Next-Gen vs. Traditional VAs

Over the past several years, I've published a few articles questioning the value of deferred variable annuities with income guarantee riders (VA/GLWBs), including one in the Summer 2013 issue of the Journal of Retirement. These VAs with guarantees are marketed as offering upside potential, downside protection, and liquidity all in one convenient package. But my concern is that the impact of compounding fees over time creates an overwhelming cost to the VA/GLWB user, such that one could be better off by just combining stocks and simple income annuities. That was a conclusion in my article about the efficient frontier for retirement income. 

Along these lines, Jefferson National asked me to write a sponsored white paper [you are supposed to be a financial advisor to gain access to the paper, and this VA is not otherwise available directly to consumers] about their new Monument Advisor investment-only variable annuity designed to be used by financial advisors.  Fees for the Monument Advisor variable annuity add up to $240 per year for an account of any size, and the reason to consider this VA is because the advisor and client have already determined that there is value to the financial plan by seeking the tax deferral offered by variable annuities. Besides these tax deferral aspects, the VA otherwise basically behaves like a traditional investment account, at least after age 59.5.

So the question becomes: should someone seeking tax deferral through a variable annuity use the low-cost Monument Advisor investment only approach, or should they go ahead and proceed with traditional VA with guarantee riders, which may also include commissions, insurance, and guarantee fees?

In the article I reinterpret this question in terms of: what will be the difference in outcomes for someone using a lower-cost investment-only VA vs. someone using a higher cost VA with guarantees, in terms of the retirement income that can be supported.

Obviously, if the investment portfolio is depleted, someone will be happy to still have guaranteed income. But the value of that guaranteed income can be oversold. Fees in the VA/GLWB will have eaten away the account value more quickly, so that there will be no growth in benefits to keep pace with inflation, and there will be no liquidity either. As inflation erodes the value of the VA/GLWB's guaranteed income, the real value of this guaranteed income could become much less than the user realizes. When markets are down, a VA/GLWB ends up behaving like a SPIA, but with a lower payout rate. And the question is: what has the VA/GLWB user given up in the process of seeking the somewhat illusionary upside potential and liquidity for their assets? Again, the illusionary nature of these is that compounding fees eat away at the potential for either upside or liquidity when it may be most needed later in retirement. 

In the white paper, I first review the points made in favor of using VA/GLWBs, including tax deferral, the ability to lock in growth for a hypothetical benefit base during accumulation, the ability to guarantee income for life, and liquidity.

Then I construct a composite hypothetical VA/GLWB, based on the characteristics of those offered by 5 major companies. I look at maximum allowed equity allocations of both 60% and 100%, a guaranteed roll-up rate in the deferral period of 5.3% compounded, a 1.29% mortality and expense fee on the account value, a 1.35% rider fee on the benefit base (which can end up being a much higher percent of the remaining assets when the account value is less), a 4.8% guaranteed income withdrawal rate for a 65 year old, and annual fees of $39.

This compares to the investment-only VA with annual fees of $240 and a 1% annual advisory and investment fee applied to the account value.

With Monte Carlo simulations, I then investigate the amount of lifetime guaranteed income supported by the VA/GLWB, and then try to replicate the same withdrawals from the investment-only VA. 

Let me provide one example based from the article. Consider a 10 year accumulation period followed by a 30 year distribution period. There is a 58% chance that the VA/GLWB contract value will have hit zero (though the guaranteed income would still be provided). There is a 12.1% chance that the investment-only VA would be depleted while replicating the same payments. In the median outcome, the VA/GLWB would be depleted. In the median for the investment-only VA, the VA would have supported all of the income provided by the VA/GLWB and still have a real value of $159,709, relative to $100,000 at the start. Real wealth actually grew by 60%.

Ultimately there is no single answer about what is right, as it depends on a person's preferences, but this analysis helps to make clear about the tradeoffs involved. With the VA/GLWB, some guaranteed income will always be provided, though in real terms it may end up being quite low. The effect of the compounding fees is dramatic. It is much less likely that there will be any liquidity, because fees eat away at the contract value.  In this example, there was a 12.1% chance of running out of assets with the investment-only approach. But in the median outcome, the investment-only approach could have matched income from the VA/GLWB and still experience real asset growth of almost 60%. That's the true cost of the guarantee. And making these costs more clear is the point of the white paper.