Our current plan design models—defined benefit (DB) and
defined contribution (DC) plans—are inadequate and unnecessarily restrict
retirement plan design.
Traditional defined benefit (TDB) plans very nicely provide
an adequate retirement income for participants, especially those who work their
entire career for one employer/plan sponsor. But they create a fundamental
problem; because they provide a fixed benefit, not payable until retirement and
without regard to asset performance, TDBs create a fixed corporate liability.
Generally, investors insist that this liability must be reduced to a present
value, using current interest rates.
During the 1980s and 1990s, when interest rates were relatively
high, funding TDBs was somewhat inexpensive. With the rates’ steady decline
over the last 12 years—with a 100 basis-point decline in 2011 alone—for many
companies, the “cost” of TDB plans has become intolerable.
This problem is solved, generally, by freezing or
terminating the TDB plan and replacing it with a 401(k) plan, effectively
taking retirement savings liabilities, at least with respect to future
accruals, off the employer’s books.
However, 401(k) plans create their own problems. To produce
an “adequate” retirement benefit, these plans depend on participants
contributing early in their careers. Despite 20-plus years of advertising and
matching contribution programs—and, more recently, default contributions
(automatic enrollment) programs—a sizable percentage of participants remain
left behind or have seen their savings “leaked out.” Because 401(k) plans
provide for participant asset-allocation decisions (recently mitigated by the
use of target-date funds) and a retail mutual fund investment model, with
accompanying retail fees, 401(k) plan investments underperform DB investments.
(The data here is problematic, but the rate of underperformance appears to be
at least 100 basis points.) And, because they pay lump sums, they create a
longevity risk problem that has, thus far, proved impossible to solve.