For returns, they’re still not worth shooting.
But there is an inkling of hope.

    The subject is variable annuities. Yes, it’s time for an update of my annual Variable Annuity Watch. This year the purchase of a simple broad index fund has again beaten a large majority of variable annuity sub-accounts.
The index product did this before taxes. It did it after taxes. It did it for domestic equities, international equities and taxable bonds. Here are the results for the 10-year period ending May 31:

Domestic equities. 

Over the last ten years the average return on the 545 surviving large-blend domestic equity sub-accounts with 10-year performance records was 6.14 percent a year. That’s a full 1.56 percent a year less than the 7.70 percent return of the Vanguard 500 Index fund.
Only 52 variable annuity sub-accounts in the category provided higher returns, so the odds against selecting a winning sub-account were 10-to-1.
 Not very favorable.
This comparison doesn’t count tax liabilities. Since major index funds such as the Vanguard 500 Index and its expense-competitive peers are tax-efficient, accumulating most gains, an investor in the 25 percent tax bracket would have ended the 10-year period with a $20,160 accumulation on an original investment of $10,000. He would have a cost basis of $12,382 on net reinvested dividends and capital gains distributions.
That means $7,777 in unrealized capital gains and a tax liability, at 15 percent, of $1,167. A decision to cash out would have netted $18,993. That computes to an after-tax return of 6.63 percent annually. To provide the same result in a variable annuity, assuming a 25 percent tax rate, would require a pre-tax return of 8.84 percent.
Only 20 of the 545 surviving sub-accounts provided 8.84 percent or better, so the odds that you or your sales representative would pick one of those funds were 25-to-1 against you.
 Not good.

Taxable Bonds.
Over the same period, the Vanguard Intermediate Bond index fund returned 6.45 percent excluding taxes. The average variable annuity intermediate term bond sub account returned 4.80 percent. Only 4 of the 297 surviving funds in the category provided a higher return---so the odds against you are approaching100-to-1.
Taxes, of course, loom larger here. The after-tax return of the index fund would be only 4.95 percent. With a final market value of $16,209 after taxes and a cost basis of $17,907, the index bond fund investor actually has a tax loss to declare and no tax liabilities on the $16,209. To duplicate the after-tax net, the variable annuity sub-account would have needed to have a pre-tax return of 6.60 percent. Only 4 of the sub-accounts did that, with returns ranging from 6.61 percent to 6.73 percent. The odds are dismal.
International Equities.
If you had invested in the Vanguard Total International stock index, your pre-tax return would have been 8.31 percent compared to the 6.49 percent return of the average of the 277 surviving variable annuity sub-accounts. Only 26 of the 277 provided a return higher than 8.31 percent so, again, the odds of selecting a winning VA fund were 10-to-1 against you.
The original $10,000 index fund investment has grown to $21,145 in value, including $8,802 in unrealized capital gains taxable at 15 percent. So the after-tax net value of the investment is “only” $19,824, or an annualized return of 7.08 percent. To equal that return, the variable annuity sub-account would have needed to provide a return of 9.44 percent before taxes for an investor who is in the 25 percent tax bracket. Only 15 of the 277 surviving sub-accounts in the category provided a return that high, so the odds against doing better in an international variable annuity sub-account were 20-to-1.
 Since we’re talking about funds that survived the period--- and not counting the funds that disappeared during those ten years--- these figures understate how difficult it is for variable annuity sub-accounts to beat simple indexing.
It’s pretty much the same old story. So why am I holding out a ray of hope?
Because variable annuity contracts have changed.

As recently as 5 years ago the vast majority of the variable annuity contracts offered a “death benefit.” In the event of your death, your beneficiary would receive an amount equal to the original investment or the accumulated value, whichever was greater. You had to die to get the benefit, which made it unattractive to most people.
Today, more than 85 percent of all variable annuity contracts are sold with a “living benefit” feature that guarantees some form of withdrawal. When they were first introduced on a handful of contracts, I examined this feature. I found its benefits illusory.
But things change.

Competition has forced improvement in living benefits. In the current issue of the Journal of Financial Service Professionals, Moshe Milevsky, an associate professor of finance at York University in Toronto, notes that living benefits (such as the Guaranteed Minimum Withdrawal Benefit, or GMWB) work, in effect, to reduce investment volatility. As a consequence, a living benefit may increase the odds that you will not outlive your money.

With millions of workers approaching retirement without a lifetime income, living benefit contracts may evolve into a useful retirement tool--- “assuming the insurance fees charged for this protection are not too high,” professor Milevsky notes in one study. In another article, he worries that “the living benefit arms race is leading to promises and guarantees that might become difficult to keep.”

Bottom line: While traditional “death benefit” variable annuity contracts continue to be poor choices, some “living benefit” contracts may provide a level of financial security not available in a direct investment.

The hard part--- perhaps impossible part--- is knowing which contract offers that security.  
On the web:

Moshe A. Milevsky (website)

Portfolio Survival

Variable Annuity Watch, 2006:

Liz Puliam Weston: The Worst Retirement Investment You Can Make