Fund companies designing life-cycle or target-date 401(k) funds are using overly optimistic assumptions such as how much participants contribute over time and how many of them make withdrawals, according to a study released today by JPMorgan Asset Management.

The study also examines the returns associated with the different asset-allocation approaches and suggests that using alternative investments in target-date funds can help retirees improve their 401(k) returns.

Mutual fund companies that envision defined-contribution plan participants who start by deferring 6 percent of their salaries, then boost that to 10 percent by age 35 are being unrealistic, according to the report.

Based on data on the 1.3 million participants in 401(k) plans that the company administers, JPMorgan researchers said that while workers start by saving 6 percent, they don’t defer 8 percent until they’re 40 and don’t hit 10 percent until 55.

Industry models assume participants get annual salary increases, but JPMorgan’s data showed they receive raises in only two out of every three years. The report also pointed to other factors that it said are not accounted for in industry models: 20 percent of 401(k) participants borrow from their plan, and 15 percent make withdrawals once they reach the age of 59½, when tax penalties for early withdrawals cease.

As an employee’s flow of savings into the 401(k) plan rises and falls, that can interact with ups and downs in the financial markets to create bigger variations in the returns employees realize from target-date funds, according to JPMorgan.

“If you’re pulling out 10 percent of your balance and [the market’s] down 20 percent, that hurts,” says Anne Lester, a senior portfolio manager at JPMorgan and co-author of the report.

According to JPMorgan, the best way to minimize the volatility in employees’ returns and improve their chances of accumulating sufficient retirement savings is to diversify by adding alternative investments. By using assets like real estate, high-yield bonds and emerging market securities, “you get less volatility for comparable or slightly higher levels of return,” Lester says. “To us, the goal is to see whether we can approach comparable levels of excess return with less volatility.”

Real estate investments have performed poorly during the recent stock market sell-off, but Lester says that over the long run, she expects “real estate, both direct and REITs, will continue to provide an attractive diversification to both stocks and bonds. Direct real estate, in particular, has shown a tremendous ability to provide diversification during periods of market stress for fixed income and equities.”

JPMorgan compared the performance of four different approaches to target-date funds: aggressive, in which 65-year-olds have a 59 percent allocation to equities; concentrated, with a 50 percent equity allocation at age 65; conservative, with a 17 percent equity allocation at 65; and JPMorgan’s SmartRetirement design, in which 65-year-olds have a 30 percent equity allocation plus a 22 percent allocation to alternative assets, including high-yield bonds, real estate, and emerging market debt and equities. The other three have only minor allocations to alternative assets.

Simulations using JPMorgan’s data about participants show the aggressive design offers the most upside, with the top quartile of participants accumulating more than $800,000. But it also produces a wider range of results at the low end of the scale: With the aggressive design, participants at the 95th percentile saved just $228,000, just above the $224,000 that represents the 95th percentile in the concentrated design and below the 95th percentile results of $236,000 for the conservative design and $260,000 for SmartRetirement.

The median amount of savings that employees accumulated under both the aggressive and SmartRetirement designs was around $550,000, while the median for the concentrated design totaled $512,000 and for the conservative design totaled $462,000.

JPMorgan also assessed the likelihood that participants would be able to accumulate $400,000, the amount it calculated would buy an annuity to replace 40 percent of a pre-retirement income of $65,000. Its modeling indicated that 76 percent of plan participants would reach that goal with the SmartRetirement design, 72 percent would achieve it with the aggressive design, 69 percent with the concentrated design and 65 percent with the conservative design.

Filed by Susan Kelly of Financial Week, a sister publication of Workforce Management. To comment, e-mail editors@workforce.com.