Roth conversions

[Corporateburnout]

A warning for using i-orp: put the same percent equities for all buckets, or it will preferentially spend out of the lower equity percent bucket and mess up the analysis.

But isn't it the function of I-ORP to tell me which account to spend from and how much to convert based on actual allocation?

 

[MrLoco]

8. Do you have reasons to not do Roth conversions?[/QUOTE]


Yes. Most definitely. Since there is a chance myself or my wife will need assistance later in life and this assistance remains tax deductible as a medical expense...we can offset taxes due from RMD's on a Trad. IRA with future medical/ assisted living/nursing home expenses as a medical deduction on Sch. A. If you convert all to a Roth....you may not be able to offset taxes due on RMD's. For 2018 and 2019 medical/dental expenses that exceed 7.5% of AGI will be tax deductible. After 2019 , medical expenses that exceed 10.0% of AGI will be tax deductible. Something to consider.

I think the best approach is to have some in taxable, some in tax deferred, and some in tax free. Then withdraw on a tax efficient basis for whatever needs may arise later in life.

 

[Hermit]

I'm left wondering how many of those reporting in this thread have run i-orp and if that agrees with their plan. A warning for using i-orp: put the same percent equities for all buckets, or it will preferentially spend out of the lower equity percent bucket and mess up the analysis.

i-orp recommended I transfer all assets to Roth at considerable tax ratios. I plan to do charitable distributions using QCDs after 70.5. I don't recall QCDs being considered by i-orp. I also don't think i-orp considers state taxes or Medicare premiums in its calculations both of which can be very big annual hits when they apply to you. Please let me know if I am wrong on these areas.

 

[whaleknives]

I'm left wondering how many of those reporting in this thread have run i-orp and if that agrees with their plan. A warning for using i-orp: put the same percent equities for all buckets, or it will preferentially spend out of the lower equity percent bucket and mess up the analysis.

The thing to remember about i-ORP is that it maximizes your spending or your final balance:

"If you fix your estate ORP will maximize your annual, after tax, disposable income. If you put an upper bound on your spending ORP will maximize your final total asset balance."
https://i-orp.com/GOPtax/faq.html
​

 

[FI_RElater]

...it’s called an “individual” retirement account

Thanks! I am "lucky" (ahem) to not have to do the SEPP, since I can withdraw from the spousal inherited IRA without penalty before 59.5.

That's why I was wondering if it's possible to maybe split the iIRA, and keep half of it as inherited and claim the other half as my own, and then start Roth conversions from there.

I also need to start spending more money.

Sorry, Googily

an IRA, even an inherited one, is an individual account...and you CANNOT split it

with your original info, I would max conversions from YOUR IRA up to the top of the (now) 12% limit, paying tax from your taxable account

since you have a while, continue doing so for the next years and don’t convert all of your IRA, as you will have 0 and 12% brackets to fill later
By not going too high, you can determine which years you need to “tax gain harvest” at 0 long term capital gains

obviously, take the surviving spouse SS at 60, and still convert to the top of the 12%. consider if it is best to continue doing Roth conversions at 67 and waiting until 70 for your SS to delay the higher tax limit... you’ll have to wait until you get there with your then current values to know.
(for us, I’ll be converting until the January after FRA and then starting SS , it’s almost break even and I’ve got the slightly lower PIA, and it leaves the max in the taxable accounts for whoever survives) for us, there’s no way we could possibly convert all the IRA’s without going way up the tax table, as our IRA’s are over a mil combined, but we also have more than that in taxable ...plus a pension... and hefty SS for both as well. couldn’t do conversions before retirement when in 33% bracket, as it made no sense.


(I wouldn’t even consider any future inherited IRA from parents, they might use them for long term care and not have tax consequences if enough costs above 10% MAGI)

 

[kaneohe]

Sorry, Googily

an IRA, even an inherited one, is an individual account...and you CANNOT split it

.....................................................................

A different interpretation of an inherited spousal IRA......appears that part can be retained as inherited and the rest can be made owned by surviving spouse...
https://www.bogleheads.org/forum/viewtopic.php?f=1&t=248259&newpost=3903813

 

[sengsational]

But isn't it the function of I-ORP to tell me which account to spend from and how much to convert based on actual allocation?

This is a general "beef" I have with i-orp, so I mention it whenever possible My premise is that asset allocation should be treated as liquid across account types, because, in most cases, it is. Of course there are some cases where an investment is "locked-in", like a rental property. But more often than not, you have choices in a Roth account, you have choices in a tIRA, and you have choices in after tax. i-orp models spending the bonds first, no matter which account the bonds are held in. Then, what happens in reality, if you follow the i-orp guidance along with maintaining your asset allocation, is that you spend the bonds, then rebalance across all three account types. The asset allocation in each account type changes over time, whereas in the i-orp model, those stay static and so it tends to suck the funds out of the one that happened to have the most bonds in it at the beginning. If you add up all assets across all three account types and rebalance irrespective of the account type, as I do, this results in a non-optimal solution. The "solution" is to lie to i-orp, telling it your asset allocation is exactly the same across all three account types, and so not giving it any excuse to drain one account type over the other, and let it base it's "decisions" upon the tax consequences.

i-orp recommended I transfer all assets to Roth at considerable tax ratios. I plan to do charitable distributions using QCDs after 70.5. I don't recall QCDs being considered by i-orp. I also don't think i-orp considers state taxes or Medicare premiums in its calculations both of which can be very big annual hits when they apply to you. Please let me know if I am wrong on these areas.

I also do not recall the charitable distributions being in the optimization. It does allow you to specify the tax rate in your retirement state, though.

 

[Brat]

I am going to throw another complexity into this decision making process: long term health care insurance payouts.

Our plan is "Non Qualified" and per-diem based. I am not sure exactly what this means but I am guessing that it means that there is a risk that benefits are not excluded from taxable income. Should, heaven forbid, one of us is using that benefit it would amount to $125,000 annually. DH and I are both in MRD status.

The more I can stash in our Roth IRAs the lower our MRDs at the end of life. We can always tap the Roth if need be.

 

[Hermit]

I am going to throw another complexity into this decision making process: long term health care insurance payouts.

Our plan is "Non Qualified" and per-diem based. I am not sure exactly what this means but I am guessing that it means that there is a risk that benefits are not excluded from taxable income. Should, heaven forbid, one of us is using that benefit it would amount to $125,000 annually. DH and I are both in MRD status.

The more I can stash in our Roth IRAs the lower our MRDs at the end of life. We can always tap the Roth if need be.

Another twist on long term care would be the tax deductability if you self insure. Are any of those costs tax deductable as medical costs? Would I be shooting myself in the foot to move all assets to Roth and later have very high medical/assisted care/nursing home care costs later? Isn't there a quote that the average spent on medical, when over 65, is north of $200K?

 

[Dtail]

Another twist on long term care would be the tax deductability if you self insure. Are any of those costs tax deductable as medical costs? Would I be shooting myself in the foot to move all assets to Roth and later have very high medical/assisted care/nursing home care costs later? Isn't there a quote that the average spent on medical, when over 65, is north of $200K?

From my understanding, LTC unreimbursed costs would be tax deductible above the 7.5% threshhold (next year 10%).
I believe Fidelity has quoted around 275k for medical costs over 65 yo.

 

[Brat]

I can see that we need a discussion of how LTC reimbursement of Home and Community Based Care is treated under the internal revenue code when a substantial portion of the services are not medical in nature.

 

[orplanner]

i-orp withdrawal strategy.

This is a general "beef" I have with i-orp, so I mention it whenever possible My premise is that asset allocation should be treated as liquid across account types, because, in most cases, it is. Of course there are some cases where an investment is "locked-in", like a rental property. But more often than not, you have choices in a Roth account, you have choices in a tIRA, and you have choices in after tax. i-orp models spending the bonds first, no matter which account the bonds are held in. Then, what happens in reality, if you follow the i-orp guidance along with maintaining your asset allocation, is that you spend the bonds, then rebalance across all three account types. The asset allocation in each account type changes over time, whereas in the i-orp model, those stay static and so it tends to suck the funds out of the one that happened to have the most bonds in it at the beginning. If you add up all assets across all three account types and rebalance irrespective of the account type, as I do, this results in a non-optimal solution. The "solution" is to lie to i-orp, telling it your asset allocation is exactly the same across all three account types, and so not giving it any excuse to drain one account type over the other, and let it base it's "decisions" upon the tax consequences.

I also do not recall the charitable distributions being in the optimization. It does allow you to specify the tax rate in your retirement state, though.

ORP treats all 3 accounts (IRA, Roth, taxable) as liquid but with only 3 options for moving money between accounts:

1. IRA to Roth Conversions
2. IRA to taxable distributions
3. Roth to taxable distributions (never done because there is no economic benefit).

All other permutations will bring a nasty-gram from the IRS. 2 of the 3 cases have tax consequences that affect optimization.


Generally speaking ORP does not distinguish between stocks and bonds for distributions. Generally but not always, ORP will distribute from the account with the lowest rate of return first, it doesn't matter what is in the account. So ORP will hang on to the Regan Era bonds in your taxable account paying 18% interest and distribute the Apple stock in your IRA first.



My preferred "solution" is to maintain the same asset allocation in all three accounts so that the distribution decisions are made more on tax economics and less on chasing yield. There are reasons for doing asset allocations in the three accounts differently but they are outside ORP's scope. It is the job of the analyst or advisor to manage the interface between ORP's model and the real world. i.e. your situation.


ORP offers a glide path option that let's you change asset allocations over retirement.



A retirement calculator that tightly controls asset allocations using rules other than economics is a simulator, not an optimizer. Its function can be performed much more simply with a spreadsheet.


This is the 2nd time that QCDs have come up, qualifying them for membership on ORP's wish list.


ORP considers Medicare Income Related Monthly Adjustment Amount to be on the expense side of the ledger having no impact on ORP's retirement income management. I refer you to ORP's FAQ.

 

[SevenUp]

ORP considers Medicare Income Related Monthly Adjustment Amount to be on the expense side of the ledger having no impact on ORP's retirement income management. I refer you to ORP's FAQ.

I guess that depends on the definition of "expense". E.g., one could say taxes are expenses and thus have ORP ignore them.

Just as with taxes, Medicare premiums increase as a function income. Is it the stepwise nature of IRMAA that prevents the ORP engine from including that functional behavior?

 

[Hermit]

ORP treats all 3 accounts (IRA, Roth, taxable) as liquid but with only 3 options for moving money between accounts:

1. IRA to Roth Conversions
2. IRA to taxable distributions
3. Roth to taxable distributions (never done because there is no economic benefit).

All other permutations will bring a nasty-gram from the IRS. 2 of the 3 cases have tax consequences that affect optimization.


Generally speaking ORP does not distinguish between stocks and bonds for distributions. Generally but not always, ORP will distribute from the account with the lowest rate of return first, it doesn't matter what is in the account. So ORP will hang on to the Regan Era bonds in your taxable account paying 18% interest and distribute the Apple stock in your IRA first.



My preferred "solution" is to maintain the same asset allocation in all three accounts so that the distribution decisions are made more on tax economics and less on chasing yield. There are reasons for doing asset allocations in the three accounts differently but they are outside ORP's scope. It is the job of the analyst or advisor to manage the interface between ORP's model and the real world. i.e. your situation.


ORP offers a glide path option that let's you change asset allocations over retirement.



A retirement calculator that tightly controls asset allocations using rules other than economics is a simulator, not an optimizer. Its function can be performed much more simply with a spreadsheet.


This is the 2nd time that QCDs have come up, qualifying them for membership on ORP's wish list.


ORP considers Medicare Income Related Monthly Adjustment Amount to be on the expense side of the ledger having no impact on ORP's retirement income management. I refer you to ORP's FAQ.

Thanks for the detailed info. That helps.

A QCD inclusion would help, but I think it might be complex. In my case, I plan to do QCDs up to a specified percentage of my portfolio and then maintain that percentage by tracking amount given through QCDs over the years. I guess you could either give the RMD amount or a percentage of tax deferred accounts up to $100k or something like that. In my case maybe be able to switch from the RMD amount to a percentage of tax deferred in a later year. Use of QCDs may not actually optimize tax planning in terms of having the largest portfolio possible, but it sure does optimize providing me with the largest portfolio when considering giving I have already committed to.

 

[sengsational]

All other permutations will bring a nasty-gram from the IRS. 2 of the 3 cases have tax consequences that affect optimization.

Right. I didn't mean actually moving money across those boundaries. Just selling one asset class and buying another within that particular taxable account type in order to put the asset allocation back on target.

Generally speaking ORP does not distinguish between stocks and bonds for distributions.

Yeah, I was a bit sloppy saying "sell bonds first". It's about the return.

My preferred "solution" is to maintain the same asset allocation in all three accounts so that the distribution decisions are made more on tax economics and less on chasing yield.

In a perfect world, yeah. But it would be a pain for me to do that "for real". But as I said, I think bending the truth to convince the optimization that all three have the same return is a good idea.

 

[Scuba]

8. Do you have reasons to not do Roth conversions?

Yes. Most definitely. Since there is a chance myself or my wife will need assistance later in life and this assistance remains tax deductible as a medical expense...we can offset taxes due from RMD's on a Trad. IRA with future medical/ assisted living/nursing home expenses as a medical deduction on Sch. A. If you convert all to a Roth....you may not be able to offset taxes due on RMD's. For 2018 and 2019 medical/dental expenses that exceed 7.5% of AGI will be tax deductible. After 2019 , medical expenses that exceed 10.0% of AGI will be tax deductible. Something to consider.

I think the best approach is to have some in taxable, some in tax deferred, and some in tax free. Then withdraw on a tax efficient basis for whatever needs may arise later in life.

This is my approach as well. So far, our tax accountant has advised that we not convert anything to Roth as we are still in a pretty high bracket between former employer deferred comp payouts, dividends and interest income. Our TIRA’s are only about 25% of our total portfolio. There may come a time that it will make sense for us to do some Roth conversions, but it hasn’t happened in our first 2 years of ER.

 

[orplanner]

I guess that depends on the definition of "expense". E.g., one could say taxes are expenses and thus have ORP ignore them.

Just as with taxes, Medicare premiums increase as a function income. Is it the stepwise nature of IRMAA that prevents the ORP engine from including that functional behavior?

You are correct to accuse me of having a somewhat arbitrary definition of spending. "Spending" is the wrong word. From the model's point of view anything that does not affect taxable income is not modeled. Medicare premiums don't change your AGI so they are not modeled. That is a decision on my part. I am open to arguments to the contrary. The step wise formulation is not the cause. (I think I could deal with the step wise IRMAA. It just doesn't fit in the income section of the model. )

 

[orplanner]

Right.

In a perfect world, yeah. But it would be a pain for me to do that "for real". But as I said, I think bending the truth to convince the optimization that all three have the same return is a good idea.

My dear boy, in the Operations Research World we do not bend the truth. We adapt the quantitative tool to fit the situation. In polite circles this is called "modeling".

 

[bw5972]

I've been thinking about doing conversions for several years now but this post finally motivated me to dig a little deeper and run some numbers using some online Roth conversion calculators. I entered my current marginal tax rate and expected future tax rate and was surprised to see that I am better off not converting. The incremental loss of ACA subsidy (approx. 10%) on top of my federal + State (CA) rate makes the conversion worse off than just staying put.

Ditto, I used the calculate at Schwab and get a result:

"What this means for you
Based on the information you provided, if maximizing the future value of your portfolio is your key goal, converting some or a portion of your assets to a Roth IRA may not make sense for you. The estimated future value of your converted assets is $105,088 less than if you left the assets in your Traditional IRA. This illustration is hypothetical in nature. Actual results will vary. See Calculator Assumptions for more information."

 

[Hermit]

You are correct to accuse me of having a somewhat arbitrary definition of spending. "Spending" is the wrong word. From the model's point of view anything that does not affect taxable income is not modeled. Medicare premiums don't change your AGI so they are not modeled. That is a decision on my part. I am open to arguments to the contrary. The step wise formulation is not the cause. (I think I could deal with the step wise IRMAA. It just doesn't fit in the income section of the model. )

I am concused with saying it does not affect taxable income. I would expect it to be handled as a tax driven soley by MAGI for everyone over 65. I agree that the base cost for Medicare is an expense. The premiums are simply a tax on higher income. It must be paid by everyone that meets the criteria.

Thanks so much for taking the time to discuss this with us.

 

[SevenUp]

You are correct to accuse me of having a somewhat arbitrary definition of spending. "Spending" is the wrong word. From the model's point of view anything that does not affect taxable income is not modeled. Medicare premiums don't change your AGI so they are not modeled. That is a decision on my part. I am open to arguments to the contrary. The step wise formulation is not the cause. (I think I could deal with the step wise IRMAA. It just doesn't fit in the income section of the model. )

I am concused with saying it does not affect taxable income. I would expect it to be handled as a tax driven soley by MAGI for everyone over 65. I agree that the base cost for Medicare is an expense. The premiums are simply a tax on higher income. It must be paid by everyone that meets the criteria.

Thanks so much for taking the time to discuss this with us.

Orplanner,

Yes, thank you!

Don't know if you have an ORP "wish list" somewhere, but IRMAA would be a nice addition.

There are a bunch of credits that can affect marginal rates but those may not be so applicable to retirement planning.

For taxes, besides IRMAA, there is Alternative Minimum Tax, Net Investment Income Tax, capital gain taxes, SS benefit taxation, etc., that might move the optimum strategy away from what "tax brackets only" would suggest.

With the new tax code, AMT is much less prevalent. Don't know how many people are affected by the others.