Using SWR Analysis to Compare Asset Classes: Edited

Here are some of Rob Bennett’s thoughts about using SWR analysis to compare asset classes. He has done this for years. I have extracted this from a series of recent emails. This is edited.

Keep in mind that all asset classes have more than a single dimension.

Say that it is January 2000, and you identify the SWR for TIPS as being 5.85 percent and the SWR for 80 percent S&P stocks as being 1.6 percent. It does NOT follow that you should be 100 percent in TIPS.

SWR analysis has often been employed in one-dimensional investing strategizing exercises. This is unfortunate. It has tarnished people’s assessments of the tool. Many make their investment choices PRIOR TO doing the SWR analysis. They use the results of the analysis to reassure themselves.

When you use the tool in this way, and you see that the number generated supports your earlier decision, the tendency is to use the number generated as the be-all and end-all. The 4 percent number for stocks almost always beats the numbers generated for alternate asset classes, so those using the tool to back up their earlier choices have gotten into the practice of using the SWR number as a one-stop decision-making tool. If the stock SWR number is higher, then investing in stocks is “optimal.”

The conclusion does not follow from the findings.

It is entirely possible for a lower-SWR asset class to be the better investment choice.

Say that the SWR for stocks is 3.5 and the SWR for TIPS is 4.0. Does that mean that TIPS are “better” than stocks? It does not. Stocks offer more potential for long-term growth. And the 3.5 number is a worst-case-scenario number for stocks while 4.0 is a certainty for TIPS. A good case can be made for going heavily with stocks. You do not hurt yourself too much in the SWR area by doing so and you help yourself a good bit in the long-term growth potential area.

Personal circumstances need to be taken into account.

There are some who cannot afford even the little bit of risk involved in moving from 4.0 percent SWR TIPS to 3.5 percent SWR stocks. But few should be going with ALL stocks or ALL TIPS in this scenario. You need some TIPS (or some alternative with TIPS-like characteristics) to counter the volatility of the stocks. Another factor is how close the investor is to retirement. Speaking generally, however, TIPS shine in the SWR area and stocks shine in the potential long-term growth area, and if you can get stocks at the same SWR as TIPS, you are going to want to go largely (but never entirely!) into stocks.

The key to analyzing this at a deeper level is always to keep separate in your mind the different things being analyzed.

A conventional SWR analysis is one component of an analysis to determine which asset class is best. It is not the entire analysis. A complete analysis requires conventional SWR analysis PLUS other things. What people miss is that a good number of these other things can also be examined through the use of SWR-type analyses.

A good example is whether the investor will hold an asset class for the long-term or not.

Buy-and-hold investing has become popular in recent decades. There are wonderful gains to be had from buy-and-hold investing. In a theoretical sense this is a good development. In reality, however, the current-day understanding of buy-and-hold investing is very much misguided. It is fantasy-world stuff. You don’t become a buy-and-hold investor by saying the words “buy-and-hold.” It is not a magical incantation. To obtain the benefits of buy-and-hold, you must invest in such a way as to realize the benefits. An SWR-like analysis can help you do this.

How? We have an historical record that we can use to obtain clues to the critical question—How likely is it that I will actually hold my stocks for the long term? The historical data are incomplete. That’s unfortunate. But we are not entirely without clues.

It slants things terribly in favor of stocks to just ASSUME that all investors will maintain their current-day allocations for 30 years. This is a silly la-la land assumption. It is necessary to make some assumptions. But of all the assumptions that one could make, why is it that many have elected an assumption that all investors will be true buy-and-hold investors, no matter how unrealistic their decision-making processes are at the time of their purchases?

Everyone knows that such assumptions are fallacious. But there is a great reluctance to doing anything about the problem. I believe that this will change when stock prices drop. Then people will be using assumptions biased AGAINST stocks. The goal should be to determine what sorts of returns stocks will provide to real live investors in the real live world. When you know that, you are on your way to making a good comparison between stocks and alternative asset classes.

Let’s say that you determine that the SWR for stocks is 3.5 and the SWR for TIPS is 4.0.

Let’s say that you further determine that in your circumstances long-term growth is far more important than safety. So, based on your SWR analysis, you decide on an 80 percent allocation to stocks. Have you done your job? NO!

You need to think a bit about what sorts of assumptions the 3.5 number is based on. Does it presume that you will be a 100 percent buy-and-hold investor? Is this assumption realistic for investors going with 80 percent stock allocations at the valuation levels that prevail at the time this decision is being made? If not, you need to make adjustments.

In order to make a good comparison of asset classes, you need realistic assumptions. Comparisons need to be FAIR. Stocks really are better than TIPS in some important respects. Stocks really do offer more potential for high long-term growth. But obtaining the high long-term growth in the real world requires sticking to a buy-and-hold strategy. You need to look to the historical data to determine how wild a ride you are planning for yourself and how likely it is that your buy-and-hold strategy will survive the ride being elected.

The lower your stock allocation, the less the volatility of stocks hurts you and the more likely it is that you will stick with a buy-and-hold strategy. These things can be expressed with numbers, just as the valuation insights were expressed in numbers. I don’t say that they can be expressed perfectly with numbers. The data-set is not good enough to permit that. But I believe that valuable insights can be generated through use of the historical data on this point as with many others.

Another factor that is often overlooked is the growth potential of a safe asset class like TIPS when the intent is to transfer those funds to stocks when the SWR for stocks becomes attractive.

When we say that the SWR for TIPS is 4 percent, we are making an assumption that the investor will hold the TIPS for 30 years. What if that is not his intent? What if his intent is to move the money to stocks as soon as the SWR for stocks rises to a pre-specified level? In that case, the SWR for the money initially invested in TIPS is not 4 percent, but something higher. [Or at the least there is a chance that it is higher—it could be that (worst-case) the number for stocks would never rise enough and the investor would be stuck with TIPS at their original SWR.]

Again, it is possible to look to the historical data for guidance on this question. Again, assumptions are required to do the analysis. (You need to assume that stock prices will drop in x number of years). But again, there are ALWAYS assumptions being made in any type of SWR analysis. The question is never, “is there an assumption being made?” The question is, “is the assumption being made a reasonable one?” Reasonable assumptions are good. Unreasonable ones are bad. The assumption being made in most analyses of TIPS is that they will be held for 30 years REGARDLESS OF WHAT HAPPENS TO STOCK PRICES. This is unreasonable.

Why would an investor who was smart enough to limit his stock ownership at times of low SWRs for stocks not also be smart enough to increase his stock allocation when the SWR for stocks goes higher? He would, of course, make the switch. He would obtain better returns over 30 years from the money originally allocated to TIPS than a conventional SWR analysis would show. Again, the reason is that the conventional analysis was rooted in an assumption [i.e., that the investor would stay in TIPS for 30 years] that is not supported by the historical data. The historical data show that, whenever stocks are at highly unappealing prices, there is a strong likelihood of better prices in the not-too-distant future.

There are all sorts of things that can be done by making references to the historical data.

Think about SWR analysis as a probability assessment tool. It does not answer questions directly. It supplies guidance. It provides the probabilities of various outcomes. This is no small thing. All investing decisions are rooted in assessments of the probabilities of various outcomes. They MUST be. What else is there to go on in making a decision for one asset class over another?

The unfortunate thing is that most assessments of probabilities are really just guesses. People are saying, “Stocks seem generally to do kind-of good, so I am going to go with all stocks and maybe that will work out.” The better way is to look to the historical data and see how likely it is that things will work out. The answer that comes back might be that there is only a 20 percent chance that things will work out as you hope. In those circumstances, you should be coming up with a different plan.

The job of an SWR analyst is to assess the probabilities of various outcomes, presuming that stocks perform in the future somewhat in the way in which they have in the past. It’s always possible that the 20 percent-probability outcome will come through. So the SWR analyst can never say with absolute confidence, “this is the best thing to do.” What he can say is that, IF stocks perform in the future as they have in the past, the odds of such-and-such happening are good or the odds of such-and-such happening are bad. It’s a big step up in the development of investing strategies because it is rooted in something objective—the historical stock-return data. Our goal is to move people away from basing their investing strategies on subjective impressions, which often cause them to do just the wrong thing at just the wrong time. (Reference: Chapter Two of “The Four Pillars of Investing” by William Bernstein).

Rob

Rob Bennett is the author of Passion Saving: The Path to Plentiful Free Time and Soul-Satisfying Work.
His web site is: Rob Bennett's web site

Edited by John Walter Russell
July 24, 2005