Defined: The quantity of money, expressed as a percentage of the initial investment, which can be withdrawn per year for a given quantity of time, including adjustments for inflation, and not lead to portfolio failure; failure being defined as a 95% probability of depletion to zero at any time within the specified period.
Usage: Typically, SWR is utilized as an approximation of the probability that a given portfolio can support a given annual spending component for a required period, with a reasonable confidence. To do this, variables such as the allocation of assets within a model portfolio, the beginning balance, and/or the number of years expected in retirement are varied, a model is applied, and results of these alterations in the variables are observed and compared, in order to optimize for the maximum.
Controversy: Unfortunately, the term "Safe Withdrawal Rate" is necessarily an ambiguous term.
This is because initial methods utilized historical data to statically determine what would have been safe given the actual results that past portfolios would have generated with the variables given. The next logical step, of course, was to use that information to predict
future SWRs. Either use is technically correct, but one should always be sure to be clear whether the use is in reference to past or projected SWRs, so that unnecessary argument can be prevented.
One scenario backtested in the Trinity study suggests that a retiree with a suitably allocated $1 million portfolio could withdraw $40,000 the first year, give herself a cost-of-living adjustment every year afterwards, and have a 98% chance of the portfolio lasting at least 30 years.
Taken literally, such a plan has been criticized as unrealistic. Even if the tests showed that the plan had a 98% success rate over all past time periods, would a prudent person blindly go on steadily increasing withdrawals in a prolonged bear market? It also leads to apparent absurdities. Say that retirees A and B have saved $1 million in 2008, and the market crash reduces their portfolios to $800,000 in 2009. A, however, retires in 2008 while B waits until 2009. The Trinity study bases withdrawals the dollar value of the portfolio at the start of retirement. The value fluctuates with the vagaries of the stock market. Thus, even though their situations are almost identical, in the Trinity scenario, retiree A, by virtue of having retired in 2008, is allowed to withdraw $40,000 plus COLA in 2009; while retiree B, despite being in an almost identical situation, would be allowed only $32,000.
The authors of the paper, however, did not mean for their scenarios to be applied rigidly or uncritically. The article makes this very important statement:
The word planning is emphasized because of the great uncertainties in the stock and bond markets. Mid-course corrections likely will be required, with the actual dollar amounts withdrawn adjusted downward or upward relative to the plan. The investor needs to keep in mind that selection of a withdrawal rate is not a matter of contract but rather a matter of planning.
Nisiprius requested clarification from Professor Philip L. Cooley, senior author of the Trinity study:
What the "4% SWR" means is not that you can treat a portfolio as if it were a guaranteed annuity. I think all the [Trinity] authors meant is that if it is late 2008 and your stocks halve in value, you don't need to halve your spending instantly. It's OK to cross your fingers and continue spending according to the 4%-then-COLAed plan, even though it means dipping into capital, and it's OK to go on doing that for a while.
Professor Cooley's response:
You have hit the nail on the head! I've tried to explain that thought to journalists but they don't seem to get it. You've got it. Stay flexible my friend!, which is the advice we should give to retirees.
