We all know that The Retirement Risk Evaluator shows that the Old School studies get the safe withdrawal rate wrong. The number isn’t always 4 percent. At the top of the bubble, it was less than 2 percent. Today, it’s 3.6 percent.
That’s only one side of the story. When the historical stock-return data is examined using an analytically valid methodology, it tells us something else of equal importance. When stock valuations are low, the Old School studies are wildly off the mark not on the high side but on the low side. When the P/E10 level drops to 8 (this is what usually happens in the aftermath of out-of-control bull markets), the SWR rises to — 9 percent?
A lot were shocked to learn that the number can drop to 2 percent in an out-of-control bull. Can it really rise to 9 percent in the aftermath?
That’s what the historical data says. That’s the number you get when you use an analytically valid methodology. That’s the rate that applies at that valuation level if you are willing to assume that stocks will perform in the future at least somewhat as they always have in the past.
But would I advise a retiree actually to withdrawal 9 percent from his or her portfolio? Is that not a crazy-with-risk idea?
I don’t think so. There are caveats that apply for those using the 9 percent number. But that is the number you get when you do the analysis and, when you come to an informed understanding of how stocks work in the real world, you come to see that that number actually makes a good bit of sense.
The fair-value P/E10 value is 14. When the P/E10 value is 8, stocks are selling for a price not much above half of their fair value. Given the fantastic 10-year return that stocks provide when selling at fair value, would you not expect a super-fantastic 10-year return when they are selling at only half of fair value?
The Reversion to the Mean phenomenon exerts a powerful pull on stock prices, bringing them in the direction of fair value. When the price drops all the way down to 8, that pull is very powerful indeed. For the P/E10 value to move up even a notch, stocks need to pay a return in excess of the normal 6.5 percent real. For the P/E10 value to move up enough notches to take us from 8 to 14, stocks need to pay a return in excess of the normal 6.5 percent real for a good number of years in succession.
One of our most important findings is that it is just about always the first 10 years of a retirement that determine whether it will succeed in the long term or not. If you do well in the first 10 years, you will see enough of an increase in wealth to get you through 30 years, especially after the compounding returns effect is counted in. If you do poorly in the first 10 years and continue taking out a withdrawal that seemed to make sense back when you had much more in the way of assets, you stand a good chance of sinking your plan somewhere down the line.
Leverage. That’s the hidden factor. Do well early on and leverage works for you. Do poorly early on, and leverage kills you. The retiree very much wants to get leverage on his side.
When stocks are at extremely low prices, the odds of leverage working against you are almost nil. The conventional wisdom when we are at a P/E10 of 8 will be that stocks are for losers. But think about it from the standpoint of leverage. What are the odds that the P/E10 level is going to drop still lower when it is already at one of the lowest levels ever seen in U.S. history? Stock prices might not go quickly up when we are at 8. But they are almost certainly not going to go down much. If prices merely remain stable, you will be earning an annual return of about 6.5 percent real. That translates into a safe withdrawal rate well above 6.5 percent (the withdrawal rate is higher than the return earned because SWR studies assume that the portfolio will be gradually depleted over the course of a 30-year retirement).
There’s a good chance, of course, that returns are going to go up from very low levels. The odds are that the investor handing in his resignation at a time of low prices is going to enjoy extraordinary leverage. His first 10 years are going to be good years. Even if a worst-case return pattern happens to pop up, a plan calling for a 9 percent withdrawal will work.
I noted above that there are caveats. The most important caveat is that the entire U.S. economy could go kerplooey. Most people laugh at mention of that possibility today. Not too many will be laughing when we get to a P/E10 of 8. The middle-class will have suffered the largest loss of wealth in U.S. history at that point. Lots of people will be predicting that our economy will soon go under and lots of others will be saying that that sounds right to them. Neither the Old School not the New School SWR studies factor in that possibility. If the whole shebang goes down, your retirement could go down too.
How big a risk is that? Bill Bernstein has put it at about 20 percent. It happens to every economy sooner or later, and so sooner or later it is going to happen to ours. So it’s not entirely right to say that a 9 percent withdrawal is 95 percent safe at a time when the P/E10 level is 8. If you count in the possibility that the entire economy could go down, the odds of success are less than 95 percent.
That’s so for those taking a 3.6 percent withdrawal on a retirement beginning today too, though. There’s always a chance that the economy is going to go bust, both at times when stock prices are high and when they are low. The fact that people will be talking about it when we get to a P/E10 of 8 doesn’t make it more likely that it will happen in the course of a 30-year retirement, it only makes you more aware of the possibility.
Remember, a retirement that lasts 30 years is going to see times of high valuations and low valuations. Just about all 30-year retirements will see times when everyone is saying that national bankruptcy is just around the corner and times when everyone is saying that anyone who sees national bankruptcy as even a remote possibility belongs in the loony bin. What people are saying about this topic on the day you retire does not affect the safety of your 30-year plan (although it of course affects your feelings about the safety of your plan — which is something very different).
The bottom line is that neither the 3.6 percent number that applies today nor the 9 percent number that applies for retirements beginning at a P/E10 of 8 are perfectly safe. But they are equally safe. The 3.6 percent number is the number that comes up if you look at what happens in a worst-case scenario for someone retiring today and the 9 percent number is the number that comes up if you look at what happens in a worst-case scenario for someone retiring when the P/E10 value is 8.
Does it make sense that the SWR could vary so much? Yes, it makes sense.
The stock investors of the 1990s borrowed massive amounts from future returns to finance the sickest and most twisted bull ever experienced in U.S. history. The investors of today are the investors whose returns were being borrowed. We’re like the people who get stuck covering someone else’s credit-card bill. The need to pay off the huge debt holds us back. We cannot be sure of earning returns large enough both to cover that debt and to support a retirement plan calling for a withdrawal of more than 3.6 percent.
When we get to a P/E10 value of 8, the debt will have been fully paid off and then some. When prices drop that low, investors are leaving a windfall inheritance to those who come after them rather than a huge credit-card bill. It’s not fair (that’s one reason why I don’t view wild price swings as a good thing), but that’s the way it goes.
Retire today, and you are taking on at least some risk if you go with a withdrawal of greater than 3.6 percent. Retire when the bull-market debt is paid off and when prices have dropped low enough to provide an inheritance windfall for stock investors, and you can take a withdrawal of 9 percent with the same level of safety that applies for a 3.6 withdrawal for those retiring today.
Today’s Passion: There’s a thread at the Early Retirement Forum showing that more people are losing confidence in the phony baloney Old School SWR studies all the time. Please note the poll results — there are a lot more people at that forum who understand why ignoring valuations cannot work than they are who are willing to say so for so long as there are Goons present in the room. I found the same thing to be so at the Financial WebRing Forum. There was a poll in which a large percentage of the board community expressed a desire that honest posting on Techncal Analysis be permitted (I do not personally believe in Technical Analysis but I certainly believe that honest posting on it should be permitted) despite the fact that there were few willing to say so in posts put to the board. When the ownership of a board makes clear that it will not tolerate honest posting on any particular topic, the board becomes a corrupt enterprise. You can never know what a board community really thinks about a topic for so long as a ban on honest posting on that topic remains in effect.