Decumulation Factors AFI & AAFI (Inflation-Adjusted)

[yogibearbull]

Decumulation Factors have been updated to AFI & AAFI that are inflation-adjusted (previously, unadjusted AF, AAF). Examples have the recent data.

For 01/1987 to 10/2023 (almost 36 yrs), VWELX could support AFI = 152.09, AAFI = 170.68 for inflation-adjusted $12/yr or $1/mo withdrawals. Be aware that Bengen dealt with the worst case survivability that led to AFI = 300 >> 152.09. So, if one followed Bengen Rule, one now has a huge portfolio balance (but that is better than running out of money).

ybbpersonalfinance.proboards.com/post/1249/thread

[yogibearbull]

Jan 2020 was a tough start for withdrawal programs in the recent history. I was curious what would have worked? PortfolioVisualizer (PV) came handy for SWRs:
4-funds 5.76% (AMCPX, AGTHX, AIVSX, ANWPX) (Dave Ramsey controversy on Twitter – his 8% failed)
VWELX 7.25%
FBALX 6.83%
ABALX 7.40%
VFINX 4.34%
ybbpersonalfinance.proboards.com/post/1253/thread

[Mustang]

I waded into some of the details. I worked through some of the numbers. It appears to me that the Safe Withdrawal Rate (SWR) is actually the maximum withdrawal rate since the final balance equals zero.

EOY withdrawals beg the question, what are the retirees living on during the first year of retirement? Assuming a 4% initial withdrawal on $1M, is the first $40,000 set aside outside of the calculations? With the first withdrawal taken at the end of the year is it also adjusted for inflation for the second year of retirement? I used January withdrawals with the second year's withdrawal adjusted for inflation in my spreadsheets to avoid this type of complication. But when coming up with a Safe Withdrawal Plan to cover the next 30 years the difference is probably insignificant.

What is significant is picking the dates for coming up with a plan. Its no surprise that different time periods favor different asset allocations. It is a balancing act to find the funds that are the safest to use in the future. Picking 8% might result in prematurely running out of money. When testing Wellington even using 4% for a 1968 30-year retirement caused the fund to run out of money after 25 years.

I like Portfolio Visualizer. It is easy to use.  It makes comparing funds easy.

Just a few comments from a non-professional. 

Edit: Recent history (January 2020 start) has been kind to two of my funds.

[yogibearbull]

Mustang , great points.

I have tried to stick with the PV SWR definition for the related factor AFI. It has 1st withdrawal at the end of the 1st yr only, and that is also inflation-adjusted. Also, the industry assumption for SWR is that the ending balance at the end of the period is zero and failure is only when the money runs out before the end of the term. I suppose the assumption is that people would die before 30-40 year term. If I was starting from scratch on this topic or PV programming, I would define things a bit differently.

Left unanswered is what do people live on during the 1st year, or after the term-end?

If the withdrawals are started right away, then the money may run out 1+ year before the term-end. So, some adjustments can be made for earlier starts.

I was also bothered by the "after" scenario. That is what my AAFI calculations are for - it provides inflation-adjusted initial principal at the term-end (but the withdrawal rate is lower than the SWR). In some cases, a valid AAFI > 0 may not exist. People can then restart another withdrawal program or just do it just ad-hoc.

One unexpected benefit of this is that it removes the principal-lost aspect of the regular SWR. For example, SWR over 10 years may be deceptive for its high withdrawal (due to principal-lost) but the money would be gone after 10 years; but SWR over 30-40 years may as well be lifetime for most. With AAFI, one can do appropriate withdrawals over 10, 15, 20,...,30, 35, 40 years without worrying about principal gone.

Note that,

SWR = 1,200/AFI

Reduced SWR, or SWRM = 1,200/AAFI.

My objective has to provide people easily useable tools for analysis. They can decide on their own magic withdrawal number looking at examples.

[archer]

Mustang, "What is significant is picking the dates for coming up with a plan. Its no surprise that different time periods favor different asset allocations."

I like to use 2000 for the starting date when I do analysis for PF survival. This is definitely erring on the worst case scenario, and likely to leave a lot of money on the table when I die. However, while there isn't much that can be done if one runs out of money, there are more possibilities for adjusting withdraws as PF balances increase and life span decreases, and for many of us there are heirs to consider.