The Dangerously Miasmic Myth Of A 4% Safe Withdrawal Rate

by: Tim Richards


One - rare - area where academic economic research intersects with the interests of private investors is on the topic of Safe Withdrawal Rates - the maximum amount you can safely withdraw from your investment pot each year following retirement. Opinion is divided on exactly how much is safe, but the general consensus is somewhere in the region of 4%.

Of course this is utter baloney. As usual, when academic analysis meets commonsense understanding it creates a perfect miasma of miscomprehension and, no doubt, bad retirement planning. Your safe withdrawal rate is dependent on you, not on some mythical, half-baked rule of thumb.


In Greek mythology a miasma is a contagious infection with an independent life of its own. In finance we see these all the time in the form of memes, ideas that infect and pollute the investing mindset of many an erstwhile trader (see Memes, Money, Madness and Diworsification is Good For You for other examples). And the only way to deal with the pernicious little blighters is to track them back to their lairs and ruthlessly exterminate them. Many a lurking meme fails when exposed to the cold light of rationality.

So, the original research behind the Safe Withdrawal Rate came from William Bengen in 1994, in a paper entitled Determining Withdrawal Rates Using Historical Data. Bengen's paper is full of reasonable assumptions, but the main ones are that the portfolio asset allocation is 50/50 US equities and US treasuries and the Safe Withdrawal Rate is one in which your portfolio lasts for 30 years before possibly running out of money. And the magic number turns out to be 4%: you can withdraw 4% of your starting value each year, up-rated for inflation, regardless of the underlying performance of your portfolio.

One of the critical things here is that the 4% is based on the starting value of the portfolio - so if you're unlucky enough to retire on the cusp of one of the great stock market collapses then you very rapidly end up withdrawing a huge percentage of that starting value each year, and thus diminishing your pot of capital very quickly. Unsurprisingly Bergen's research shows that, based on historical market returns, as you increase the rate of withdrawal then you increase the likelihood of running out of funds. And this, my friends, is where the recommended Safe Withdrawal Rate of 4% comes from.

Bond Busting

Only, as Bengen points out in his original paper, the actual evidence suggests that you'd be stupid to go for a blended 50:50 portfolio of stocks and bonds:

"I think it is appropriate to advise the client to accept a stock allocation as close to 75 percent as possible."

Yet, despite the data clearly supporting a more aggressive allocation towards equities, you can almost sense the unease he feels recommending this:

"An asset allocation as high as 75 percent in stocks during retirement seems to fly in the face of conventional wisdom, at least the wisdom I have heard. But the charts do not lie, they tell their story very plainly".

Now define "conventional wisdom" ...

The Cult of the Equity

As we saw in the discussion of psychological investing trends across history that is The Zeitgeist Investor, the historical consensus was that bonds were safe for retirement and equities weren't, because of the relative risks of individual stocks and bonds. It took George Ross-Goobey to change all of that, as he recognized that, in the long-run, portfolios of stocks could offer better and safer returns than government bonds:

"Studies going back 80 years and including several depressions show that Common Stocks have increased in value at a rate which offsets the long-term rate of inflation and on top of this have shown a real yield in terms of purchasing power of about 4% or 5%".

That's from the UK in 1956, just as the so-called Cult of the Equity was beginning to swing into action. And just as negative real interest rates were beginning to cut into real bond returns - as we saw in How Sneaky Governments Steal Your Money.

In fact, in his landmark paper Bengen looks at various other historical scenarios including what an investor should have done following the Wall Street Crash if they'd gone into it with a 50:50 bond/equity portfolio. Someone retiring in 1929 with a portfolio of $500,000 and withdrawing at 4% would have been down to less than $200,000 by the end of 1932. If they'd run for the safety of a bond portfolio at this point they have run out of money in 1947. On the other hand ...

"But what if our client had the audacity to demand, on December 31, 1932, that we increase the stock allocation to 100 percent, and hold that allocation for the remainder of his life? ... by 1992, if he were still alive, he would have amassed $42 million in his retirement fund!"

Mind you, they'd have probably been past caring as well, if they'd retired at the expected age of 60. Although, as we saw in earlier article Dying to Invest, that's a dangerous assumption to make, as the unfortunate Andre-Francoise Raffray discovered.


Of course, this is all historical revisionism, being run against the only history we happen to have. If we were to rerun the tape we'd get a different result: whatever happens in the 21st century it won't be the same as what happened in the 20th (see: Investing in the Rear View Mirror). History is told from the viewpoint of the victors - remember U.S. stock market returns aren't generally applicable across the world - an investor in German markets back in the 1920s would have been wiped out along the way by hyperinflation and the Second World War (see Triumph of the Pessimists or Stuck in a Weimar World).

In fact, some recent analysis by Wade Pfau on international diversification makes exactly this point - U.S. safe withdrawal rates don't necessarily apply to the rest of the world:

"While global diversification did improve outcomes, there is still a 22.7% failure rate for the 4% rule with global portfolio returns measured in terms of the local currency. Clients can expect better outcomes with international diversification, but even with this greater diversification, there is still a chance that 4% will not work and adjustments to spending patterns or asset allocations will be needed in retirement."


Which is quite grim, really: a 22.7% failure rate on a 4% starting withdrawal rate isn't exactly comforting. But, once again, the target portfolio is 50:50 bonds to equities. When Philip Cooley, Card Hubbard and Daniel Walz looked at the historical success of a 100% stock portfolio in Retirement Savings: Choosing a Withdrawal Rate That Is Sustainable, they found:

"A 100% stock portfolio supported 100% of all 15-year periods in which annual withdrawals were made based on an initial withdrawal of 3% of portfolio value. The portfolio success rate drops to 98% for a 5% initial withdrawal rate, reflecting the failure of the all-stock portfolio during one of 56 15-year periods (1929 to 1943). Not surprisingly, as the withdrawal rate rises, the portfolio success rate declines."

So if you can avoid the one period encompassing the Wall Street Crash and the start of the Second World War you can invest in 100% equities and safely withdraw at 5% ... or maybe not. The point is that the 4% rule is a mental short-cut with a dubious provenance and it's now ricocheting about between advisers, investors, bloggers and sundry other ancillary experts as though it's a magic bullet. It's not..

No Magic Bullets

As we've seen before the sheer power of memes can override any attempt at rational analysis because people just love simple rules which they can apply to complex subjects to avoid thinking about them. In an area as complex as investment, which is populated by lots of behaviorally challenged investors who are convinced they know what they're doing and are heavily resistant to being told that they don't, this is dangerous stuff.

There is no such thing as a Safe Withdrawal Rate and believing otherwise is dangerous. What there is are investors who know what they're doing and those who believe in mythical rules. On balance I'd recommend being one of the former, but there's no accounting for taste.