by Rob Bennett
I gained my fame on the internet by being the person to discover the errors in the Old School safe withdrawal rate studies. The studies (uncorrected to this day) say that a 4 percent withdrawal is always “100 percent safe.” Studies done using an analytically valid methodology (that is, one adjusting the safe withdrawal rate for the valuation level that applies on the day the retirement begins) show that the safe withdrawal rate can drop to as low as 2 percent at times of high valuations and can rise to as high as 9 percent at times of low valuations.
Stocks were insanely overpriced at the time (this was 2002). So they were also of course insanely overpopular. I took a lot of heat for saying that there were circumstances in which a retirement plan calling for a 4 percent withdrawal (one that permits those with a portfolio value of $1 million to take out $40,000 for living expenses each year) is not safe. The historical stock-return data shows that, when stock prices are where they were at the top of the bubble, retirement plans calling for a 4 percent withdrawal have only a 30 percent chance of surviving 30 years. That’s not safe. It’s not a close call.
Today, I am seeing more and more people come around to a more prudent point of view on stock investing. I recently wrote a guest blog that mentioned the safe withdrawal rate issue and the owner of the site commented that he would not advise retirees to go with high stock allocations (the safe withdrawal rate studies generally assume a stock allocation of 80 percent). I do not approve. This is not the answer.
At least I do not approve entirely. The 80 percent stock allocation that is generally assumed in retirement studies is too high. That assumption was a product of the stock mania that fueled the bull market and it should be changed to something more reasonable (I offer a calculator at my site that permits the user to calculate the safe withdrawal rate that applies for any stock allocation). Still, I see great danger in the idea of reacting to the millions of failed retirements we are going to see as a consequence of the errors made in the retirement studies by concluding that stocks are more dangerous than we thought they were at the time these studies were developed.
The post-Shiller research does not show that stocks are more risky than we thought they were. It shows that stocks are less risky than we thought they were. We need to avoid the temptation to give in to fearful reactions when we see bull market irresponsibility ruin millions of lives. We help the people whose lives were ruined by the dangerous studies not by becoming more fearful but by demanding more accuracy in our retirement studies.
The safe withdrawal rate wasn’t anything close to what the “experts” were saying it was back in the days when I was being hit over the head with bricks for reporting the number accurately. But valuations have fallen since then. The safe withdrawal rate is close to 4 percent today. My calculator shows that the withdrawal rate that has a 95 percent chance of surviving 30 years for a retirement beginning at today’s valuation level is 3.75 percent. If you were willing to go with a retirement that is only reasonably safe (an 80 percent chance of working out) and you were retiring today, you could go with a withdrawal rate of 4.35 percent.
It’s obviously true that a portfolio going with a low stock allocation is generally safer than a portfolio going with a high stock allocation. So, yes, we can solve the problem of our having caused millions of failed retirements by encouraging retirees to go with low stock allocations. That solution doesn’t come free of cost, however. Investors who go with low stock allocations when that is not a good idea lower their returns by doing so. Telling people to go with unnecessarily low stock allocations delays their retirements.
The purpose of retirement research should not be to delay retirements any more than it should be to cause failed retirements. The idea should be to report the numbers accurately. Reporting the numbers accurately permits aspiring retirees to retire as soon as is possible consistent with their desire to enjoy safe retirements.
We are likely going to see valuation levels of one-half of fair value sometime within the next five years or so. At a P/E10 level of 8, the safe withdrawal rate for an 80 percent stock allocation is 9.13 percent. That permits the retiree with a $1 million portfolio to pull out $91,300 each year to cover his living expenses. Lowering that retiree’s stock allocation from 80 percent to 40 percent percent would lower his safe withdrawal rate to 6.62, covering only $66,200 of annual living expenses. Following the cautious advice of my blogger friend would cost this retiree $25,000 of spending in every year of his retirement. For what purpose?
We are hurting retirees in two ways by failing to correct the Old School studies and by failing to promote the New School studies that would take their place once we acknowledged the errors that brought us both Buy-and-Hold and the economic crisis. We are encouraging some of them to enter retirements all but certain to fail and encouraging others of them to live dramatically diminished lives for no good reason whatsoever.
The safe withdrawal rate is sometimes a low number and sometimes a high number. We need to stop twisting ourselves into logic pretzels trying to justify the unfortunate convention of pretending that it is always the same number. We shouldn’t overstate the safe withdrawal rate but we shouldn’t understate it either. We should aim to report this important number accurately and honestly and realistically.
Those performing investment research should begin making a serious effort to get the numbers right.
Rob Bennett describes the practical implications of the investing theories of Robert Shiller. His bio is here.