I assume this is a post about the pitfalls of the 4% rule
WFJ wrote: ↑Sun Apr 24, 2022 10:23 pm
4% Pitfalls
Sample vs Population, Shape of distribution (volatility assumption, normal vs skewed), fat tails, Point estimate vs confidence intervals, SORR vs run probability, extrapolation beyond 30 years.
Can you please explain what this means ?
My take is that you are trying to state that results can change with different data-sets. I don't believe that many people believe differently. If you have believed differently I think you need to reassess your understanding of data and statistics.
WFJ wrote: ↑Sun Apr 24, 2022 10:23 pm
Sample vs population
The 4% rule assume all possible market returns are contained in the sample period.
No it doesn't. It doesn't do anything of the sort.
WFJ wrote: ↑Sun Apr 24, 2022 10:23 pm
Shape of distribution
The 4% rule assume returns are normally distributed and not skewed.
No it doesn't.
I'll try and explain what a WR is. A withdrawal rate is the amount that you withdraw from your portfolio, So if I have a million dollars and I withdraw 40k then I have a 4% WR for that year.
Various people have completed studies on different types of portfolios and different WR"s. These are studies looking at historical data. Results have been produced from various studies. The Trinity study stated that a WR of 4% was considered safe in that it had a 95% probability of success within that data-set and that portfolio over 30 years.
That 4% has become a figure that people use as a guideline to working out how large their portfolio has to be to survive.
No one who understand statistics or data in any way shape or form should come to the conclusion that a 4% WR is guaranteed to be successful over 30 years. This show someone who doesn't understand statistics.
Now there are other problems with thinking like this as well. You cannot accurately set your WR for life. So in the real world your expenses are variable and life changes etc.
All you have is a guideline. All you have is some metrics to help you gauge how you are tracking.
The key point that I think that you aren't understanding is that there is a massive difference between a back tested data-set with certain parameters and believing that those conditions will hold true in the future. You appear to believe that if the data-set and the results are not completely the same there is a problem but that is an inaccurate assessment.
So if you think a 3% WR with extremely low expenses is safe then you don't understand how data works and specifically how back-testing data is different to managing a real time scenario.
It's all just a guideline.
Those aren't pitfalls at all. They are just how data and statistics works.