Portfolio Longevity through Annual Distributions

[Markola]

I was fooling around on Portfolio Visualizer and stumbled on something pretty remarkable to me, if it is true. It seems that taking a single annual distribution from a portfolio each year to cover living expenses is superior to quarterly or monthly distributions. In my reading about ER, I haven’t encountered this phenomenon before.

Check these graphs out. Each is for a $2 million 50% stock/50% bond portfolio from which $120,000 with inflationary increases is with drawn each year. I used the oldest data I could on PV but the phenomenon seems to hold no matter what the initial year is.

I wonder if I’m making an embarrassing input mistake or if there is some other kind of error? Intuitively, I would have thought that monthly distributions would be superior to annual, since more money would remain invested longer, or that there would be no significant difference across the three. However, annual distributions outperform. The annual distributions across the three over time grow essentially in tandem, so that’s not it. What do you think is going on here?

1) Monthly distributions of $10,000
2) Quarterly distributions of $30,000
3) Yearly distributions of $120,000

 

[N02L84ER]

My guess, the $10,000 January withdrawal remains invested an additional 2 months in the quarterly scheme and an additional 11 months in the annual scheme. The $10,000 February withdrawal remains invested an additional 1 month in the quarterly scheme and an additional 10 months in the annual scheme. In other words, the monthly withdrawals happen at the end of each month, the quarterly withdrawals happen at the end of each quarter, and the annual withdrawal happens at the end of the year.

 

[Plantman]

My guess, the $10,000 January withdrawal remains invested an additional 2 months in the quarterly scheme and an additional 11 months in the annual scheme. The $10,000 February withdrawal remains invested an additional 1 month in the quarterly scheme and an additional 10 months in the annual scheme. In other words, the monthly withdrawals happen at the end of each month, the quarterly withdrawals happen at the end of each quarter, and the annual withdrawal happens at the end of the year.

I agree with this assessment. I bet your initial expectations would be realized if you took the withdrawals at the beginning of each period.

 

[pb4uski]

Yes, if the first withdrawal is taken out of the $2m so in each case the starting portfolio is $2 million less one withdrawal, the resuts are inverse... monthly is best, followed by quarterly then annual.

 

[Markola]

Thanks all and you’re right. I dug in to the data to see that the annual contributions are taken in month 12, giving that option the creeping advantage over time. “Time in the market…”