The 4% Retirement Withdrawal Rule: International Data Casts Doubt

There is a rule of thumb that says a person can withdraw 4% of the value (as of the date of retirement) of an investment portfolio and adjust the amount every year for inflation without fear of running out of money. A prime proponent of this idea is William Bengen (review of his book here) who based his conclusions only on US data.

Not all countries are created equal, so a key question is how people elsewhere might have fared using the 4% rule. Along has come Wade Pfau and his provocatively titled paper An International Perspective on Safe Withdrawal Rates: The Demise of the 4% Rule in the February 2011 issue of the Journal of Financial Planning. (Author Pfau has his own blog here)

The results are not very encouraging: "... from an international perspective, the 4 percent real withdrawal rule has simply not been safe." Using data over a longer period from a different source than Bengen (but which validates Bengen's results about the US), Pfau calculated what would have happened in 17 developed countries using quite generous assumptions and found that in only three other countries - Canada, Sweden and Denmark - the 4% rule would have worked out.

The paper's main aim is really only to get a feel across different countries. It deliberately excludes several extremely important factors that could dramatically alter the specific maximum Sustainable Withdrawal Rate in realistic conditions:

• mainly downwards - no adjustment for rebalancing trading and fund management costs to the index data, no taxation on returns, using the stock vs bond allocation that gave the best result, as if the retiree could perfectly forecast the best combo for the next 30 years;
  
• or upwards - possible other asset classes like REITs, foreign stocks, small cap stocks, value stocks; longer rebalancing intervals than yearly
  

The study confirms another Bengen US finding, though exact figures vary country by country: stocks always comprise a substantial part - at least 50%, some up to 100% - of the portfolio that had the best SWR; for Canadians that was about 50%.

A scary bit to the paper is table 4 (which appears in the blog article version but not in the FPA version). Table 4 shows that for an arbitrary commonly cited 50% equity/50% bond portfolio (as opposed to the stock bond split that gave the best result), in no country would a retiree never have run out of money using the 4% withdrawal rate.

The key question highlighted in the conclusion applies to every country. The past is all well and good but for any particular country, which past will most resemble the future? Will the US or Canada continue to outperform compartheed to everywhere else
or will they revert to some lower international mean? One can simply blindly lower the SWR e.g. to 3.5% or 3%, or try to adjust expected future stock and bond returns using current relative valuation.

The even larger issue for the retiree is whether an investment portfolio with systematic withdrawals is the most appropriate way to fund retirement expenses. I don't believe so. I've already noted objections to the 4% SWR approach by Scott, Sharpe and Watson. The critical correction needs to be that assets should be matched with liabilities in terms of income risk (variability, default) and timing of income.