Part-Year Return Capital Loss Allocation

A taxpayer changes domicile from West Virginia to Missouri. There is a $24,374 carryover loss and a current year loss of $334 incurred during the portion of the year the taxpayer was a Missouri resident.

I am thinking that for federal purposes, there is a definite answer as to the amount of each loss that forms a part of the overall $3,000 loss allowed. The three possibilities as I see them are:

1. Recognize net losses chronologically; thus, the entire amount comes from the carryover loss.
2. Recognize current year net losses before applying carryover amounts; thus, the $334 is recognized first, then $2,666 of the carryover is recognized.
3. Combine all losses without regard to how or when they were incurred, then deduct a pro-rata share of all of them. Thus, $24,374/$24,708 x $3,000 = $2,959 of the carryover loss is recognized.

I believe that whichever is the correct answer would also control for state allocation purposes.

So, which is it?

I'm leaning towards option #3 because it would follow from the conclusion I came to in this post.

I believe the oldest are used first.

I believe that whichever is the correct answer would also control for state allocation purposes.

Why do you think that? I would look to Missouri law to see whether there's any loss to carry forward or if it's just $334. I'd assume not as many states do it that way. There isn't anything in federal regarding the vintage of a capital loss, it's just the excess from the prior-year return.

If at the federal level the taxpayer has both short term and long term capital loss carryovers the short term is applied first .

I agree that you need to look at Missouri law to see if you can carryover any of the loss incurred while a West Virginia resident to Missouri (I doubt it )

Why do you think that? I would look to Missouri law to see whether there's any loss to carry forward or if it's just $334.

I think that because all states, with the notable exception of Pennsylvania, derive their income tax laws from federal tax laws. This means that unless expressly provided to the contrary, items are recognized at the state level to the extent that they are recognized at the federal level.

This is true even in the case of a nonresident. See for example the following definitions of nonresident income:

The net amount of items of income, gain, loss and deduction entering into the nonresident’s federal adjusted gross income that are derived from or connected with sources in this state...
OR Rev Stat § 316.127(1)(a)

The net amount of items of income, gain, loss and deduction entering into his federal adjusted gross income, as defined in the laws of the United States for the taxable year, derived from or connected with New York sources...
NY Tax L § 631

The net amount of items of income, gain, loss, and deduction entering into his or her federal adjusted gross income which are derived from or connected with sources in this state...
MO Rev Stat § 143.181.1.(1)

Oregon, which doesn't allow carryovers of losses incurred while not a resident, uses a definition substantially similar to the definitions used by MO and NY. The latter two states don't have any provisions that undo carryover losses prior to a change in residency, which means the losses are recognized for state purposes when they are recognized for federal purposes.

The uniformity in wording between states means that even states that fiddle with carryover amounts don't take it upon themselves to increase the amount of a loss that a taxpayer can recognize in a given tax year. So, in my facts, if only a portion of the $334 loss is allowed for federal purposes, then only a portion will be allowed for Missouri purposes.

you need to look at Missouri law to see if you can carryover any of the loss incurred while a West Virginia resident to Missouri (I doubt it )

Most states don't have any specific provisions regarding carryovers, which means that for a resident, they get taken into account by default no matter what state they were incurred in.

There isn't anything in federal regarding the vintage of a capital loss, it's just the excess from the prior-year return.

I think I'll count this as a vote in favor of option #3. Noobie's opinion from post #2 differs, but I think he's only right in cases where carryover amounts expire (meaning, corporate capital losses, excess FTC, unused business credits, NOL's, and charitable deduction carryover amounts).

My vote is for something else entirely.

For West Virginia tax purposes, the capital loss deduction is $3,000. For Missouri tax purposes the capital loss deduction is $334. It doesn't matter that these don't add to $3,000.

So, which step in my reasoning to you disagree with: that states only allow losses to be recognized if they are also recognized for federal purposes, or that there is a definite amount of each loss recognized for federal purposes?

I disagree with "there is a definite amount of each loss recognized for federal purposes". The rule for carryovers is that the loss "shall be a [...]loss in the succeeding taxable year." IRC 1212(b)(1). On the 1040, there is simply a 24,708 loss, of which only 3,000 is allowed. IRC 1211. There is no provision for using up losses in any particular order, they're just combined in determining how much loss there was in the current year.

For WV tax purposes, on WV Form IT-140 Schedule A, there is a 24,374 WV-resident loss, of which only 3,000 is allowed under IRC 1211. Line 9, Column A: (3,000); Column B: (3,000).

For MO tax purposes, on Form MO-NRI, there is a 344 MO-resident loss, of which a maximum of 3,000 is allowed under IRC 1211, so instead 344 is allowed. Line G: (344).

As a separate example, if taxpayer had lost 3,000 in January while a WV resident and 3,000 in December while a MO resident, there would be a 3,000 loss on the federal return, and a 3,000 loss on both state returns.

There is no provision for using up losses in any particular order, they're just combined in determining how much loss there was in the current year.

Sure, they're combined. But the character of the carryover amount generated is determined in proportion with the character of the original losses that gave rise to the carryover amount. At the federal level, see 1.172-3(b) regarding business and nonbusiness capital losses. At the Missouri level, see 12 CSR 10-2.010(2) regarding losses allocated between a husband and wife.

Although my citations don't cover every aspect of the character of a capital loss carryover, they shouldn't have to. When we resolve ambiguities in the Code, we should do so in a way that avoids inconsistency with other established areas of the Code. This is why Missouri's regulation matches with the federal regulation. It applies pre-existing principles rather than establishing new principles.

When we look at the character of the remaining $21,708 loss going into the next year, Missouri would tell us that if the original $334 loss was from the husband and the original $24,708 was from the wife, then the husband should have $293 of loss remaining. If the $334 was a business loss and the carryover was nonbusiness, federal regulations would tell us that $293 of the loss carryover should be counted as a business loss for NOL purposes. What reason is there to break with these approaches and say that the entire $334 loss was recognized in the first year?

Why do you think that? I would look to Missouri law to see whether there's any loss to carry forward or if it's just $334.

I think that because all states, with the notable exception of Pennsylvania, derive their income tax laws from federal tax laws.

Derive, sure, but with the potential for significant differences. HSAs don't exist in CA, there's no carry-forward of capital losses in NJ (no matter where you lived when they were realized), etc. I don't know anything about MO. You have to check this stuff state by state and a common construct is "carry forward the loss from last year's this-state return, and if none was filed, the carry-forward amount is $0."

§11-21-44. West Virginia source income of part-year resident individuals.
(a) Individuals. -- The West Virginia source income of a part-year resident individual shall be the sum of the following:

(1) Federal adjusted gross income for the period of residence, computed as if his or her taxable year for federal income tax purposes were limited to the period of residence.

(2) West Virginia source income for the period of nonresidence determined in accordance with section thirty-two of this article as if his or her taxable year for federal income tax purposes were limited to the period of nonresidence.

[emph added]

1. An individual who is a resident for only part of his taxable period shall be treated as a nonresident for purposes of sections 143.011 to 143.996 .  His Missouri nonresident adjusted gross income (Missouri adjusted gross income from sources within this state) shall consist of

(1) All items that would have determined his Missouri adjusted gross income if he had a taxable period as a resident consisting solely of the time he was a resident, and

(2) All items that would have determined his Missouri nonresident adjusted gross income if he had a taxable period as a nonresident consisting solely of the time he was not a resident.

[emph added]

This means to me that you have to (conceptually) re-do the 1040 for each state, ignoring the nonresident portion of each return. What would the 1040 look like if we ignored the dates of Missouri residence? There would be a capital loss of 24,374, of which 3,000 is allowed. That is how the WV-PY-resident income is computed. What would the 1040 look like if we ignored the dates of West Virginia residence? There would be a capital loss of 344, all of which is allowed. That is how the MO-PY-resident income is computed. Yes, it may seem odd that more than 3,000 is deducted in total on the two state returns, but that is how it works.

Another example, suppose taxpayer had a 344 capital gain after changing residency. Would you then expect to be able to deduct 3,344 on the WV return?

you have to (conceptually) re-do the 1040 for each state, ignoring the nonresident portion of each return.

I agree with you here. By structuring the laws the way that they have, states have essentially created a scheme of subordinate taxable periods. When calculating the income for each taxable period, all relevant federal laws apply.

Even so, it seems dubious that the state-level taxable periods would have the effect of causing items of income, loss, etc. not recognized at the federal level to become recognized for state purposes. When dealing with short periods, no federal provision is prorated unless expressly provided for. See Gregg v. United States. So, if we really decided to re-do everything for just the short year period, all sorts of things would happen. Additional passive losses could easily become disallowed or allowed based on fluctuations in AGI and the amount of time a taxpayer has to meet material participation tests in the short year. A different amount of net operating loss might be recognized for state purposes. Both of these results would give rise to state-federal differences that would need to be tracked and might take years to get back in sync.

Most states, including West Virginia and Missouri, do not have any provisions that contemplate anything like that. They don't even have any specific provisions regarding carryover amounts at all. As a result, the states simply recognize the amounts when they are recognized for federal purposes. They try to make things simple all around. Decoupling provisions related to things like bonus depreciation are generally applied in the current year as opposed to being added to or subtracted from a suspended passive loss.

The mere existence of two state-level subordinate return periods for a single year should not change anything. Instead, the amounts to be recognized in both subordinate return periods should be decided at the federal level first and then allocated to the different periods by applying the law as if those were the only amounts to be considered.

The application of this principle to pass-through income results in an allocation between the two short-year periods. This is due to the federal treatment of the income as constructively received by the owners of the entity at the time it is received by the entity. If we completely redid the 1040 suggest, then all of the income would be recognized in whichever state's period the last day of the entity's taxable year fell into. Illinois does it this way, but I can name seven other states that specifically require an allocation between the two periods.

Applying the principle to your example of the $344 capital gain: we have a total of $3,344 in capital losses to be taken into account in the West Virginia short year period. Because all relevant federal laws apply, $3,000 of the loss is recognized and $344 is carried forward to the Missouri period. The $344 loss is then netted with the $344 gain for an allocation of $0 to Missouri.

I do understand that some states actually provide for state-federal differences in carryover amounts. But these are specific provisions, and here we are talking about what happens with states that follow federal law without such provisions in place.