Yes, indeed last year due some one time tax credits, her effective tax rate was about 18%.
She loses $22000 of the tax break in the 403(b) contributions as those are getting converted to the Roth,
I've already done all those calculations. There is no withdrawal tax rate to consider. Her retirement tax rate will be higher than her current tax rate, not to mention that the Roth is never taxed.
That's a bogus rule. Tax deferred growth in the Roth will, at some point, make up for higher tax rates at withdrawal.
I hope that readers understand that 'effective tax' (tax paid divided by one's income) is interesting, but not very useful. When it comes to taxes, life happens at the margin. The next dollar you earn (or convert to Roth) is taxed at your marginal rate. The last dollar you deposited pre-tax saved you tax at the marginal rate. If you are right on the line these two rates may differ.
I find Elle choosing her words carefully, 'general rule'. Fewer and fewer people are going to retire in a higher bracket, and since it's not an all-or-none, those who are in a lower bracket at retirement will be best served by converting enough just to top that bracket off each year. This may very well complete a full conversion of all IRA assets before your "at some point" has passed.
While it's none of my business, you hinted that she'll never spend all the money. If she's leaving any to charity, the IRA would go to them tax free, so she's just shortchanged them. If she's leaving to a younger generation, unless they are all well paid professionals, they might have stood to inherit a larger sum. Given that medical expenses exceeding
7.5% of one's income can be used as Sch A deductions, a very bad health year can turn into an opportunity for a low tax rate withdrawal or conversion. Given that your sister is not at the top of the 28% bracket, and would have at least $70K/yr to get there, I fail to see how putting her in the 35 can be better in the long term.
As Phil stated, you'll have a better chance of understanding the 2012 tax rates by October '11. That's the answer to the exact question you asked. No one can be sure today.
It's an 8 page PDF and worth reading. I read this earlier today. It is indeed worthwhile and, as others, you and I have discussed, it points out that the decision on whether to convert to a Roth will hinge largely upon current tax rate and retirement tax rate. Expected rate of growth is irrelevant.
The math is simple enough. Yet I do not want to browbeat you or anything. For the interested reader: What I presented (and backed up by other posts here) is repeated in many online fora which get questions about whether to convert. The math is simple enough. Folks doubting the point others and I are making here about current and future tax rates being the main criteria may google and see it re- stated.
I realize you're done with it. :-) I am posting for the archives yada, for others trying to decide whether to convert.
I am a little curious about your statement that "the tax hit will be the estate tax." I am not sure how this is relevant to a decision on whether to convert or not. Unless you are thinking, 'Well my sister's estate will now be less by some $175k (the tax paid on the conversion)." But that $175k is taxes paid and so money given away. That is, your way, your sister may be paying taxes (the conversion tax) to minimize the estate tax, which may be self-defeating.
Both a Trad IRA and a Roth IRA get counted as part of one's estate when one dies.
If indeed your sister never touches the Roth and so it goes to the beneficiaries she designates, then interestingly non-spouse beneficiaries do have to take distributions. If a five-year limit yada is not met, non-spouse beneficiaries may even have to pay taxes on the Roth IRA's distributions.
- Non-spouse designated beneficiaries may take RMDs over their lifetime. No (income) tax due regardless of how much they take out. Any estate tax is paid for by the estate of deceased, not by the beneficiaries.
It's relevant in the sense that whether the Roth has the desired positive effect on future net wealth will in some common scenarios depend upon the rate of return.
The scenario I described before -- 35% tax rate during year of conversion, and 28% during subsequent years -- is probably representative of reality for a lot of people.
But back to the PDF file on Joe Taxpayer's site, I think the following statement is misleading, possibly so misleading as to be wrong:
"Translated into pragmatic terms, if the client's tax rate in retirement will be the same as their tax rate while working (for example, 35 percent) and, regular or Roth, they expect to draw on their retirement funds at the same rate, over the same period, then in the language of economics, this client is indifferent to the choice between a regular 401(k) and a Roth 401(k)."
The reason it is misleading is that it assumes the TP has an certain amount of money, and that if they choose Roth they are necessarily contributing less. That's not a very valid assumption. It is often more valid to compare traditional vs. Roth when contributing the maximum permissiable amount in each case.
On Apr 26, 8:55 pm, Elle wrote: MODERATOR'S COMMENT:
The following is what I had in mind, from
formatting link
others:
"One rule that does not change is the requirement for the Roth IRA to exist at least five years before earnings can be withdrawn tax-free. To be more precise, earnings can be withdrawn tax-free beginning on the first day of the fifth taxable year after the year the Roth IRA was established. That means January 1, 2013 for Roth IRAs established in 2008.
As a result, a beneficiary may have to pay tax on earnings withdrawals if the original owner's death and the beneficiary's withdrawal both occur shortly after the Roth IRA is established. This result isn't as harsh as it may seem, however. The tax only applies to earnings that built up after the contribution to the Roth IRA."
Right you are. Hopefully the beneficiary isn't going to deplete the account at such a pace as to take earning sooner than the five years from conversion have passed. If, for whatever reason, they did, the above is a factor.
Well the effect a Trad IRA and any non-tax protected accounts have on future net wealth also depend in the same way upon the rate of return.
It may be a scenario for some people, but let's be clear that (1) it goes against the standard advice (advice based in not difficult financial mathematics) to convert only as much as will "top off" one's current bracket (= the bracket before the conversion amount is considered); (2) if wisdom is behind such a decision, then the decision will likely emphasize one will be in a bracket (or as Joetaxpayer's link points out, at an effective tax rate) higher than
35% at retirement; and (3) the person is confident of his/her crystal ball abilities for his/her situation.
Honestly, if someone is in this high a tax bracket with such a large income and IRA assets, and with so much money at risk of being thrown away on potentially unnecessary taxes, I would hope the someone would hire a specialist with experience in this kind of decision-making, someone who is capable of good communications to make the case for whatever decision she or he favors. Again, I am posting for those readers (of sizable means!) and do not wish to burden the OP, who has made his decision.
I do not see this assumption in the paragraph from the paper above.
You do not get as high a value in the non-Roth case. To get the same value in both cases requires deferring less in the Roth scenario (by a factor dependent on the tax bracket), and this was the assumption made in the PDF file.
Put another way, under Roth the amount you spend on tax up front is effectively invested tax-free. If you do not do the Roth, you must invest this amount in a taxable fashion.
Finally, a nice numerical example. I follow case 1. The first C2 result (taxing the $132665K) yields $95519. But the $14K should only have the growth taxed, and result in $30665. Total = $126184.
The $7K shortfall is caused by the tax on the growth in the $14K as you pointed out. A 15% cap gain tax improves the number to $129184, or a $3472 shortfall.
Agree so far? The real question is this - since 20 years of exactly 28% produces so slight an advantage to Roth, are you willing to bet it all that way? If I tinker further and assume withdrawals at just 25% marginal, I gain a slight advantage. Will the pension income increase at a pace that the client will stay at 28% or will there be years with the potential to convert at a lower rate? Note - if a 3% lower marginal rate breaks me even, they do you agree the OP's conversion right into the 35% bracket would have been better off spread out at 28%? 7% spread? You can easily change your numbers to show $17,500 needed in the side account, etc. About 10% of the IRA balance each year getting converted, but at 28%.
Your example is excellent and I concede shows a Roth advantage given the assumptions for the reasons you've stated. But I'll maintain that much can happen to throw a wrench into those assumptions. I've mentioned a high medical bill year. How about a high taxable account stock loss? Maybe the sale of appreciated assets in that $14K kitty are offset by that.
However, this is applying the 28% tax only at the very end, whereas in the base case it's applied each year -- if the investment generates regular income (i.e. interest).
If the investment generates unrealized capital gain, that defers taxes so yes. Generally capital gains treatment subtracts from the relative advantage of having investments in a qualified plan; and Roth is just a way of effectively shoving more money into the qualified plan. Preferential capital gains rates skew it even further.
Not me; I have only Roth-converted a small fraction of my qualified plans. Maybe 10% of the total. It amplifies risk too much to be compatible with my overall planning. I'm always surprised by the large fraction of people who are totally convinced they need to Roth-convert everything.
I'm also surprised by the large fraction of people who think that Roths will never, ever be taxed! Mark my words . . ."the "wealthy" have amassed large sums of money in Roths. It would simply be unfair not to receive a reasonable amount in taxes from such money . . ."
The paper starting on page 2 with the John and Jane example does what you did, but with, IMO, more reasonable assumptions about taxing the annual growth of the money in the taxable account. The paper arrives at the same result you did:
"Table 1 provides a sample spreadsheet showing one possible implementation of a side-fund analysis. A key issue is the amount of tax to subtract each year from the side fund. A typical approach is to assume that the side fund, like the other accounts, is invested in a
60/40 stock/bond mix, with each year's 8 percent return consisting of
2 percent interest taxed at 28 percent, 2 percent dividends taxed at
15 percent, and 4 percent of (unrealized) capital gains, on which no tax is paid until the end. In this application of the side-fund analysis, John and Jane are shown to be better off with the Roth 401(k) ?the after-tax value of their accumulation in the Roth is about $76,000 greater than what would have been the sum of their after-tax accumulations in the regular 401(k) and the side fund. If we assume that after 2010, dividends will again be taxed as ordinary income, and the capital gains tax will revert to 20 percent, the Roth advantage is even greater: just over $84,000."
The paper rebuts this result as follows:
"Had John and Jane been unable to save any more than 20 percent of their income, the side-fund analysis would not apply (fully funding the Roth requires almost 30 percent of their earnings). Likewise, if the government had set the tax-sheltered contribution level at or above the maximum that John and Jane could set aside for savings, the side-fund analysis would not apply. Similarly, the side-fund analysis would cease to be applicable, for any client, if dollar limits on
401(k) contributions were removed. The first problem with the side- fund analysis, then, is its limited applicability.
The second problem with the side-fund analysis in Table 1 is that its result, favoring the Roth, is special to the assumptions used. By drawing on current thinking about asset location, and by altering a few other assumptions, we can ultimately reverse the results of Table
1, and produce a side-fund analysis that shows a regular 401(k) plus side account to be superior to a Roth 401(k)."
There is more on page 3, and I think it is worth studying closely. I think its objections to the way you wish to analyze Roth vs. Traditional are pretty clear.
No, it isn't: at a 0 growth rate, the only thing that matters is the relative tax rates. At a 50%/year growth rate, the taxable account grows at 30%/year, so converting now is better, the more so the longer the account will be active.
Yes, if tax rates are constant, and returns are positive, the Roth always wins under the assumptions in my analysis, or in the author's "Side Fund Analysis".
This is a non-rebuttal. It ignores the simple premise that you're only doing a Roth because you've already deferred the maximum legal amount, and you would like to gain the tax advantages of deferring more, which you can gain with a Roth.
The author's "self funded analysis" is simply a statement that, if you defer less than the total allowed, and then Roth some or all of it, you are then in a wash with a non-Roth approach with the same outlay. This is important to know, but it is not the situation under which (most) people are attracted to Roth, which is when you're already at the contribution limit.
I disagree here too. "its result, favoring the Roth, is special to the assumptions used." The assumption of all income being regular and taxable is not a "special" assumption, it is the baseline for many investors, and other assumptions with different tax qualities are evaluated case-by-case. As I have said if there are cap gains in the mix, the favorability of deffering those gainst drops.
I still don't much like the analysis, and I still don't think it manages to crystallize the main advantage of a Roth and when you would want to use it.
where the employed tax rate = the retired tax rate.
Many web sites make this point (yada) as well. The only way it is unreasonable is with a lot of assumptions. For example, your assumption that 20% of each year's growth goes to taxes. The paper addresses myriad assumptions and how they can be manipulated to argue both for and against the Roth.
Steve, we do not agree on how to interpret the paper. I have generally not followed your reasoning from the first post of yours I addressed. I leave it to readers here to peruse the paper and judge for themselves. I continue to feel it is a worthwhile read. It's not like the authors had something to gain from the points they make.
Yes. Consider someone who has $X in an IRA, and $Y in cash (and the cost of converting the IRA to Roth is paying the $Y in taxes).
Money inside the IRA (whether or not converted) grows at some rate, tax-free. Money outside grows at the same rate, less taxes.
That formula has many assumptions built into it. It is false if they're wrong.
I'm in a 40% bracket, maybe higher with new tax rates. (That's state and federal combined.)
Tax rate after retirement lower: against Roth. Much growth between now and retirement (long time and/or high rates): for Roth.
When you claim that a factor is irrelevant, and I show that it matters, claiming my assumptions are unreasonable isn't meaningful; it still matters. If my assumptions are provably false, then I have to show that it matters under other assumptions. For someone in a 30% bracket, earning 10%/year, the after-tax net is 7%/year; if he has 30 years before retirement, that's the difference of about 2.5 times as much money.
BeanSmart website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.