Variable Life Insurance: Be Wary of Policy 'Delusions'

    by Peter Katt

    “My stocks ‘averaged’ 4.1% but my variable life account is down 40%!”

    This is one individual’s situation seven years into his variable life policy that he employs for the dual purposes of family protection and tax-deferred investing to enhance his retirement resources.

    Everything was done right for Dr. Smith:

    • He has a pure defined-contribution design with maximum premiums relative to very low initial death benefits while maintaining it as a non-MEC (modified endowment contract) policy, which allows it to retain certain tax advantages.

    • He was shown simulations displaying the difference between volatile variable life equity returns versus far more tranquil whole life returns over his lifetime, the-hare-and-the-tortoise of life insurance comparisons.

    • A low-load policy was used. He put 100% of his premiums and cash values into Vanguard’s growth fund where he intends to let it ride until retirement. And there is none of that risky market timing going on in this fund.
    What happened?
    The returns for the fund over the last seven years were: 18%, 33%, 24%, 20%, –43%, –36% and 13%. The fund’s arithmetic average—calculated by adding up the year-to-year return results and dividing by the number of years—is 4.1%; the arithmetic average is what you often see in variable life illustrations.

    However, an investor’s actual average return is not an arithmetic average, but a compounded (geometric) average. As an example of how different the arithmetic average is, consider an investor who starts with $100 and doubles his money to $200 in one year (a 100% return) but then loses 50% the following year, and is back to his original $100. The arithmetic average is 25% [(100% + (–50%)) ÷ 2], yet the investor has made nothing.

    An investor’s actual return can also be affected by the timing of cash flows into and out of the policy, as was the case here.

    The bottom line is that Dr. Smith’s policy has actually suffered a 12.5% average annual loss over seven years, which has resulted in 40% less value than what he put into the policy in the first place.

    This loss was not solely caused by the volatile stock returns in the portfolio. A second, but much less important, factor is the expense associated with variable life.

    Dr. Smith’s defined-contribution design used Option B with a specified death benefit plus the cash value. Because he really doesn’t need these larger death benefits at his age, 61, we have changed his Option B to Option A (cash values included in the death benefit) with its smaller death benefits and, therefore, lower premium costs. Over his lifetime, this change from Option B to A will reduce his costs by about the amount of his equity losses to date, and future investment results may also get his actual returns on the plus side.

    Variable life as insurance—not!
    Dr. Smith’s situation is well worth remembering the next time you are shown the arithmetic average for equities as a way of promoting variable life as the preferred insurance asset. The arithmetic average is visually supported by variable life illustrations that must show a constant yield.

    Variable life illustrations create an illusion of predictability that simply doesn’t exist. An illustration showing a 4.1% gross yield to demonstrate safety to Dr. Smith would have been very misleading. And if Dr. Smith had been sold a defined-benefit design with large level-to-maturity death benefits relative to some known premium cost, his policy would be insolvent right now without him facing the equivalent of a margin call. (See my November 2001 AAII Journal column about the horrors of using variable life for defined-benefit life insurance designs, “Variable Life Insurance Policies and Stock Market Volatility.")

    As an aid to deciding between variable and participating whole life for the purchase of a defined-contribution type policy, the following analysis may be helpful.

    Generally, variable life’s expenses are higher than participating whole life by an equivalent of 65 to 90 basis points in yield. That is, variable life has to earn 65 to 90 basis points more than whole life to be even.

    Since 1926 large-cap equities have had an average annual compound return of 10.2%, while bonds have had an average annual return of 5.9% but with much less volatility.

    If you factor in that the best participating whole life companies (for instance, Northwestern Mutual and Guardian) have the liquidity ability to commit up to 25% of their general portfolios to private and public equities, but protect their policies from year-to-year losses, the real historical difference between a long-term investment in participating whole life and variable life is around 175 to 200 basis points.

    This is especially important for older clients whose time horizons may make variable life far more risky because they may not have the time to earn back their investment mistakes.

    Three rules for using variable life
    Because of unpredictable and volatile equity results, as well as high policy expenses we can fashion three rules about variable life insurance:

    • One, never use it with a defined-benefit life insurance design that has large level-to-maturity death benefits and the fantasy of knowing the amount of premiums.

    • Second, never believe the tranquil illustrations showing constant equity results as a guide to making a decision between variable and participating whole life for defined-contribution life insurance designs that have minimum initial death benefits relative to expected premiums.

    • And, never buy variable life with visions of huge equity returns and then chicken out and select fixed-income investments, because those higher policy expenses will make it a sure loser compared with using participating whole life.

    Rating Financial Strength

    Nearly every article about life insurance in the mainstream press has the obligatory reminder for readers to check the financial strength ratings of companies before buying life insurance from them. This mantra has also been picked up by many advisors.

    Paying attention to financial strength ratings is a prerequisite, but should not be determinative in selecting a life insurance company and policy any more than hiring a heart surgeon into a practice should not be based on their having an undergraduate degree. Of course they have an undergraduate degree and of course such-and-such insurance company has sparkling financial ratings—but what are their real attributes? Regarding the purchase of permanent life insurance among the list of companies with outstanding financial ratings, you need to find those very few companies with outstanding current pricing that give fair pricing treatment to old as well as new policies. These few companies stand in sharp contrast to the many companies with outstanding financial ratings and excellent current pricing that do not provide old policies with the same excellent pricing. Because the probabilities are that permanent life insurance is a very long-term asset, it is critical that you judge companies by their fair treatment of old policies relative to the current pricing on newer policies.

    Almost all companies selling universal life have a history of repeatedly coming out with newer versions of their basic policy, with each newer version improving upon the old with lower insurance costs and higher interest rates. Many mutual companies (but not all) have done a much better job of providing old policies with fair pricing relative to new ones.

    Using financial strength ratings as the sole determinative measurement for life insurance companies is foolish. Financial strength should only be a prerequisite for developing a list of companies whose fair treatment of old policies can be tested.

    Peter Katt, CFP, LIC, is sole proprietor of Katt & Co., a fee-only life insurance advisor located in Kalamazoo, Michigan (269/372-3497; His book, “The Life Insurance Fiasco: How to Avoid It,” is available through the author.

→ Peter Katt