“Starting to save at age 25, rather than age 45, cuts the required saving rate by about two-thirds,” and,

“Delaying retirement from age 62 to age 70 also reduces the required savings rate by about two-thirds.”

The study illustrates the dramatic impact of saving early and retiring later. It shows an employee needs to save only 7 percent of his income if he starts saving at age 25 and retires at age 70. On the other hand, an employee who begins saving at age 45 and retires at age 62 must save 65 percent of his salary—a pretty unrealistic goal. Delaying retirement helps a little. If this 45-year old saver retires at age 70, he would “only” have to save 18 percent of his salary—still a lot, but at least it's within the realm of reason.

This delayed retirement can significantly help those concerned about saving enough to retire. One of the paper's authors told me more than 70 percent of all households could maintain their pre-retirement standard of living if they waited until age 70 to retire.  

But the truly astounding revelation is the relative unimportance of asset allocation. The researchers compared the real-life impact between investing in the “typical” asset allocation and investing in the “optimal” asset allocation. What they found would startle even the most committed product salesman: Any advantage offered by the “optimal” asset allocation can easily be erased merely by working four months longer.

This, in the end, might have the greatest bearing for 401(k) investors. The importance of investment performance has long been touted and fees have been understood as being critical to investment performance. For a study to suggest that asset allocation—and its subsequent influence on investment performance—is only of secondary concern to retirement success suggests only one thing: 401(k) plan sponsors and investors should pay more attention to the levers they can control and less attention to their investments (within reason, of course).