Diminishing Returns
I have seen arguments for relaxing safety requirements on the basis of diminishing returns. For example, William Bernstein mentions diminishing returns on page 236 of "The Four Pillars of Investing." Considering that the average nation lasts for only 200 years, he questions whether it is meaningful to plan for 95%+ safety for 40 years.
If you estimate your odds of portfolio failure due to political instability to be 10%, you should calculate your overall portfolio safety as 0.90*(the safety level in the absence of political disaster). If the safety level at your calculated withdrawal rate were 99%, your overall level of safety would be 89.1%. If the safety level at your calculated withdrawal rate were 95%, your overall level of safety would be 85.5%. If the safety level at your calculated withdrawal rate were 90%, your overall level of safety would be 81.0%. If the safety level at your calculated withdrawal rate were 85%, your overall level of safety would be 76.5%.
All of these increments make sense. But suppose that your calculated level of safety is 99.9%. Your overall level of safety becomes 89.91%. Going out of your way to reach the 99.9% calculated level of safety does not make sense.
Similarly, if you estimate your odds portfolio failure due to political instability to be 20%, you would calculate your overall portfolio safety as 0.80*(the safety level in the absence of political disaster). If the safety level at your calculated withdrawal rate were 99%, your overall level of safety would be 79.2%. If the safety level at your calculated withdrawal rate were 95%, your overall level of safety would be 76.0%. If the safety level at your calculated withdrawal rate were 90%, your overall level of safety would be 72.0%. If the safety level at your calculated withdrawal rate were 85%, your overall level of safety would be 68.0%.
Does it make sense to try for a calculated level of safety of 99%? Possibly. Going out of your way to get a 99.9% calculated level of safety does not make sense. Your overall level of safety would be 79.92%.
When the odds of political instability are 10%, it does NOT mean that you should reduce your calculated level of safety to 90%. Doing so reduces our overall level of safety all of the way down to 81.0%. When the odds of political instability are 20%, reducing your calculated level of safety to 80% reduces your overall level of safety to 64.0%.
It is better to plan for higher levels of safety and then to address the issue of political instability. If your own odds of political disaster are 10% or 20%, you need to make contingency plans. Then you need to integrate them into your investment strategy.
Another area in which I have seen remarks about diminishing returns is life expectancy. I dislike the notion that dying is a way to LUCK OUT. Our dividend-based research leads to a much better solution. We have increased today's safe withdrawal rate to 4.8% (plus inflation). We have extended the time period far into the indefinite future.
Have fun.
John Walter Russell February 13, 2006
|