Thoughts on the Future of Retirement Income Products

This fly-fisherman, financial advisor and former insurance company executive surveys the retirement income product landscape from his lake house in central Maine.

In the past few years, we have seen the emergence of many new retirement income products. With the exception of guaranteed minimum withdrawal riders on variable annuities, however, we have seen very little in the way of sales of any of the products. In this article, I’ll discuss pros and cons of the various products from my perspective as a financial planner, assess the prospects for these products, and recommend what I believe would be an optimal solution for many people. 

The Immediate Annuity

This product has been the “product of the future” for more than a decade, but the future never seems to arrive. Sales limp along at an annual rate of $10 billion or so, compared to $250 billion for deferred annuities. One rationale for the low sales is consumer aversion to large, irreversible financial commitments, particularly where shorter-than-expected longevity would result in a retrospectively bad investment. Sales may also be low because financial salespeople are loath to sell a financial product that kills the prospect of any future rollover commissions. Given the lack of sales push or customer pull, insurance companies don’t devote a lot of resources to developing or promoting such products. 

However, despite all the negatives, the immediate annuity is a simple product that can match up well with retirement income needs, particularly if payments increase with inflation. Prospects for this product may improve if 401(k) plans are enhanced to make it easy to convert savings into retirement income. It could become attractive for fee-based planners, but few fee-based planners serve the low- and middle-income clients who could best use the mortality-pooling benefits of an immediate annuity to stretch scarce retirement savings. Also, most financial planners have focused on accumulation and not retirement income products.

The following changes might allow immediate annuities (and other retirement products I’ll discuss later) to play a more prominent role in retirement planning:

  • Requiring that retirement savings plans offer an immediate annuity option.
  • Providing tax subsidies for financial planning to spur the development of a fee-based financial planning industry serving participants in retirement plans.
  • Providing planners with the training and software they will need to incorporate immediate annuities and other guaranteed products in retirement plans.

Longevity Insurance

This product also goes by the longer name of Advance-Life Delayed Annuity (ALDA).  It’s an income annuity where the first payment is delayed for a number of years. For example, a 65-year-old man could purchase longevity insurance that pays an income for life that begins at age 85. If the individual dies before reaching age 85, no payout occurs. The individual would need to plan on making retirement savings last to age 85 and then rely on the longevity insurance for income thereafter. The individual would be freed from the planning challenge of trying to make retirement assets last for an unknown future lifetime.

Because of the income deferral, this product costs less than an immediate annuity. An immediate annuity paying an inflation-adjusted $20,000 per year might cost about $290,000, while longevity insurance paying the same benefit would cost only about $40,000.

This product first came on the market a couple of years ago and I’m aware of two companies that offer it. Anecdotal evidence suggests that sales have been miniscule. As with the immediate annuity, the retirement marketplace doesn’t seem to be set up to pay much attention to this type of product. Also, the financial crisis and the problems faced by a formerly AAA company like AIG have made individuals nervous about buying a product that delays income for 20 or more years.

The product also has a fundamental design problem. Going back to our example, let’s say the individual buys the product and then needs to make other savings last until age 85. If individuals make overly aggressive assumptions about investment growth, they may run out of assets before age 85 and then face a period without income. If they are too conservative, they may reach age 85 with more assets than they need, and may regret not spending more money when they could have enjoyed it.

The fix for this problem would be to combine longevity insurance with a product that guarantees withdrawals until age 85 regardless of investment performance. This would be an acronymic marriage of a limited-term GMWB or RCLA (discussed next) with the ALDA.  I believe that longevity insurance has potential as a component in building retirement guarantees, but I don’t foresee strong prospects for it as a stand-alone product.

The Guaranteed Minimum Withdrawal Benefit (GMWB)

This product is offered as a variable annuity rider and has gone through a number of transformations over the decade or so of its existence. The latest version, called the Guaranteed Lifetime Withdrawal Benefit (GLWB), guarantees the annuity purchaser certain minimum withdrawals from the annuity for life (5% of the initial purchase amount is typical), regardless of investment performance. The purchaser pays .50% to 1.00% per year for this rider.

I like this concept for its guarantees, but I have these concerns:

Complexity. This is an “actuaries-gone-wild” product where instead of offering a simple income guarantee based on an initial purchase amount, the product’s guarantees vary with a ratcheting provision based on underlying fund performance. As a financial planner, I would rather show the client something simpler-“Pay X, and y ou’re guaranteed Y-end of story.”

Lack of Inflation Protection. At 2.5% inflation, a 5% fixed guarantee is worth only about 3% in 20 years, and 3% isn’t much of a withdrawal guarantee.  I would like to see a product whose guaranteed payouts increase based on actual inflation. Products offered today provide some upside based on underlying fund performance, but the correlation with actual inflation is tenuous.

Cost. A typical variable annuity costs about 2.25% per year. If we add .50% to 1.00% for the rider, we’re near 3.00%. Recent estimates for the premium of future stock returns over bond returns range from 3% to 6%. If that’s the case, a client who uses a variable annuity with a GLWB rider might end up paying out between 50% and 100% of the equity premium in fees.

Rider Utilization. At least one study has shown that, to make best use of the GWLB rider, the customer should take withdrawals at the maximum allowed level immediately after purchase and continue them for life—that is, turn the product into an immediate annuity with an equity-linked refund feature.  My impression is that most customers will use the product for accumulation and never call on the income features at all—thereby turning the product into a very expensive mutual fund investment.

In some ways, the GLWB is ideal for guaranteeing retirement income. But its effectiveness in meeting customer needs is seriously compromised. Given the attractive commissions offered on variable annuities, I expect sales to remain strong. But I think there is a better way forward for the customer, and that takes the form of the next product.

Ruin Contingent Life Annuity (RCLA)

This product is the brainchild of Professor Moshe Milevsky of York University in Toronto. He has noted that it is basically an unbundled version of the GLWB. The product was described in detail in Retirement Income Journal so I’ll provide a simplified example here.

Begin with a block of assets that will be used to provide inflation-adjusted withdrawals at some set percentage of the initial amount. The customer purchases a guarantee that if the funds are depleted-due to longevity, poor investment returns, or an adverse sequence of investment returns-the guarantee will kick in and inflation-adjusted payment will continue for life.  Because payouts are a function of multiple contingencies-longevity and investment related-the product will cost less than a longevity insurance policy that pays out based on longevity only.

Before illustrating how the RCLA works, I’ll show results for other product solutions. We’ll start with a 65-year-old man with $325,000 of savings and the desire to withdraw an inflation-adjusted $20,000 per year for life. One solution would be simply to invest the money, take systematic withdrawals, and hope the money lasts a lifetime. Assuming 2% inflation, 4% bond returns, 8% stock returns, and a 50/50 stock bond mix, the portfolio would last until age 91. The risks with this approach are:  (1) living beyond age 91, (2) experiencing returns lower than assumed, or (3) experiencing a bad sequence of returns, i.e. low returns in the early years of retirement.

A very different solution would be to guarantee a lifetime income by purchasing an immediate annuity. At current rates, an inflation-adjusted annuity generating an initial income of $20,000 might cost about $290,000. That would meet the need for guaranteed lifetime income, but would lock up all but $35,000 of the available funds and leave disappointed heirs if the retiree were to die early.

What about longevity insurance? An ALDA purchased at age 65 and providing $20,000 per year (in today’s dollars) beginning at age 85 would cost about $40,000. That would leave $285,000 to provide income for 20 years. Based on the inflation and investment assumptions used in the systematic withdrawal example above, we would expect the $285,000 to last 22 years-not much margin to guard against bad investment experience.  Based on some Monte Carlo analysis, there’s a 44% chance the assets would not last the needed 20 years.

Now let’s look at the RCLA. An RCLA able to produce the required income might cost roughly $20,000, leaving $305,000 as the portfolio on which guaranteed withdrawals would be taken. In effect, we have created the same lifetime guarantee as the immediate annuity, but with $305,000 of initial liquidity for heirs vs. $35,000 if we purchased an immediate annuity. Among legacy-minded retirees, this may lessen the resistance to the initial purchase. (Of course, the $35,000 would be expected to grow over time and the $305,000 to decline, making the immediate annuity potentially the better investment if the retiree lives a long life.) I have not provided an example based on a variable annuity with a GLWB because I am not aware of any that provide a simple inflation guarantee. But with their high expense charges, GLWBs would not produce results as attractive as with the RCLA.

Compared with the RCLA, a disadvantage of annuities and longevity insurance is that the pricing reflects the insurance companies’ cost of holding reserves for such products in fixed income investments. Even though these represent long-term commitments, insurers are not allowed to support them with a mix of stock and bond investments. GLWB and RCLA pricing allows for equity investing and equity option pricing, which means a lower price for consumers.


Both immediate annuities and the RCLA can help clients build secure retirement plans. But neither of them, as I suggested earlier, is likely to get off the drawing board and into retiree portfolios without major changes, such as mandating that retirement savings plans offer the product, subsidizing the growth of the fee-based planner industry, and providing software and training that supports the use of this product. Existing product delivery systems won’t suffice.

© 2009 RIJ Publishing. All rights reserved.