Tuesday, October 29, 2013

Safe Withdrawal Rates: Is 60 the New 95?

Safe Withdrawal Rates (SWR) strategies, spawned by the so-called Trinity Study1 in 1998, are based on a dangerous assumption — that the future will look like the past.

Now, most everyone knows that is a ludicrous assumption. Nonetheless, in finance we sometimes pretend that is it not, and in our closets we Baby Boomers hang on to our plaid bellbottoms, just in case.

The widely-publicized interpretation of those studies is that retirees can withdraw 4½% of their nest egg the first year of retirement and continue to withdraw that constant dollar amount, increased for inflation, every year and have a 95% probability of funding at least thirty years of retirement. (That's the widely-publicized interpretation, not mine.2)

More and more, researchers are raising serious doubts about that "history repeats itself" assumption. Dr. Wade Pfau studied withdrawal rates in other countries and found that the U.S. was an exception. The widely-advertised 4% to 4½% withdrawal rates weren't safe globally. In subsequent studies, Dr. Pfau concluded that more frugal withdrawals would probably be needed for today’s retirees to achieve a 95% chance of their portfolio lasting 30 years or more in this country. Maybe closer to 3%.

A recent study by Pfeiffer, et. al., entitled The Benefits of a Cash Reserve Strategy in Retirement Distribution Planning, investigated the benefits of maintaining a one-year cash reserve account, but I found the results of the Reverse Dollar Cost Averaging (RDCA) scenarios most interesting. RDCA is another name for constant-dollar withdrawals, or Safe Withdrawal Rates (SWR).

The study (excellent video explanation by Dr. Pfau here) compared monte carlo simulations of RDCA and a similar strategy that held one year’s expenses in a cash account to minimize having to sell stocks when prices were low. A major difference between the Pfeiffer study and the Trinity and William Bengen studies earlier is that Pfeiffer considers transaction costs and taxes and “a future of lower investment returns than seen in the historical data” based on recent capital market projections (that's current predictions of future stock market returns for those of you who speak English).

The results? In a tax-deferred environment (401(k), IRA, etc.), with a 4% withdrawal rate the cash reserve strategy improved 30-year portfolio survivability by nearly 5% from . . . wait for it . . . 55% to 59.4%. 

Not 95% to 99%.

55% to just under 60%.

In a taxable environment, cash reserves increased 30-year survivability with 4% withdrawals, but from a very risky 60% to a still risky 66%.

The following chart is a summary from the Pfeiffer study. The first pair of columns represents an unrealistic environment with no taxes or transaction costs. The second pair represents a ROTH IRA scenario with only transaction costs. The third pair represents a tax-deferred (IRA, etc.) environment and the fourth pair a taxable environment.

95% isn't even on the y-axis.

What happened to 95% survival rates and 4½% withdrawals? If Pfau and Pfeiffer are correct, history won’t repeat itselPf. . . sorry, itself . . . and 4½% won’t work. If history does repeat itself and the U.S. stock market continues to outperform the rest of the world, Bengen was correct and 4½% will work.

Who’s right?

It’s unknown. And unknowable. Check back with me in thirty years. But, the newer research nowadays seems to point to significantly lower sustainable withdrawal rates.

Maybe they’ll all be wrong. In his early writing at EfficientFrontier.com, William Bernstein wrote that retirees shouldn't expect any rate of success greater than 80%.

That "everyone being wrong" thing happens a lot with financial projections. A quite successful money manager used to predict the coming year’s market returns by waiting for other money managers' predictions and then picking the range that no one else picked. It worked amazingly well until the others caught on and they all started waiting.

If you’re basing your retirement funding primarily on portfolio withdrawal strategies, you should heed the words of the authors of the Trinity study that started it all:

"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."

95% survivability of your investment portfolio with 4½% annual withdrawals isn’t carved in stone somewhere.

In fact, 60 may be the new 95.

1Retirement Savings: Choosing a Withdrawal Rate That is Sustainable

2I think the studies show that your probability of success is a function of the value of your portfolio at any point in time (not its initial value when you retire) and the number of years remaining in retirement. In other words, you have to reassess the safe withdrawal amount periodically, and the safe withdrawal rate increases as retirement progresses.


  1. I don't recall ever seeing someone using 4.5% as a Safe Withdrawal Rate. The Trinity Study doesn't seem to even include that as an option - they only tested whole percentages (3%, 4%, 5%, etc). So I would argue that your interpretation of the "widely-publicized interpretation" is at best, not so widely-publicized.

    On the flip side, one oft-ignored facet of the Trinity Study is that they only studied time periods up to 30 years. Even if 4% is a SWR, it's a SWR if you don't mind having no money left after 30 years. If you're planning on retiring longer than that, it's not even purported to be safe.

  2. AND a 5% probability (or 40%?) that it won't even last 30 years.

    By the way, references to 4.5% aren't that hard to find. Here's one from Forbes last June, quoting Bill Bengen, no less. I'd provide more if it changed the argument. http://www.forbes.com/sites/nextavenue/2013/06/10/how-much-to-withdraw-from-retirement-savings/

    Thanks for reading and commenting!

  3. I think what is being overlooked is the current environment with 0% cash, sub 3% 10-yr T-bonds, and sub 3% dividend yields on stocks. I don't believe this combination has ever occurred in US history until the past 3 years. Bengen and others did excellent work but the Fed has deliberately orchestrated an investing environment where major asset classes have very low yields.

    Why would anybody expect 4% to provide a safe 30-yr+ portfolio success rate in this environment when not a single major asset class yields over 3%? At this moment, I would expect a safe withdrawal rate of 3%-3.5% for a classic 60/40 portfolio.

    1. The problem with the work of Bengen et. al. is precisely that it assumes that we've already seen everything the market has to offer in its brief history. I think you just provided another example that we haven't.

      The Fed holding down rates artificially does, of course, hurt yields, but it's been a huge boon to equities.

      As for expecting 4% to be safe, I've never expected withdrawing from a volatile portfolio to be safe.

  4. Enjoyed the read and absolutely agree that there are a tremendous number of unknowns and unknowables. With regard to the 4% (or 4.5% rule) I do believe it is as good a starting point as any when first developing a retirement plan. I believe the key is for people to understand the concept of a SWR, develop one, and apply it to their situation. Perhaps they will develop an initial plan for a 4% rate and modify down (3.5%, 3%) over time as they draw closer to retirement and have a clearer picture of their unique retirement portfolio, their assumptions about the number of years the money has to last, the investment environment, etc.

    1. You make a couple of excellent points, James. 4% to 4.5% is a good starting point because a TIPs ladder would historically have paid out about 4.5% for 30 years and 4.5% is in the ballpark for long term inflation-protected payout rates for lifetime annuities, though they currently are around 3.5%.

      The other important point you make is that if you spend down a stock portfolio, you're going to need to adjust your withdrawal rate periodically. Fixed dollar withdrawals are extremely risky.

      Thanks for reading and commenting!