Announcement

Collapse
No announcement yet.

Can Traditional 401K/IRA be better than Roth as resident? (Loophole?)

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    Originally posted by wa2106 View Post
    As far as I can tell, the calculations in that thread didn't take into account the taxes owed on the Roth contribution. Any comparison between Roth and traditional has to assume either lower ability to contribute to Roth (19,500*(1-marginal tax rate)) OR assume that tax savings from traditional (19,500*marginal tax rate) are invested in taxable account.
    One of those pesky considerations to which I alluded to having likely forgotten, it seems Thanks. Updated numbers below.

    Click image for larger version

Name:	Screen Shot 2021-06-05 at 12.49.09 PM.png
Views:	145
Size:	23.3 KB
ID:	277097

    Comment


    • #17
      Originally posted by Gamma Knives View Post
      I'm sorry but you are clearly looking at marginal rates incorrectly. The discussion was resident contribution to roth versus traditional IRA not converting a traditional to roth as an attending (edit: or switching to roth contributions as an attending). These are not equivalent to a resident roth contribution as tax rate will be different.

      I've provided details answers and your response has simply been you're wrong. I am trying to help you understand (or see what I'm missing) but your response is not helpful.
      How am I "clearly" looking at marginal rates incorrectly? Keep in mind I didn't join this thread until someone pointed out that Roth in residency was mathematically superior. I contested this assertion. I am simply stating that as a resident you'll probably be making a marginal decision in the 22% bracket. The question is about the marginal rate on the back-end for THOSE DOLLARS CONTRIBUTED. I kept trying to ask you to answer my question, which you didn't do, that inquired about a late-stage attending with a high 401k balance. The point of that inquiry was to motivate you to think on the margin in that situation. The person has a nest egg. They will be withdrawing money per their plan or mandated RMDs, and the marginal tax rate on withdrawal has grown considerably. To put a finer point on it, if someone has $5M in pre-tax late in their career and wants to spend or convert this before they die and do so over a 25 year period from, say, 62-87 then their annual withdrawal will be $354k, assuming a 5% growth on underlying funds. More if a higher rate assumed. Add SS to the mix from 70-87 and you're talking a 28%+ tax rate on withdrawal (assuming this bracket goes back after 2025 and doesn't get affected by future tax policy, also inflating the current brackets for inflation). So any ADDITIONAL money contributed (on the margin) in that late attending's career will be saved on whatever their current rate but get taxed at 28%+ on the margin later. This would increasingly favor a Roth 401k for many attendings. The point of this whole exercise and my question was to state the obvious - that decisions are made on the margin and you don't just get to assume that the dollars contributed late in attending-hood get mixed in and come out at some effective rate. They affect the margin.

      So applying the same concept to someone starting out fresh you don't get to all of a sudden apply some other concept not involving a marginal assessment. The resident is likely in a 22% marginal bracket and has little to nothing saved. The question is whether or not THAT MONEY CONTRIBUTED will come out at less than or more than 22% ON THE MARGIN later when they retire. As they have nothing saved indeed this is the first money to come out, and not all at once. You seem to want to say (and maybe I'm misreading things) that the resident is going to have their money that's currently being contributed taxed at some effective rate later or perhaps at the top marginal rate on withdrawal in the future. That is not the case. Those currently contributed dollars and their growth will come out on the marginal tax rate based on their draw down plan, which may be dynamic based on their actual dollars saved. Whatever way you slice it the $19.5k of pre-tax contributions and its growth will not be coming out at a high marginal rate, whether they be drawn down starting at 62 (fully recognizing this won't be enough money to retire on and SS will need to come out earlier) or delaying retirement to allow for more pre-tax growth and better SS amounts. Again, you should always be thinking on the margin, whether it be for later amounts contributed as a late-stage attending, or as a resident just starting out. I'm happy to see your math describing otherwise. And don't forget to adjust tax brackets for inflation. And when you say "marginal rate at retirement" you have to look at different marginal rates depending on when you are contributing and how much you have in underlying funds. There is no one "marginal rate at retirement".

      Comment


      • #18
        Originally posted by bovie View Post

        One of those pesky considerations to which I alluded to having likely forgotten, it seems Thanks. Updated numbers below.

        Click image for larger version

Name:	Screen Shot 2021-06-05 at 12.49.09 PM.png
Views:	145
Size:	23.3 KB
ID:	277097
        Not sure why you would use 20% LTCG rate if you're expecting to be in 25% tax bracket - that would imply 15% LTCG rate. Also, are you assuming an effective tax rate of 25% or marginal tax rate of 25%? I assume marginal and it's hard to imagine that all of your traditional contributions would be withdrawn at a 25% rate. A large portion would be 0-15%.

        Comment


        • #19
          Originally posted by ENT Doc View Post
          The resident is likely in a 22% marginal bracket and has little to nothing saved. The question is whether or not THAT MONEY CONTRIBUTED will come out at less than or more than 22% ON THE MARGIN later when they retire. As they have nothing saved indeed this is the first money to come out, and not all at once. You seem to want to say (and maybe I'm misreading things) that the resident is going to have their money that's currently being contributed taxed at some effective rate later or perhaps at the top marginal rate on withdrawal in the future. That is not the case.

          There is no one "marginal rate at retirement".
          Every year you file income tax there is one marginal rate for you that year (retired or not). That marginal rate is dependent on how much taxable income you have that year. Money is fungible. The concept of money you contributed as a resident comes out first and other money comes out later does not make sense. The pre-tax money is all in one pool.

          If a choice as a resident results in higher taxable income decades later in retirement that additional income is effectively at the marginal rate when compared to making a choice that results in less taxable. When viewing a change in income it is viewed at the margin.

          You are arguing for potential roth contributions at later career when they will be more expensive (due to higher marginal tax rate) and against the roth contribution when they are less expensive as a resident?

          edit: your description would seem analogous to someone saying I pick up lots of extra shifts in January because then I'm paying a low income tax rate as I haven't gotten to the higher brackets. I never pick up shifts in December because at that point I'm in a high bracket.

          edit2: if you are changing one variable you should consider the effect on the margin. If it is doing extra shifts you should consider those at your margin. If it a plan that will result in a higher pre-tax account you should consider the difference at the margin. It is the difference to consider not tracking individual dollars or contributions.
          Last edited by Gamma Knives; 06-05-2021, 12:24 PM.

          Comment


          • #20
            The math for pretax versus Roth likely favors Roth as resident but not by as large of a margin as one would assume. The main reason to do it from my standpoint is diversification - we don't know how these accounts will be treated in the future so I'd rather have Roth in residency at a relatively low marginal rate, knowing that I will be doing traditional later on at a higher marginal rate. Worst case scenario I'm wrong by a few thousand dollars but at least I'm diversified.

            https://www.whitecoatinvestor.com/ro...ditional-pslf/

            Comment


            • #21
              Originally posted by wa2106 View Post
              Not sure why you would use 20% LTCG rate if you're expecting to be in 25% tax bracket - that would imply 15% LTCG rate.
              For now, sure. But that may change. Or it may not. The vast majority of returns here are being taxed at income rates anyway, not LTCG. A 25% tax bracket doesn't even exist now. I just made a not-crazy assumption about something very far into the future, which is currently unknowable. I certainly expect to be in the highest LTCG bracket in retirement though, so I used that. And the conclusion is the same either way.

              Regarding marginal rate, I agree that not all would be taxed at 25%. But considering additional sources of income which I anticipate to be not-insignificant, such as real estate and even social security, and my anticipated spending (i.e., income) requirements in retirement (high), I would expect the lion's share of withdrawals to be taxed at a rate much closer to 25% than 0-15%.

              And while much of this is academic--in that it is highly dependent on assumptions, small changes in which lead to large changes in outcomes--I certainly agree that numbers aside the benefit of future tax diversification and the ability withdraw money tax-free down the line, essentially allowing for a choose-your-own-tax-rate retirement, should perhaps even be the main consideration.

              Especially given that Roth space is limited and once it's gone, it's gone. Can't imagine many retirees would look back and wish they had fewer Roth dollars.

              Comment


              • #22
                Originally posted by bovie View Post

                For now, sure. But that may change. Or it may not. The vast majority of returns here are being taxed at income rates anyway, not LTCG. A 25% tax bracket doesn't even exist now. I just made a not-crazy assumption about something very far into the future, which is currently unknowable. I certainly expect to be in the highest LTCG bracket in retirement though, so I used that. And the conclusion is the same either way.

                Regarding marginal rate, I agree that not all would be taxed at 25%. But considering additional sources of income which I anticipate to be not-insignificant, such as real estate and even social security, and my anticipated spending (i.e., income) requirements in retirement (high), I would expect the lion's share of withdrawals to be taxed at a rate much closer to 25% than 0-15%.

                And while much of this is academic--in that it is highly dependent on assumptions, small changes in which lead to large changes in outcomes--I certainly agree that numbers aside the benefit of future tax diversification and the ability withdraw money tax-free down the line, essentially allowing for a choose-your-own-tax-rate retirement, should perhaps even be the main consideration.

                Especially given that Roth space is limited and once it's gone, it's gone. Can't imagine many retirees would look back and wish they had fewer Roth dollars.
                I think it's safe to assume a 25% effective rate in retirement too when making these assumptions. Due to the fungibility of money I can only chuckle at some of the discussion here about this $$ being first dollar 0% rate.

                I think the key here in my calculations is that people on this board are going to die very very wealthy because of how conservative this group is. As an example, someone correct me if I'm wrong but isn't RMD on a $2.5m 401k/IRA like $100k or something at 72. Thats really quickly rocketing beyond these low marginal brackets everyone has stated so it's kinda silly that people are so zeroed in on the low brackets. SS, dividends on taxable - this group has a serious spending problem (for many). I think a lot of ppl here are going to have way more than that in the traditional accounts so not accounting for that seems weird contortionist type math.

                And I completely agree with wa2106 that diversification has a non-calculable value but it's still valuable.

                Comment


                • #23
                  Originally posted by Gamma Knives View Post

                  Every year you file income tax there is one marginal rate for you that year (retired or not). That marginal rate is dependent on how much taxable income you have that year. Money is fungible. The concept of money you contributed as a resident comes out first and other money comes out later does not make sense. The pre-tax money is all in one pool.

                  If a choice as a resident results in higher taxable income decades later in retirement that additional income is effectively at the marginal rate when compared to making a choice that results in less taxable. When viewing a change in income it is viewed at the margin.

                  You are arguing for potential roth contributions at later career when they will be more expensive (due to higher marginal tax rate) and against the roth contribution when they are less expensive as a resident?

                  edit: your description would seem analogous to someone saying I pick up lots of extra shifts in January because then I'm paying a low income tax rate as I haven't gotten to the higher brackets. I never pick up shifts in December because at that point I'm in a high bracket.

                  edit2: if you are changing one variable you should consider the effect on the margin. If it is doing extra shifts you should consider those at your margin. If it a plan that will result in a higher pre-tax account you should consider the difference at the margin. It is the difference to consider not tracking individual dollars or contributions.
                  Money being fungible doesn't really encapsulate what I'm talking about here. You still haven't answered my question about the late stage physician with a massive 401k and large expected withdrawals. You really think that additional 401k savings just goes into a pot and doesn't affect things on the margin on the back end? You think that additional savings just goes in at a marginal rate and comes out at an effective rate? I'm not sure what you are referring to as one marginal rate. There were multiple brackets you went through. And additional savings vaults you into higher brackets later. Earlier savings had you in a lower bracket on withdrawal because not as much could be withdrawn. More saving and time vaults you into a higher bracket. Different marginal effects based on amount and time and drawn down plan.

                  Take it to the extreme as I did in another example. The resident puts $19.5k into a traditional 401k and saves taxes at 22%. Oh but wait. What if their $19.5k straddles the 22% and 12% brackets and they go into the 12% tax bracket and their final tax filing had them in the 12% bracket? Are you saying they have a single, marginal rate of 12%? Again, you can't look at marginal decisions based on the marginal rate on the tax year filing. Anyhow, back to my point. That resident we'll say contributes all of that $19.5k in the 22% bracket. It grows at 8% for 30 years. They now have $196k. They take it out all at once, go for broke. At that time the $196k would put them in a 22% tax bracket (assuming no changes to this bracket). Are you saying the $196k came out at 22%? I surely hope not.

                  As far as your Jan vs Dec analogy this is incorrect. If you have basal income or # of shifts and are at $XXX salary then any additional shift will gets taxed at your highest bracket, perhaps vaulting you into a different bracket. Presuming you are paid the same per shift in Jan and Dec it would make no difference come tax time. But I'd take the Jan shift simply because I'd want the money earlier to invest.

                  I'm not sure what your second and last paragraph are saying. And I'm not sure what you mean by "more expensive" re: Roth contributions. All that matters is whether or not the marginal rate at the time of contribution is higher or lower than the marginal rate at withdrawal for THOSE DOLLARS.

                  Comment


                  • #24
                    Originally posted by ENT Doc View Post

                    Money being fungible doesn't really encapsulate what I'm talking about here. You still haven't answered my question about the late stage physician with a massive 401k and large expected withdrawals. You really think that additional 401k savings just goes into a pot and doesn't affect things on the margin on the back end? You think that additional savings just goes in at a marginal rate and comes out at an effective rate? I'm not sure what you are referring to as one marginal rate. There were multiple brackets you went through. And additional savings vaults you into higher brackets later. Earlier savings had you in a lower bracket on withdrawal because not as much could be withdrawn. More saving and time vaults you into a higher bracket. Different marginal effects based on amount and time and drawn down plan.

                    I'm not sure what your second and last paragraph are saying. And I'm not sure what you mean by "more expensive" re: Roth contributions. All that matters is whether or not the marginal rate at the time of contribution is higher or lower than the marginal rate at withdrawal for THOSE DOLLARS.
                    Money being fungible is why your fixation on specific dollars does not make sense. I have no idea what "affect things on the margin on the back end" but yes 401k savings all go in the same pot. I have not used the term effective rate so I'm not sure where you are getting that from. Yes there are tax brackets and we both know how those work. The tax brackets are not your marginal rate. If you have a large amount that spans multiple brackets you have to do a weighted average of the spanned brackets but that is deviating from the point.

                    That is why I have not engaged the other scenario as it is not relevant to the question. Sure there can be scenarios where it makes sense for an attending to contribute to Roth instead of traditional. What I was attempting to point out by saying they are more expensive is that if you are doing roth instead of traditional at a higher tax rate you forego more tax savings (from the decreased traditional contribution) at a higher marginal tax rate compared to the same action at a lower marginal tax rate (aka attending versus resident).

                    Doing roth instead of traditional you forego more tax savings (at the time of contribution) if your marginal rate is higher; it is more expense.

                    Comment


                    • #25
                      Originally posted by ENT Doc View Post
                      As far as your Jan vs Dec analogy this is incorrect. If you have basal income or # of shifts and are at $XXX salary then any additional shift will gets taxed at your highest bracket, perhaps vaulting you into a different bracket. Presuming you are paid the same per shift in Jan and Dec it would make no difference come tax time. But I'd take the Jan shift simply because I'd want the money earlier to invest.
                      The shift analogy was describing how you are viewing the situation. It is obviously incorrect as is your argument. That was the point. You identified the error the hypothetical physician was making. That is the error you are making when you assign the first 401k contribution a lower tax rate.

                      When discussing roth v traditional for a resident we are presuming the contributions they make for the rest of their career are static irrespective of the resident contribution (same as assuming the physician in the analogy will maintain the same base salary). The physician following the argument you are maintaining would say I'm not paying marginal tax on "THOSE DOLLARS" because they were the first ones I earned this year. Again that is an error; it the same error as saying the 401k dollars contributed as a resident are 0% because of personal exemption they are additional in this scenario so they are at the marginal bracket.

                      I don't know if this has cleared anything up. Based on the responses you have given, I fear I am doing an inadequate job explaining. Perhaps someone else will weigh in with more clarity.

                      Comment


                      • #26
                        This is now the second thread this week that ENT Doc has made the same argument. I don't think he's budging despite as far as I can tell significant push back from other users, and no backers agreeing with him (nor even likes on his posts for whatever that's worth).

                        not to say that the discussion is not worthwhile - I appreciate getting pushback, but this one is such a bizarre hill to die on.

                        Comment


                        • #27
                          This is again a situation where we're all going to have to agree to disagree. And if I don't have any likes or support then so be it.

                          Comment


                          • #28
                            Just wanted to throw out this Kitces article.

                            https://www.kitces.com/blog/how-to-e...inal-tax-rate/

                            “In principle, calculating a marginal tax rate for future income is the same as in the present – evaluate how the tax obligation changes as, at the margin, income is added or subtracted from the base amount of income that will already be present.”

                            This is the point I’m making. You have a specific allowance of income based on your current savings and draw down plan. Money contributed at the margin on the front end affects the margin at the back end. It’s not some pool of fungible money at some effective rate.

                            Comment


                            • #29
                              Originally posted by Gamma Knives View Post
                              1)The downside that immediately jumps out to me is any gains you have made in the tIRA will be taxed as income when you convert.

                              2) You will likely need to do the conversion the calendar year before you graduate to avoid being taxed at a higher bracket. Will that bump negatively affect you loans that year (I am not familiar with loan repayment rules)?

                              I would not do the conversion after becoming an attending as the income will be taxed at your marginal rate which will likely result in more taxes overall.
                              I've read this thread and while it was hard to follow due to my current level of financial knowledge. I might be wrong but I felt it spiraled a bit off topic and may have been made more complicated than necessary. But, your bolded statements are key (this is in reference to a non-PSLF scenario as that is my situation).

                              1)You'll get hit with a bit of a tax bomb following the conversion (so save accordingly for this event), but the $ in taxes owed following conversion at the end of your residency will be 'less' due to inflation making that money 'less valuable' than the money you saved in previous years on taxes (that money was worth more). ex: $10,000 in taxes 4 years ago was 'worth more' due to inflation than owing $10,000 in taxes today. Did that make sense? It is hard to word.

                              2) Most people convert to a private loan after residency as the rates are now cheaper than when they were in residency on an IDR, such as REPAYE. You are correct, making that conversion after you become an attending would be very costly. However, if this conversion were made prior to becoming an attending, it would be a huge benefit based on my my OP / #1 I wrote above. Someone correct me if I am wrong but, in the final 6 months (may a bit earlier, maybe later, but lets stick with this), as long as you have an attending contract in hand, private companies will refinance at a lower rate.

                              So you do 2 things during your final year of residency:
                              1) During the first 6 months of your final year of residency convert your Traditional to Roth (after already receiving in previous years tax deductions, lower IDR monthly payments, and taking this money and saving / investing during those years. And also the money saved in prior years being more valuable than money in the future due to inflation). Then
                              2) During the last 6 months of your residency, assuming you have a contract in hand for an attending position, sign on to have your loans converted to a private loan as being on an IDR will now be financially unfavorable relative to a low rate private loan.

                              In the end, for simplicity, assume you have 3 residents, all maxing out their 401k and IRA after 4 years totaling to $102,000. Resident A did 100% Roth, Resident B 100% Traditional, and Resident C 100% traditional but converted to Roth prior to attending salary. By my OP and my rationale stated above, wouldn't resident C come out the farthest ahead?

                              Anyone feel free to answer or contribute your thoughts thanks for the interesting discussion everyone

                              Comment


                              • #30
                                Originally posted by asdfgghk View Post

                                I've read this thread and while it was hard to follow due to my current level of financial knowledge. I might be wrong but I felt it spiraled a bit off topic and may have been made more complicated than necessary. But, your bolded statements are key (this is in reference to a non-PSLF scenario as that is my situation).

                                1)You'll get hit with a bit of a tax bomb following the conversion (so save accordingly for this event), but the $ in taxes owed following conversion at the end of your residency will be 'less' due to inflation making that money 'less valuable' than the money you saved in previous years on taxes (that money was worth more). ex: $10,000 in taxes 4 years ago was 'worth more' due to inflation than owing $10,000 in taxes today. Did that make sense? It is hard to word.

                                Anyone feel free to answer or contribute your thoughts thanks for the interesting discussion everyone
                                There is inflation but hopefully the money in your tIRA will have grown at a better rate than inflation. If your money has grown beyond inflation your tax bill will be higher even adjusted for inflation (assuming same tax rate).

                                Edit: just realized I didn't consider that you would have more growth (assuming you put the extra tax savings in the tIRA when contributing). So, if you are in the same tax bracket when you convert would be a wash. The potential downside would be if the conversion was large enough to push you to a higher tax bracket.

                                Last edited by Gamma Knives; 06-27-2021, 07:24 AM.

                                Comment

                                Working...
                                X