy = 0. The IRR is then calculated based on the expected premiums, expected surrender values, and expected death benefits.

The MSIRR is 4.95% for Product A and 4.99% for Product B. The reason the value is higher for Product B is due to its larger surrender values as shown in Table 3, which is important when the probability of surrender is factored in. In other words, Product B is more “expensive,” but at a 3% lapse assumption, that additional cost is worth it to get the higher cash values.

On the other hand, if we assume the probability of surrender is zero, Product A has an MSIRR of 6.02% vs. 5.80% for Product B. This finding makes sense. If we assume surrenders never occur, the larger surrender values found in Product B are not beneficial to the client. Table 4 shows the impact of differing surrender rate assumptions on the MSIRR calculation.

Table 3. Illustrated Ledger Values for Products A and B

Table 4. MSIRR Comparison for Varying Surrender Rate Assumptions

A Comprehensive Approach

This analysis demonstrates the importance of factoring in the possibility a client will want to access the policy’s cash value for any of the reasons described earlier: the policy is not needed, a more suitable policy is available, or the funds are needed. By focusing exclusively on the death benefit rather than taking a more comprehensive approach to the uses for life insurance and the reality that needs can change, a buyer may be overlooking better choices for life insurance.

• ¹ IRC §1035 allows for a tax-free exchange of life insurance policies that are considered to be of a like kind. Gains from the surrendered policy are not recognized for tax purposes, and the new policy retains all of the tax elements of the prior policy.
• ² Term insurance policies have no cash value and cannot be surrendered for cash value, but any unearned premiums may be returned if terminated early.
• ³ Based on a variety of factors including sex, age, and health condition.
• ⁴ This can be calculated by multiplying the probability of the policy remaining in force by (1 - r_surr) where r_surr is the assumed rate of surrender. The probability of the policy remaining in force for a period of time (“t”) can be calculated by the formula (1 - r_surr)^t.
• ⁵ Ignoring the probability of death occurring.
• ⁶ Illustration solving for 10 premium payments to make the policy account value equal to the face amount ($2 million) by age 120, based on a 7% gross rate of return with premiums allocated to the policy’s S&P 500 index fund. The insured is assumed to be in the best underwriting class.
• ⁷ The illustrated internal rate of return with the probabilities of mortality (death) and surrenders factored into the rate of return calculation. The mortality component of the calculation can be determined using an industry mortality table such as the Society of Actuaries 2015 Valuation Basic Table.