Trad to Roth conversion - more or less

[eyesonthehorizon]

Yeah, the difference is if it's within the same institution vs going between companies. Internal movements at one company (or those between companies but where the account title isn't changing) usually can be done in-kind.

Some people don't want to pull more out of their 401ks than needed because their state leaves IRAs open to creditors.

[Mr. Green]

I had a bit of an epiphany this morning and I'm paging @boarder42 for his thoughts since he has also spent quite a bit of time thinking about this process.

With the market poised for a considerable drop I have been doing some thinking about whether it would be advantageous to max out Roth conversions when the market is down since it may well be the last year that has no income cap for ACA subsidies. For me personally that still limits the maximum theoretical conversion to about $100,000 because I would never want to push into the 22% ordinary income tax bracket under voluntary circumstances.

My initial thinking was that I would set some type of threshold for a market drop (say 30%) after which I would automatically make the maximum conversion that I was considering for the year. The thinking being that the same number of absolute dollars would represent a larger percentage of my portfolio,  thus getting as much of our stash as possible into accounts that will no longer be taxed.

However I ultimately realized there is a fairly large flaw in this thinking under certain scenarios, and this flaw extends so far as to influencing whether or not I would contribute lots of money to a Roth IRA early in my career if I had the choice or not. My focus with my FIRE spreadsheet projections has been about how one can maximize their yearly spending over their lifetime in a bell-shaped curve while also minimizing taxes through long-term tax smoothing. In the posts I've made upthread I have commented on how for the very early retiree (think < age 40-45) the biggest hurdle to increasing spending actually becomes available Roth IRA principal prior to age 60.

The critical flaw here is that increased Roth conversions over decades require a large traditional IRA balance and depleting that balance more quickly reduces the amount of funds available to convert (duh). At a certain point, taking advantage of the ACA subsidy cap removal and converting a larger amount now could actually reduce our maximum annual spending via a Roth conversion pipeline. I suppose if we got close enough to traditional retirement age the conversation could then turn to a 72t SEPP to bridge that gap but I am still not a fan of that particular method.

As this dawned on me, I realized that it is actually quite a good thing that the majority of our money is in traditional IRAs right now because that high balance will fuel large Roth conversions for us for the next two decades. Had we instead made significant Roth IRA contributions in place of tIRA contributions when we were younger and hadn't yet passed income thresholds that allowed that, we would have significantly less Roth principle to spend over the next 20 or so years. The majority of our money would be Roth gains and not subject to withdrawal without penalty prior to age 60.

It is possible that this is merely a self-reinforcing situation in the sense that those that have typically been able to retire before age 40 have had exceptionally high incomes which would disallow Roth contributions anyway. However I know it was the case for us that my income (and our stash) only grew astronomically in the last 5 years or so before we FIREd. Had I known, we could have been making Roth IRA contributions (instead of tIRA) during most of my twenties. Looking at it now, if trying to maximize our long-term annual spending, I suspect that having made Roth contributions all those years would have actually left us with less flexibility than we have now.

[dandarc]

"ACA subsidy cap removal" - did they actually make that permanent? Last I knew, that was 2021 and 2022 only, but making it permanent was included in the Build Back Better bill that failed. Really hope I missed something related to that though.

I think the strategy to maximize Roth conversions during big downswings might work well on the "minimizing taxes" goal. In terms of maximizing available to spend across lifetime it is less clear to me. Overall I think you're right about conversions (if you can weather the 5 year holding period after conversions) allow for much more Roth available to withdraw than sum of contributions. So chalk another up to "Roth sucks", at least during accumulation years.

[Mr. Green]

@dandarc the cap removal expires after this year which is why I was initially thinking about maximizing a conversion if we see a big market crash.

[secondcor521]

There are three separate things here.

First, yes, it is generally advantageous to do a Roth conversion when the market is "low" during the year because you get more shares converted for the same tax cost.  Or alternatively, you get the same amount of shares converted for a lower tax cost.  I saw a YouTube video once that claimed that converting half of the total on 1/1, then the other half when the market was down 20% during the year, else the other half at the end of the year was historically a good approach.  Personally I just wait to see if I'm getting scared or people are grumbling or afraid and then do maybe half of what I think I'll do for the year.  Then I fine tune in December to hit my AGI target.  It may not be totally optimal but I place a fairly high value on the certainty of hitting my AGI target (and thus the tax structure I want for the year).

Second, how much to convert is always an interesting and tricky question, and one must consider all of the factors.  Of which there are quite a few, and certainly the ACA PTC currently-temporary cap removal is one that would apply to people considering Roth conversions around that 400% FPL mark.  Overall, though, the best metric I've been able to come up with after a lot of study is comparison of marginal rates now to probable marginal rates when I'm 75 or so.  And being 52-almost-53, I don't think it's feasible to construct an accurate enough model to get things to within closer than a 5% point range.  So in the end one might feel comfortable paying taxes at 20% but not at 25% or something like that.

Third, as far as Roth contributions vs. traditional, I agree with the assessment that for early retirees, traditional contributions then Roth conversion ladder is probably the better way to go in terms of spending availability in one's 50s.  My three 20-something offspring are all doing Roth contributions now because they're in low brackets.  I don't think any of my three kids will FIRE anyway for various reasons, so perhaps it's not a problem.

[Mr. Green]

I have been marinating on this for the last few days based on @secondcor521's comment and I think I'm overcomplicating the issue based on the assumed complexity of the situation.

Choosing not to take advantage of converting while the market is down leaves gains from the next market recovery in our tIRA accounts. We can reasonably expect to pay at least ~20% tax on that money. 10% is the lowest Federal tax bracket plus 5% North Carolina state taxes. It's more likely that money would end up in the 12% Federal tax bracket, which I expect to revert to 15% after the next few years. Plus health insurance "taxes" via reduced subsidies on higher income. Regardless, I can't really think of a scenario where the combined taxes are less than 10% which is the only "tax" we'd pay for taking Roth gains before age 59.5 via penalty. It's probably really as simple as that.

Now that I  think about it, I suspect if I were to retool my model to plan for taking Roth gains prior to age 59.5 we could both increase our maximum annual lifetime spending and lower total overall taxes. I think a version 3.0 is needed.

This would also change my opinion re: Roth contributions early in my career, where once again we should have been contributing the max any time we were able.

Do I have that right?

[seattlecyclone]

Regardless, I can't really think of a scenario where the combined taxes are less than 10% which is the only "tax" we'd pay for taking Roth gains before age 59.5 via penalty.

Roth gains removed prior to 59½ count as regular income toward your AGI and are taxed accordingly, plus the 10% early withdrawal tax.

[Mr. Green]

Regardless, I can't really think of a scenario where the combined taxes are less than 10% which is the only "tax" we'd pay for taking Roth gains before age 59.5 via penalty.

Roth gains removed prior to 59½ count as regular income toward your AGI and are taxed accordingly, plus the 10% early withdrawal tax.

Bah, I thought that was kicking around in the back of my brain somewhere but a quick Google search came up empty so I figured I was confusing it with something else. The marination shall continue!

[elysianfields]

Back on topic, I'd like to add my thanks to everyone else's, the value in this thread definitely provides great ROI on the cost of several years' MMM forum membership...

@Mr. Green, thanks for the great spreadsheet.  I've needed to start modeling our income & tax projections, RMDs, Roth conversion options, etc. for a while now, and your heavy lift will make that easier.  Our situation differs, so I'm adapting it to our needs.

As a US Foreign Service employee hanging on for MRA, like @MasterStache I'll receive a pension.  Since I'll retire before age 62 and therefore won't have reached SS eligibility, Uncle Sugar adds an annuity supplement for the period between my retirement and age 62 (aside: some FS folks opine that continuing to work after reaching retirement eligibility makes little sense because you never have another chance at that annuity supplement).  Uncle also contributes to FEHB in retirement at the same rate as for current employees, and my contribution share also remains the same, albeit with post-tax dollars upon retirement (while working, it's pre-tax).  My DW will also receive a small pension.

Thus I can discard with calculations of ACA eligibility, and perhaps post-retirement Medicare part B calculations (I'm not sure whether it makes sense to opt in to Medicare part B - another topic to model and analyze).

OTOH, the income from the pensions (and the annuity supplement in the first few years) will make it difficult to squeeze some tax-free or low-tax-rate Roth conversions into the pipeline.  We've both contributed heavily to taxable TSP and other retirement accounts, and are continuing to contribute to taxable through 2022, our last year to qualify for the AOTC, and we need to keep reducing AGI to maintain eligibility.

I'm using the model to decide whether to shift our TSP contributions to Roth in 2023 and beyond, because that would likely vault us into the 24% marginal tax bracket, keeping in mind that we're among those high-saving folks who will face high RMDs on our traditional funds and the accompanying federal taxes, with little room to convert at lower rates due to our pensions (boo-hoo, I know).

Also, assuming we maintain our health, we'll have so much income that taking SS before age 70 will simply increase our taxes further for little benefit (more tears fall into my beer...).

For ill or good we started contributing to Roth IRAs when they first became available in 1998 - our incomes were much lower then - and have simply continued to make IRA contributions only to Roth since then.  We may have missed out on some tax breaks in the past, and nevertheless have built up large Roth balances rather than adding to our RMD/tax problem.

Upon retirement, we plan to transfer our Roth TSP funds to Roth IRAs, because according to the document Important Tax Information About Your TSP Withdrawal and Required Minimum Distributions,

The Internal Revenue Code (IRC) requires that you begin receiving distributions from your account in the calendar year you become age 72 and are separated from federal service. Your entire TSP account—both traditional and Roth—is subject to these required minimum distributions (RMDs).

I'm definitely down with @ender 's idea of doing a cash-out refi just before bailing out - get the mortgage while we're still receiving paychecks.

I'm following up on this thread in case someone reads this without knowing about the changes resulting from the SECURE Act 2.0.  Kitces has a long read on it here:  https://www.kitces.com/blog/secure-act-2-omnibus-2022-hr-2954-rmd-75-529-roth-rollover-increase-qcd-student-loan-match/

The short version:

The age for RMDs was raised - in our cases to 75 - giving us a few more years to try to squeeze in Roth conversions.

Roth TSP funds are no longer subject to RMDs in retirement, so we'll probably leave Roth funds in the TSP after retirement.

Employers can consider the option of having matching funds go to a Roth 401(k), though I haven't heard whether TSP will do so.

On a personal note, we realized that contributing to Roth TSP didn't make sense because it would vault us into the 24% Federal tax bracket, and we also pay State income tax on top.  So we've kept with Traditional despite our high balances in Traditional TSPs & IRAs.  We probably wouldn't consider putting our matching funds into Roth TSP.

We're still not investing in the G Fund, and we probably cannot access it currently anyway due to the Department of the Treasury's "extraordinary measures" to keep from breaching the debt limit.  We continue to eschew bonds in our investment allocation.

However, SECURE 2.0 also requires high wage earners to place their catch-up contributions in Roth starting in 2024.  Since we're taxed whether we place those funds into Roth TSP or into a taxable account, it makes sense to consider whether to continue making catch-up contributions in 2024 and beyond.

We wouldn't lose any matching on catch-up contributions, so that's not a consideration.

TSP funds have extremely low costs, so if we'd invest those contributions anyway, that's an advantage for continuing catch-up contributions.  Also, there really isn't an access problem, since we'll both be over 55 when we separate from service and can therefore access our TSP accounts at any time without penalty.

If we decide to deleverage by paying down our HELOC, I suppose we might then eschew catch-up contributions to divert more funds to loan repayment.  We'll see what happens with interest rates; our HELOC currently costs us 7.25%, is only fractionally tax-deductible, and that rate is close to the long-term stock market return.

So it would seem that the trade-off is use-or-lose access to the low-cost TSP funds vs. using those taxed funds for HELOC repayment.

We'll see where things stand in 2024, I suppose.  And we'll continue to consider a cash-out refi just before bailing.

Edited to fix quoting.

[Mr. Green]

I had someone pop into my Dying With Zero journal asking for my modeling spreadsheet. It made me think I should post a link to that journal here because in our case saying with zero is heavily reliant on the Roth conversion pipeline and there is some interesting commentary there both about the effect of buying and selling real estate and the market drop over the last year. Since I'll continue updating that journal with interesting data points I find it may be useful in providing additional thinking points to others in their own Roth conversion journeys.