What We Do

Research on Safe Withdrawal Rates

Moderator: hocus2004

Post Reply
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

What We Do

Post by JWR1945 » Fri Oct 03, 2003 4:19 am

1. hocus refers to the Safe Withdrawal Rate (SWR) issue as a phenomenon. He is right. Reactions to SWR research are often emotional and out of proportion to what they should be. Studies are attacked not on the basis of the research itself, but on the conclusions that come out of the research.
2. The general notion of a Safe Withdrawal Rate can be found in this post: http://nofeeboards.com/boards/viewtopic ... 017#p11017
A more precise definition suitable for serious investigations is found in this thread: http://nofeeboards.com/boards/viewtopic.php?t=1107
3. In its most basic form a Safe Withdrawal Rate does nothing more than tell you how much you need to save in order to fund your retirement successfully. It tells you the minimum amount that you will need in order to have a sufficient degree of safety. There are a variety of withdrawal strategies to meet a variety of goals. Each has its own answer. The definition allows for every one of these.
4. What the precise definition requires is that every different case be identified clearly and as a separate problem and in a form suitable for mathematical calculation. As such, there is always a single, correct answer. At least, notionally. It is based exclusively upon information available up to the time associated with the calculation and none thereafter. As with any mathematical problem involving probability, statistics and randomness, the quality of an answer is not judged in terms of a single, actual outcome of events.
5. The definition does allow for an answer that is simply a boundary. The Trinity Study and the Retire Early Safe Withdrawal Rate Study attempted to do that. Those studies determined the results of hypothetical portfolios tested against historical sequences of investment returns. For every year in the historical record, they determined the highest withdrawal rates for which a portfolio would have survived a specified number of years, but not any longer. We refer to those results as Historical Database Rates. They took the lowest of those rates, which is the maximum lower bound of the successful Historical Database Rates, and used it to estimate the Safe Withdrawal Rate.
6. We now know that those answers are wrong. They left out at least two critical factors, both of which invalidate the results. The Historical Database Rates were based upon investments that no longer exist. Today's stock market (in terms of the S&P 500 index) is at valuations well above the historical range and it produces dividends well below the historical range. We have identified cause and effect relationships that show why the previous studies failed.
7. Keep in mind that information associated with Historical Database Rates is still useful and it is still valuable. It is just that they are not the same as Safe Withdrawal Rates. Suitable manipulations and adjustments can be made to approximate Safe Withdrawal Rates.
8. We generally prefer to use a restrictive, but much more useful version of Safe Withdrawal Rates on this board. We are not satisfied with a bound. We prefer to calculate the Safe Withdrawal Rate of each individual year. For example, the Safe Withdrawal Rate for 2003 is not necessarily the same as for 2000 or 2004. It has not always been known that doing this was possible. It had been known that we could calculate rates in terms of longer periods (e.g., decades).
9. It is important to differentiate between a Safe Withdrawal Rate, which is objective since it is a calculation, and its application, which is subjective and requires judgment. Every Safe Withdrawal Rate is an answer to a different mathematical problem. It is calculated with different degrees of accuracy. Applications are seldom identical to those involved in the calculations themselves. An individual should extract the most useful information that applies to himself and apply it carefully to his own situation. There are many factors that are necessarily external to the mathematical calculation.
10. The research on this board has included identifying cause and effect relationships and monitoring portfolio safety. It is being conducted in a manner that allows answers to be extended beyond the traditional S&P 500 and commercial paper (and other cash equivalents) portfolios that have been studied in the past. We expect to see more and more sensitivity studies. We have been able to restore the safety of portfolios with withdrawal rates that had been calculated incorrectly in the earlier studies. We expect to see the usefulness of Safe Withdrawal Rate calculations expand as we examine applications more thoroughly.

Have fun.

John R.

Post Reply