Withdrawal Rates

[audreyh1]

For taxable accounts, where one has to pay taxes on cap gains and dividends, it does reduce the return. And if one is in the 15% bracket, then there's no tax.

Still, if you have to pay taxes, most people would just account for the taxes in the portfolio return, and not in the WR. Then, when you draw 3.5% WR from it, you get to spend the entire 3.5%.

In my case, which is similar to that of many posters here, we have both before and after-tax accounts. The money drawn from before-tax accounts would be fully taxed, and with the WR of 3.5% I would have less than 3.5% to spend.

Other way round, seriously. My withdrawal is absolutely pre-tax, and I get to spend what remains after paying taxes.

None of the portfolio survival models take into account taxes. They are all based on pre-tax returns. If you do otherwise, you are ignoring the models and any pretense of "safety".

Therefore, any taxes incurred by the portfolio have to be paid out of the withdrawal.

If you have a tax efficient portfolio, you get to spend more of what you can safely withdraw.

 

[audreyh1]

If taxes affect the rate of growth of a portfolio, how can they just be considered an expense and not be included in expected returns of said portfolio?

It taxes are paid for out of what is withdrawn each year, then they have no direct affect on the portfolio performance. Only the annual withdrawal affects the performance.

And all models work this way - for good reason it's much simpler.

Given the widely varying tax situations, among retirees, it's too difficult to model after tax SWRs.

 

[NW-Bound]

Well, you pay the tax man now, or in my case, I paid him earlier.

I just looked and saw that my portfolio still has 26% of it in after-tax accounts. As that portion is not sufficient to support 3.5%WR, I have been drawing on principal in that portion. However, I have a lot of headroom in the 15% tax bracket to not having to pay any taxes on cap gains and dividends, and still have enough room left to do some Roth conversion and pay a bit of tax on that.

The Federal 15% bracket of $72,500 for a couple plus the standard deduction and exemption means one can pay little tax up to near $100K of investment income! Is that fair? I do not think so, but that's what the law is.

The way our progressive tax system works, a guy who earns (by working, not investing) $200K one year then nothing the next is nailed with AMT, while his neighbor who earns $100K each year escapes with relatively low taxes. Ask me how I know, a guy who had had sporadic income from part-time consulting work.

Roth conversion is a good way to do "income averaging", and I will be sure to make full use of it to avoid high progressive tax in a future year when I may need a higher WR than normal.

 

[tuixiu]

The Federal 15% bracket of $72,500 for a couple plus the standard deduction and exemption means one can pay little tax up to near $100K of investment income! Is that fair? I do not think so, but that's what the law is.

This is exactly what we plan to be, and I think it would be really interesting to see how many people have this situation, where high net worth and living off investment income higher at least 50% higher than median household yet having zero to little tax liability by limboing under the 15%.

Surely higher proportion on this forum (things like roth conversions aside) but I mean out of all households in the United States.

 

[Options]

It taxes are paid for out of what is withdrawn each year, then they have no direct affect on the portfolio performance. Only the annual withdrawal affects the performance.

And all models work this way - for good reason it's much simpler.

Given the widely varying tax situations, among retirees, it's too difficult to model after tax SWRs.

+1

This is exactly what we plan to be, and I think it would be really interesting to see how many people have this situation, where high net worth and living off investment income higher at least 50% higher than median household yet having zero to little tax liability by limboing under the 15%.

Surely higher proportion on this forum (things like roth conversions aside) but I mean out of all households in the United States.

I went to hell and back trying to estimate future taxes last year. Finally, settled on median of taxes between FIDO/ORP. Not the most scientific way, and probably overestimating, but any future difference between real and overestimated means more fun $ in my case.

 

[ERD50]

That is a good point, I've always just run 50 because it seemed the safest but looking at various time spans for my numbers is kinda funny in firecalc it goes back up after 45 years.

Success rate:
30 years 98.2%
35 years 95.4%
40 years 95.1%
45 years 94.9%
50 years 97.8%
55 years 97.7%

Yes, 1966 is the specific killer. Run a scenario that excludes 1966, and success % starts increasing.

-ERD50