Readers of this blog will note that I devote a
fair amount of energy ranting against the 4% Withdrawal Rule (and
other "Safe" withdrawal rates) as retirement decumulation
strategies. In addition, I'm generally not all that impressed with
proposed modifications to the 4% Rule designed to
somehow make it more workable. Vanguard recently announced its
proposed modifications in a paper entitled, "A More Dynamic Approach toSpending for Investors in Retirement." They suggest a two-step process
for determining an annual spendable amount payable from accumulated
savings: Step 1: Take X% of end-of-the-previous-year
accumulated savings. Step 2: Subject the result of Step 1 to a
corridor, the ceiling of which is (1+Y%) of the spendable amount from the
previous year and the floor of which is (1-Z%) of the spendable amount from the
previous year, where "X" depends on the "planning horizon"
and investment philosophy and "Y" and "Z" are
arbitrarily chosen upper and lower limits (they suggest a value of 5 for Y and
2.5 for Z). Readers of the paper who get as far as Appendix 3 will note
that the example set forth in this appendix describes a
slightly different approach than the approach described in the body of
the paper, which I am assuming is an error).
While the Vanguard modification of the 4% Rule does
make the approach more dynamic (i.e., it reflects actual investment
experience to some degree), I believe this approach to be inferior to the actuarial approach
suggested in this website for the following reasons:
As is the case for most "safe" withdrawal rate
strategies, it defines success as not outliving accumulated assets. It
does not adequately address the risk of under spending.
It doesn't attempt to provide constant real dollar spendable
income in retirement.
It doesn't coordinate with other forms of retirement income
such as immediate or deferred annuities and it doesn't reflect bequest
motives.
With all the adjustments required for
different planning horizons and investment philosophies, it is not
appreciably simpler than the actuarial approach set forth in this website (particularly
if you use the assumptions and algorithm I recommended several posts ago).
