This site documents some analysis that I did initially for my own purposes in the area of asset allocation and the impact of various portfolio withdrawal methodologies. I am publishing this analysis to provide assistance to people planning for retirement (or possibly already in retirement). It was intended to be a direct extension of the work done by Jaye Jarrett/Tom Stringfellow and Professors Cooley, Hubbard, and Walz at Trinity University (although it was done without their knowledge or endorsement). I want to state clearly that I do not have the financial credentials of any of these folks. See About the Author for irrelevant personal background information. Also this is not a commercial site. The ads you see are simply the cost of 15MB of free web space. My personal ISP provider does provide enough free space to hold the all the graphs associated with this study.
The question that I needed to answer was "am I saving enough to fund a comfortable retirement". The issue that I ran into was that I did not have any idea what kind of inflation-protected income stream that I could plan to generate from a given asset base (I still remember the market returns and inflation of the 1970/80's). So even though I was not necessarily ready to retire, I found myself needing to understand the answer to this question in some detail in order to evaluate the adequacy of my retirement savings plan.
There are several high quality (in other words, better than this one) sources of information that have investigated this question and have reported their results in very readable forms. Some are listed below:
- Trinity Study (summary by Scott Burns)
I will be proceeding under the assumption that the reader is familiar with at least one of these studies. Also Frank Armstrong has written an excellent paper on this topic from a broader perspective (for his home website Click Here) that is worth your time. John Greaney at his Retire Early HomePage has a huge set of resources that are quite useful including a Motley Fool sponsored message board and a downloadable EXCEL spreadsheet that will let you do your own withdrawal simulations similar to the Trinity and Jarrett studies (note that Greaney uses the Schiller data which has a smaller set of asset classes but covers a longer time period). Pause for research by the reader :-)
The Methodology chapter describes the details of the analysis.
The Fixed Annuity chapter is intended to address issue #1.
The Market Risk chapter is intended to address issue #2.
The Combining it All chapter addresses the case of risk when a fixed annuity is a significant component of your retirement picture.
The Other Withdrawal Strategies is a brief analysis of more/less aggressive withdrawal targets.
The Efficient Frontiers chapter is an attempt to map the concept of the "Efficient Frontier" for the accumulation phase to the Efficient Frontier in the Withdrawal Phase.
The Simple and Efficient chapter is a more detailed assessment of a withdrawal strategy that was found to be relatively efficient in the Efficient Frontiers chapter.
The Alternative/Next Steps chapter is a very brief discussion of an interesting alternative asset class and thoughts on further analysis.
The Omega Strategy (not) chapter discusses a modification to another withdrawal methodology suggested by Scott Burns.
Each of the chapters is linked from the navigation bars at the top and/or left side of each page. The next sequential chapter is accessible via a 'next chapter' link at the bottom of each chapter. Any graphs or additional details associated with a chapter are also accessible from the navigation bars.
This was not intended to be a rigorous study of these problems, although there is a lot of interesting data that has very precise mathematical/statistical meaning that could have been generated from the data. What I am presenting is the data that I felt would be the most useful and (more importantly) insightful to the average person planning their retirement.
A word of warning. Do not confuse precise analysis of historical data with precise prediction of future results.
2/22/01 - Omega Strategy - I did additional analysis that took the I-Bond variable out and was a straight-forward comparison of withdrawals using annual rebalancing and withdrawals using the OmegaNot strategy.
1/08/01 - Omega Strategy - I updated the description to clearly state that, while what I did shares some of the characteristics of the 'Omega Strategy', it also has very dramatic differences. I didn't change the analysis, I just changed the reference.
1/01/01 - Omega Strategy - I finally found some time to do some more analysis (this working for a living stuff really sucks). This chapter looks into the 'Omega Strategy' suggested by Scott Burns at his website here.
5/29/00 - Simple and Efficient - Added a chapter that did additional analysis of the simple and relatively efficient withdrawal strategy of a fixed percentage of your portfolio value.
5/1/00 - Efficient Frontiers in Withdrawal - Added a chapter that investigated how to optimize your total withdrawals while minimizing the volatility (more specifically the minimum) of those withdrawals. This is the analysis referenced in the 4/19/00 entry.
4/19/00 - Withdrawals Forever. Added a reference to a separate study which supports the data in this chapter. I still plan to get around to completing the analysis of a modification to the withdrawal strategy referenced in the 3/29/00 entry.
4/1/00 - Fundamentals. Added a chapter that discusses in very general terms the impact of volatility during the withdrawal phase.
3/29/00 - Withdrawals Forever. A short additional discussion of the data and a reference to a more effective withdrawal method if your goal is to maximize your income, 'withdraw forever', and are willing to accept the volatility associated with achieving these goals.
3/26/00 - Withdrawals Forever. This chapter analyzes the withdrawal rates that could be sustained 'forever'. The intent was to provide guidance to folks with retirement horizons longer than 30 years (which is probably most readers). To be quite frank I am very uncomfortable with the conclusions that could be drawn from a straight-forward interpretation of this work, but I'm presenting this data as is regardless as the data appears to be correct.