Are GLWBs All They're Cracked Up To Be?

The guaranteed for life withdrawal benefit (GLWB) in variable annuities is being hailed as the answer to income needs in the retirement market. However, its limitations and suitability for satisfying consumer income needs (particularly when compared to life contingent payout annuities), need to be more fully understood, differentiated and disclosed. Otherwise, consumer, regulatory, and legal backlash are a real possibility.

Without question, GLWBs can be useful for the right client situation. Clearly, the benefit’s main perceived value sounds pretty good: It guarantees to provide a VA policyholder with an amount of money for life, via withdrawals, while the owner continues to invest in the policy’s subaccounts–and the owner does not have to give up access to the nest egg, nor do the beneficiaries have to wait to receive its value at death.

But a closer look points up some potential shortcomings that need scrutiny, as follows.

Is the guaranteed withdrawal amount all that good?

The guaranteed amount has commonly been 5% of principal at issue ages to 65, with higher percentages guaranteed at older ages for subsequent increases in account value. In at least one case, the percentage is increased for inflation, but starts at a lower rate. To consumers, that guaranteed amount may sound better than the 4%-4.5% withdrawal rate recommended as safe by economists.

However, the “recommended as safe” withdrawal rate anticipates increases for inflation. Though ratcheting for account value increases may appear to handle this, account values during the payout phase will most likely decrease, even with reasonable returns.

So again, the question: Are the guaranteed withdrawal rates all that good? Even a female age 65 could receive 6% a year, fully adjusted for inflation and guaranteed for life, from a fixed inflation-adjusted single premium immediate annuity. A 65-year-old male could get 6.7%; a 70-year-old male 8.3%, etc. Innovation could improve these percentages by another 20%. Now, that’s the way to beat the rule–guaranteed. Such a SPIA would have no value on death, but if the primary objective is income, not legacy, this is a moot point.

In addition, after-tax income on non-qualified SPIAs benefits from exclusion ratio treatment, whereas VA withdrawals are taxed on growth first.

Will GLWBs keep up with the retiree’s inflation?

In the February 6, 2006, issue of National Underwriter, an article by Moshe Milevsky used that question as the headline. It presented valuable analysis of GLWBs relative to inflation-adjusted income, concluding that for the product to work best, it should have a ratchet feature and be aggressively invested.

But even on that basis, the GLWB does not keep up with inflation in almost 50% of the projections. It doesn’t even come close 25% of the time.

The failure rate in life contingent income annuities is much lower, by definition. That is “by definition” because life contingent annuities not only guarantee income for life, they also provide a leveraging effect from their survivorship element.

This leveraging effect can be seen in the unbundled illustration shown in the accompanying table. The table shows a life-only payout annuity for a male, issue age 70, assuming 5% interest and 2000 annuity mortality. It results in an annual payout of $9,925 per $100,000 purchase payment (9.9%). (Note: The unbundling is analogous to the way universal life insurance “unbundles” whole life insurance, except that UL’s negative cost-of-insurance is embodied in the annuity’s positive survivorship credit for payouts.)

Under this unbundled retrospective bank account-like view:

o The beginning account value comes from 1.) the purchase payment in year one; and 2.) ending values in subsequent years.

o Interest is credited at 5% on beginning account values.