From the December 2009 issue of Research Magazine • Subscribe!

Annuity Analytics: Truth in Retirement-Product Labeling

It's time for insurance companies and other financial-product manufacturers to start reporting the nutritional content of their creations in a limited and standardized format.

Most readers of this column know that the annuity is one of the most misunderstood -- and often abused -- terms in the lexicon of retirement income planning. Financial services companies currently offer variable annuities, fixed annuities, deferred annuities, income annuities, immediate annuities, longevity annuities and equity-indexed annuities, to name just a few; each comes with its own distinct and difficult-to-compare features.

This is extremely confusing to the end-user (i.e., your client) who has little or no background in actuarial science or advanced derivatives valuation. Even financial advisors themselves have a hard time distinguishing between the steak and the sizzle in the newer financially engineered products with guaranteed living income benefits (GLIB).

To make things worse, many of the retirement products I come across, christened and marketed as annuities, have little to do with the antique Roman concept of periodic income. Some offer no mortality credits whatsoever. (Remember, mortality credits capitalize on the concept of risk sharing and dictate that survivors within the pool must inherit some assets from the deceased.)

The no-free-lunch dictum applies here as well. Once you allow retirement income products to have ample liquidity, enhanced death benefits and on-demand refunds, you could be left with an expensive bond or fixed-income portfolio, perhaps with a thin veneer of life insurance. In my view that's not an annuity, regardless of what it says on the packaging.

In other words, if an "annuity" company promises to pay you whether you live or you die, that's a bond. The only way to pay you more if you live -- as annuities are supposed to do -- is to pay you less if you die. For anyone "worried" about living, it is worth giving up the death benefits.

Just as an example, a so-called income (a.k.a. immediate) life annuity sold to a 70-year-old male but with a 25-year period-certain guarantee -- which might be a safe and appealing investment from a psychological perspective -- contains very few mortality credits. It is effectively a collection of staggered and stripped bonds. Likewise, a variable annuity with a nominal 5 percent guaranteed lifetime withdrawal benefit (GLWB) for a single life, and a conservative asset mix, provides only a negligible amount of longevity insurance or protection against a market crash and negative early sequence of returns.

And so, to bring structure to this rapidly evolving market, I think it is time for insurance companies and other product manufacturers to start reporting the nutritional content of their creations in a limited and standardized format. This is akin to the labels on your favorite cereal, yogurt or granola bars which disclose the amount of vitamins, minerals and nutrients. The financial labels should disclose the amount of longevity insurance, mortality credits, stock market crash protection, and other factors contained within any and all annuity products.

I call this process Retirement Income Product (RIP) DNA Extraction.


In this context, though, instead of breaking open cells, removing membranes and dousing proteins in ethanol, the laboratory of choice is a fast computer that can simulate thousands of potential market scenarios. The extraction algorithm counts the percentage of cases in which various product features pay off and the cases in which they expire worthless, all discounted to account for the present value of money.

So, for example, if the annuity product doesn't offer any lifetime income, then its discounted payoff will be quite low for those scenarios in which retirees exceed their life expectancy. Likewise, if the product offers no inflation protection, its discounted payoff will be low in high-inflation scenarios. In the language of statistics the properly scaled "regression coefficients" would be the weights which should add up to one. The total is the sum of its parts. All of this is presented for a hypothetical case in the nearby table.

For example, such a DNA extraction algorithm confirms that a low cost S&P 500 index fund (or ETF) contains 100 percent basic market exposure with a 0 percent factor loading in all other six categories listed in the table. But, in contrast, a generic (low cost) 5 percent-for-life variable annuity with a Guaranteed Lifetime Withdrawal Benefit (GLWB) sold to a 65-year-old with a 60/40 equity/bond asset allocation consists of 90 percent basic market exposure and only 10 percent longevity insurance. And, if you increase the allowable equity exposure by 10 percentage points the longevity insurance component grows to 15 percent.

These numbers are all hypothetical, of course, and the actual percentages would depend on the case in question as well as who is actually buying the product.

However, regardless of the specifics, I believe that such an analysis -- done either by the manufacturers themselves, the distributors or ideally independent third parties -- can be used by both advisors and their clients to shed light on the underlying ingredients contained within any proposed retirement income solution. I am advocating for a standardized and limited number of comparison factors (as opposed to features) to help consumers grapple with what they are really buying.

To see the value of such a standardized system, think of the ubiquitous mutual fund style boxes popularized by Morningstar. Despite the thousands of different and unique mutual funds and money managers competing for your investment dollars, they all must classify themselves into a three-by-three matrix.

Most managers have to advertise and promote their primary holdings and strategy as either being value or growth, small-cap or large-cap, domestic or international, etc. In fact, some of these managers might actually despise being boxed in to such a grid, and would rather be given a mandate to invest in anything that catches their fancy on any given day, but the rigid classifications makes life much easier for comparison and benchmarking purposes.

A small-cap growth manager's performance is measured against a suitable index to see if they truly added value to their clients. They can't advertise themselves as offering exposure to emerging markets (read: protection against negative sequence-of-returns risk) when in fact they primarily hold large-cap European equity.

This knowledge also helps advisors construct properly diversified portfolios that cover all style factors. I think the same kind of framework would make good sense in the retirement income world.

So, to help spur this trend towards nutritional disclosure, in future 2010 columns I will change gears somewhat and start examining specific annuity and retirement income products to explicitly uncover their specific DNA. Yes, I'll be naming names. Let's see if there is any granola in these cereal bars. Suggestions are welcome. Stay tuned and have a healthy and great holiday season.

--

Moshe A. Milevsky, PhD, is an associate professor of finance at York University in Toronto, and Executive Director of the non-profit IFID Centre at the Fields Institute for Research in Mathematical Sciences. His next book, Your Money Milestones: A Guide To Making The 9 Most Important Financial Decisions In Your Life, will be published by FT/Pearson and will be available in early 2010.

Reprints Discuss this story
This is where the comments go.