Starting in 1929 invested in the S&P 500, with the four straight years of losses the $300,000 fell to $101,300 at it's Nadir but by 1959 30 years after the start the portfolio would have 2.9 Million dollars.
Worst year I found to start after was 1969 where the portfolio fell to $211,503 in 1974 but at the end of 30 years the portfolio was at $2,624,707 with a 2.9% withdrawal from the portfolio, the worst year in the 1969 sequence was 1981 where the withdrawal was 8.59% of the $311,000 portfolio at the time. Of course at that time the S&P 500 dividend yield was 5.6%.
In most cases when you start year by year the withdrawal rate never gets above 3% so that in most years you are not even spending all the dividends. Even during the great depression because of the negative inflation the withdrawal rate was under 2 percent - and the dividends then were in the 6-9 percent range, meaning you would have been purchasing stock equal to 4-7 percent of your portfolio at the nadir of all time lows in the stock market.
I do not believe there are any failures with these parameters starting in any year from 1926 forward.
Out of curiosity, did you model this annuitization/equity approach in FIRECalc, or make your own spreadsheet? The starting income requirement is $30,900 (17,500 from the annuity and $13,400 from the investments), but for each following year we need to have the withdrawal amount adjusted for inflation from that $30,900. For example, if inflation is 5% for 10 years, our withdrawals would be:
Year .. Inflation%....Total wthdrwl amnt.....from annuity.... from investmts
0.............0.............$30,900......................$17,500...........$13,400
1.............5.............$32,445......................$17,500...........$14,945
2.............5.............$34,067......................$17,500...........$16,567
3.............5.............$35,770......................$17,500...........$18,270
4.............5.............$37,559......................$17,500...........$20,059
5.............5.............$39,437......................$17,500...........$21,937
6.............5.............$41,409......................$17,500...........$23,909
7.............5.............$43,479......................$17,500...........$25,979
8.............5.............$45,653......................$17,500...........$28,153
9.............5.............$47,936......................$17,500...........$30,436
10...........5.............$50,332......................$17,500...........$32,832
Normally, if we had 5% inflation for 10 years, we'd expect the withdrawal amount to increase to be 1.71 times the initial withdrawal amount. But in the above case, we see that our withdrawal amount from the portfolio has increased to be 2.45 times the initial withdrawal amount.
If a person starts with low withdrawal percentage from the portfolio (as you did in your example--1.4%), then things might be okay. But if a person starts with the "normal" WR of about 4% from their portfolio, it might get ugly. The 4% WR generally works fine if it just has to keep up with inflation on that initial amount. But if it has to keep up with inflation on an amount that is more than double the starting amount (as in the case here), that seems more problematic.
IIRC, the research by Moshe Milevsky indicated that annuitizing something like 20-30% of a portfolio had historically provided the highest lifetime utility.
The advantage of the system is easy to see to me, it is very conservative in nature with the 3.5% withdrawal rate. The annuity provides 77% of the needed funds in the early part of the retirement, avoiding the sequence of returns shock and giving stocks the chance to catch their footing. As time goes by you are for the most part only spending the dividends from the S&P500. And in most years you are purchasing shares not selling shares.
Yes, I can see the attraction. I can see another variant that would be interesting is an annuity and a "% of year-end portfolio value" withdrawal method, with some portion of any amount above inflation being used to buy additional SPIA contracts. Given the mortality credits with increasing age, the new contracts might be enough to allow the annuity income to keep up with inflation.
Thanks very much for working the numbers.
