POST OF THE DAY
Unsafe Withdrawal Rates?

 Format for printing Reuse/Reprint

August 27, 2001

Posts selected for this feature rarely stand alone. They are usually a part of an ongoing thread, and are out of context when presented here. The material should be read in that light. How are these posts selected? Click here to find out and nominate a post yourself!

The purpose of this post is to question some the conventional wisdom on this board concerning "safe" withdrawal rates in retirement. I have several concerns regarding the interpretation of the studies often referenced on this board. I should say, however, I think that the research is very useful and I have expanded upon it myself using additional asset classes as part of my own retirement strategy.

First, I am skeptical of the entire concept of optimizing withdrawal rates based on historical asset returns. Wm. Bernstein pretty much sums up my feelings with this snippet from the Morningstar Conversations series (#1049):

"While the study of historical returns is useful, particularly when it comes to determining the risk of an asset, it only gets you so far. It's a mistake to rely too heavily on even 200 years of returns. (We actually have over 300 years of returns for British stocks.) I lean much more heavily on the discounted dividend model than I do on historical returns.

Simple example. For the 50-year period ending December 1983 the return of the long treasury was 3.48%. On the other hand, the discounted dividend model would have predicted a return of 11.7% (the bond coupon at the time). Which do you suppose provided the more accurate prediction? Similarly, the 74-year return of stocks ending December 1999 was 11.35%. With a dividend yield of 1.2% and the most optimistic estimates of long-term earnings growth at 6%, the only way you get returns higher than 7.2% (the sum of the two) is with further multiple expansion. Lotsa luck
."
Folks, if you take nothing else from this post please seriously consider the above simple but elegant statement. To amplify, the long term (i.e. decades) return of an index can be calculated thusly (from Bernstein's book):

Return = Dividend yield + dividend growth rate + PE multiple change

These are the only three factors that will account for future stock returns. Currently, the div. rate is barely 1%. Since 1920, the dividend growth rate has nominally been about 5% (approx. 2% in real terms). Thus, without PE multiple expansion or a historically unprecedented leap in long-term economic growth one could optimistically expect a real return of about 3% (maybe 6% nominal) from the S&P500. This is only if you believe that the current 1929-like valuations are sustainable on a long-term basis. If, like me, you believe that the current market PE of 29.5 is likely to eventually return to historic norms of about 15 then look out below!

The next flaw in the safe withdrawal scenario I'd like to point out is related to the nature of the S&P500 index itself. It sounds great to assume that one would've gotten the nearly 7% real returns touted since 1871 for the S&P500 but is this possible prospectively? I don't think so. The index was constructed only relatively recently and cheap indexing vehicles became possible for the small investor only in the last couple of decades. It is my feeling that returns would've been substantially lower if such popular indexing vehicles had been available in 1871. Why do I think that? Because it has become a well-known fact that as soon as a stock looks like it might be about to be included in the S&P500 index it rapidly jumps up in price several percent or more because of the inevitable massive purchases of the stock by the huge index funds that will be obligated to buy it. This clearly didn't happen during the first hundred years or so of the study. I believe that this is good evidence that returns for the index would have been lower if indexers had been forced to buy fully or overvalued stocks like they are doing now. BTW, it's not just the new S&P500 stocks that get bid up, it's all 500 of them. Indexing has become so popular that the gargantuan index funds compete for the stocks and bid them up appropriately. To me this S&P500 indexing fad looks suspiciously like the Dogs of the Dow/Foolish Four did before being exposed with regard to the datamining flaws that have recently come to light.

Okay, even if you disregard the above and stick to the historically-optimized 4% withdrawal rate touted for a roughly 75/25 mix of S&P500 stocks/bonds + cash then you must rigidly stick to that allocation. This leaves no room for individual stocks, trading strategies, mechanical investing, etc., otherwise the concept is invalid. In addition, you must religiously rebalance every year, which can really hurt if most of your assets, like mine, are in taxable accounts. I see a lot of talk on this board regarding the 4% withdrawal rate from people whom apparently are not sticking to the rigid premise of the study. This is asking for trouble if you insist on using "The Rule" blindly.

Finally, let me further rain on the parade by reminding you that, thanks to the slowly increasing effect of populist-leftist politics in this country, your taxes are only going to go up. I think that early retirees will be a particularly juicy target for the class-warfare attack being perfected by Clinton/Gore/Gephart-type politicians. You can bet that at the very least these guys are drooling over the trillions of dollars in retirement accounts and are losing sleep try to scheme ways to get at it. I suspect that they will eventually figure out a way to tax at least some of it. So, I would count on an increasing tax burden, which will take more and more of your 4% yearly draw or whatever amount you plan to withdraw each year. Social Security? I wouldn't plan on seeing much of it come your way if you can afford to live without it.

So all is lost? I don't think so but I would sure consider adding some other asset classes to your equity allocation just in case the S&P500 indexing fad really is a bubble in the process of bursting. If there is any interest I'd be glad to offer my ideas on retirement portfolio construction in a future post but my fingers are tired from typing right now. <g>