Tax Notes logo

The Deductibility of Capital Losses in a Mark-to-Market Regime

Posted on Sep. 20, 2021
[Editor's Note:

This article originally appeared in the September 20, 2021, issue of Tax Notes Federal.


Lawrence Zelenak is the Pamela B. Gann Professor at Duke Law School.

In this article, Zelenak considers the extent to which capital losses should be deductible under a mark-to-market regime applicable to the tradable assets of wealthy taxpayers, as advocated by Senate Finance Committee Chair Ron Wyden, D-Ore., in a 2019 paper.

The Wyden Proposal and Losses

In 2019 Sen. Ron Wyden, D-Ore., then ranking member and now chair of the Senate Finance Committee, released a paper advocating an end to capital gains deferral for wealthy taxpayers.1 In his paper, Wyden proposes anti-deferral accounting rules that would apply to high-income taxpayers, defined as taxpayers who in each of the three preceding years had at least $1 million of income, $10 million of applicable assets, or both.2 For those taxpayers, the anti-deferral rules would annually mark to market gains and losses in tradable property.3 For non-tradable assets, realization of gains and losses would continue to be deferred until sale or other disposition, but upon realization of gain a lookback rule would impose a deferral charge, with the charge designed to approximately offset the tax advantage of deferral (relative to a mark-to-market regime).4 In addition to the anti-deferral regime for high-income taxpayers, Wyden proposes elimination of the long-term capital gains rate preference for all taxpayers.5 Wyden does not purport to offer a ready-to-be-enacted proposal. Instead, he describes the basic idea in broad outline and solicits comments on no fewer than 49 statutory design issues, including how to prevent taxpayers from gaming the $1 million and $10 million thresholds, rules for valuing assets for purposes of the $10 million threshold, whether debt should reduce assets for purposes of the $10 million threshold, the treatment of mark-to-market losses in tradable property, the design of the lookback rules for non-tradable property, phase-in and transition rules, special treatment of some asset types (residences, farms, and tax-favored retirement savings), application of the anti-deferral regime to passthrough entities and C corporations, and information reporting.6

In this article, I limit my comments to the treatment of mark-to-market losses under the Wyden proposal. The proposal itself has little to say on this crucial topic. Apart from soliciting comments, Wyden says only that mark-to-market losses on tradable property would generally be deductible, but that “losses would be subject to certain [unspecified] limitations to prevent abuse.”7 He requests comments on two loss-related issues: (1) what limitations should be placed on mark-to-market losses, and (2) whether mark-to-market losses should be deductible against ordinary income.8

Under any income tax that takes into account capital gains and losses, whether the tax is based on traditional realization events or on mark-to-market principles, capital losses are noncontroversially available to offset capital gains of the same tax accounting period. The two thorny questions are to what extent (if at all) capital losses should be allowed to offset ordinary income, and to what extent (if at all) capital losses in excess of offsettable current-year income should carry back or forward to offset income (capital gains, ordinary income, or both) of earlier or later years. Although Wyden requests comments on the ordinary income offset question, he does not mention (let alone request comments on) the loss carryover question.9 This is surprising, considering that the carryover question is among the handful of most important mark-to-market design issues; it dwarfs in significance most of the 49 topics on which Wyden does request comments.

In this article, I offer my suggestions (solicited or not) about whether, under a mark-to-market regime along the lines outlined by Wyden, mark-to-market capital losses should be deductible against same-year ordinary income, and whether (and to what extent) those losses should be deductible against capital losses or ordinary income of earlier or later years. Not to keep the reader in suspense, I recommend that (1) mark-to-market losses should be fully deductible against all types of current-year income (mark-to-market gains, traditionally realized gains,10 and ordinary income); and (2) mark-to-market losses in excess of offsettable current-year income should not be available for carryback to earlier years, but should carry forward for an unlimited number of years (until exhausted) and should be allowed as an offset against all types of future-year income.

Recognizing that the proper treatment of capital losses is a challenging question under either a realization-based or a mark-to-market regime, and that the nation has had over a century of experience with capital losses under a realization-based regime, which can and should inform the treatment of capital losses under a mark-to-market regime, the next section of this article reviews the legislative development of rules governing the deductibility of capital losses during the first three decades (1913 to 1942) of the modern federal income tax. Drawing and applying lessons from that history, I make the case for the treatment of capital losses described above before concluding my analysis.

The History of Capital Loss Rules

Under long-standing law, capital losses realized by individuals can offset without limitation capital gains realized in the same year; also, capital losses in excess of capital gains can offset up to $3,000 of ordinary income.11 Capital losses not deductible under these rules cannot be carried back to earlier years, but may be carried forward for an unlimited number of years (until fully used).12 Congress enacted capital loss disallowance rules similar to the current regime in 1942. Under the 1942 rules, capital losses were deductible against capital gains and $1,000 of ordinary income, with a five-year carryforward of disallowed losses.13 Except for a modest increase in the ordinary income amount (far too modest to keep pace with inflation) and the 1964 replacement of the five-year carryforward with an unlimited carryforward,14 the capital loss rules have remained essentially the same for nearly eight decades.

This post-1942 stability contrasts sharply with the instability of the capital loss rules during the first three decades of the federal income tax, from 1913 to 1942. The following brief history of those first three decades is offered in the hope that it will illuminate the policy analysis of capital loss limitations under a mark-to-market regime, keeping in mind that the difference between a realization-based system and a mark-to-market system has major implications for the design of capital loss limitations.15

Capital losses were not deductible at all, not even against capital gains realized in the same year, under the 1913 income tax. The 1913 Congress did not mull over the capital loss question and decide on total disallowance. Rather, it failed to think about the question at all, with the result that the statute said nothing one way or the other about capital losses.16 In the absence of legislative authorization of a capital loss deduction, Treasury ruled that no deduction was permitted.17

By 1916 Congress was aware of the capital loss question and addressed it with a provision allowing capital losses to offset capital gains of the same year (but without deductibility against ordinary income or income of other years).18 In 1918, in reaction to the great increase in marginal tax rates necessitated by the war effort, wealthy taxpayers and their advocates urged Congress to allow capital losses without limitation against ordinary income of the same year, and Congress agreed (albeit still without any carryover provisions).19 Thus, in the span of a mere five years, capital losses had gone from not deductible at all, to deductible only against capital gains, to fully deductible against all types of income.

But even more favorable treatment soon arrived. Before 1921 capital gains were taxed at the same rates as ordinary income. The Revenue Act of 1921 introduced a dramatic rate preference for capital gains — a maximum rate of 12.5 percent, compared with a top rate of 58 percent on ordinary income.20 This innovation gave rise to a new question about the deductibility of capital losses against ordinary income (which the 1921 act continued to permit). If a taxpayer had, say, $100,000 of ordinary income in the 58 percent bracket, and a $100,000 capital loss (in excess of capital gain, if any), should the taxpayer be allowed to fully offset the ordinary income with the capital loss, thus eliminating $58,000 of tax liability? Or should the law impose rate symmetry on capital gains and losses, so that just as a $100,000 capital gain would increase tax liability by only $12,500, a $100,000 capital loss would decrease tax liability by only $12,500?21 Congress generously opted for the former, extremely taxpayer-favorable approach.22 Thus, for a top-bracket taxpayer, a net capital gain was taxed at only 12.5 percent, but a net capital loss was deductible at 58 percent.

This too-good-to-be-true treatment was not destined to last. By 1924 both progressives in the House and the decidedly unprogressive Treasury Secretary Andrew W. Mellon, despite their sharp disagreements on other tax policy issues, came out in favor of rate symmetry for capital gains and losses; the Revenue Act of 1924 continued the deductibility of capital losses against ordinary income, but with the proviso that the tax savings from a net capital loss could not exceed 12.5 percent of the amount of the loss.23 So, for example, if a taxpayer had $100,000 of ordinary income taxed at a marginal rate of 50 percent, it would take $400,000 of net capital loss to eliminate the $50,000 tax on the ordinary income.

The 1924 capital loss regime remained in place until 1932; subject to the rate symmetry requirement, capital losses could offset unlimited amounts of ordinary income. The unlimited deductibility of capital losses posed no great threat to the revenue-raising capacity of the income tax as long as the stock market remained bullish. After the stock market crash of 1929, however, unlimited deductibility of capital losses resulted in massive reductions in income tax revenues, including, of course, revenues from the taxation of ordinary income. Congress responded (finally) to these changed circumstances in 1932 by drawing a distinction between short-term and long-term capital losses (with the holding period dividing line at two years).24 For losses on assets held by taxpayers for over two years, the 1932 act continued to permit unlimited deductibility against ordinary income (subject to rate symmetry). But for losses on assets held for two years or less, the act permitted deduction against only short-term capital gain.

Thus, under the 1932 act short-term losses were deductible against neither long-term gains nor ordinary income. Still, the act for the first time allowed taxpayers to use capital losses realized in one year against capital gains of another year; in particular, the act allowed a one-year carryforward of disallowed short-term losses. The next year, however, Congress repealed the carryforward provision before it otherwise would have become effective.25

In the run-up to what became the Revenue Act of 1934, Henry Morgenthau Jr., then the acting Treasury secretary (and soon to become the secretary), commented that, even with the 1932 act’s treatment of short-term losses, “no method of tax avoidance is more widely employed” than deducting capital losses against ordinary income, and that it would be “fair to provide that capital losses can be deducted only against capital gains.”26 Elaborating on this point, Morgenthau emphasized taxpayers’ ability to choose the timing of their realizations of capital gains and losses: “Since the taxpayer may himself determine the time at which a sale of capital assets shall be made and therefore has considerable leeway in determining when he will be taxable on any gains or when he will deduct his losses, such limitations as these are not unreasonable.”27 In the end, after a rather convoluted legislative process, the Revenue Act of 1934 permitted taxpayers to use capital losses to offset only $2,000 of ordinary income (regardless of the holding periods of the loss assets).28

Except for the ability to deduct capital losses against a limited amount of ordinary income, the 1934 act constituted a return to the approach of the 1916 legislation. But this, too, was not fated to endure. Concerned about taxpayer-investors who had suffered major losses in the stock market debacle of 1937, Congress in 1938 reinstated unlimited deductibility of long-term capital losses against ordinary income (subject to a rate-symmetry rule).29 As for short-term capital losses, they could be deducted only against short-term gains, with a one-year carryforward.30

The three-decades-long legislative wandering in the capital loss tax policy wilderness finally ended in 1942, although that did not become clear until later. In that year Randolph Paul, Treasury’s leading tax policy expert, explained to the House Ways and Means Committee that “the present privilege of deducting [long-term] capital losses from ordinary income has under recent rate increases encouraged an unusually large amount of capital loss realization.”31 Agreeing with Paul and Treasury, Congress legislated deductibility of net capital loss (whether short-term or long-term) against only $1,000 of ordinary income, with a five-year carryforward of disallowed losses.32

After having experimented for three decades with nearly every conceivable approach to the deductibility of capital losses, in 1942 Congress finally settled on an approach that has now endured in broad outline for almost eight decades. Under that approach, (1) capital losses are always available to offset capital gains realized in the same year; (2) net capital losses are permitted to offset only a limited amount of same-year ordinary income; (3) carryback of net capital losses is not permitted; and (4) extensive carryforward of net capital losses against capital gains is permitted, along with limited carryforward of net capital losses against ordinary income.

So much for the “what” of the historical survey. The next section shifts the focus from descriptive to normative. It examines the “why” of the legislative treatment of capital losses under a realization-based income tax and considers what modifications of that treatment might be appropriate under a mark-to-market regime.

Losses in a Mark-to-Market System

In deciding on the appropriate treatment of capital losses under the Wyden mark-to-market proposal, we can set aside the rate symmetry question that so bedeviled Congress in the early years of the income tax. With capital gains taxed under the proposal at the same rates as ordinary income, the rate symmetry problem resolves itself. We can turn, then, to questions of the deductibility of capital losses against ordinary income and against income of earlier and later years.

In a realization-based system, taxpayers have a timing option for capital gains and losses.33 They can cherry-pick by selectively realizing capital losses while continuing to hold their gains as unrealized appreciation. If capital losses were deductible without limitation against ordinary income, this strategy would be highly effective for taxpayers and extremely costly to the fisc. Realized capital losses, although genuine enough for the particular assets sold, are often artificial in a broader sense because they are offset (or more than offset) by unrealized gains in unsold assets. This cherry-picking concern is one of the two primary reasons why Congress has, for many decades, severely limited taxpayers’ ability to deduct capital losses against ordinary income. The second reason is, simply enough, revenue; the income tax raises most of its revenue from the taxation of ordinary income, and allowing unlimited deductibility of capital losses against ordinary income would impair the revenue-raising capacity of the tax.

Interestingly, the ups and downs of the stock market have mirror-image effects on the significance of the cherry-picking and revenue concerns. On one hand, a down market makes it more likely that a taxpayer’s realized losses are not offset by unrealized gains, thus strengthening the case for allowing capital losses against ordinary income. On the other hand, a down market increases the amount of capital losses that a taxpayer could choose to realize, thus increasing the revenue threat from unlimited deductibility. During the early years of the income tax, these countervailing considerations resulted in inconsistent legislative responses to different stock market downturns. When stocks traded on the New York Stock Exchange lost almost a quarter of their value in 1917, Congress responded in 1918 by permitting unlimited deductibility of capital losses against ordinary income.34 But Congress’s 1932 response to the Great Depression was the opposite — the introduction of major new capital loss limitations in both 1932 and 1934. In 1938, however, Congress responded to the stock market crash of 1937 by reintroducing the unlimited deductibility of long-term capital gains against ordinary income.

The two crucial considerations in setting capital loss policy are differently affected by a move from realization-based to mark-to-market taxation. Cherry-picking of losses is impossible, by definition, in a mark-to-market regime. Thus, one of the two justifications for capital loss limitations vanishes. Revenue concerns, however, may actually increase. Under a mark-to-market system with unlimited deductibility of capital losses against ordinary income, in a down-market year the income tax may produce little or no revenue from the ordinary income of wealthy taxpayers subject to the mark-to-market rules.

Keeping in mind, then, both the historical development of the capital loss rules and how capital loss policy considerations are affected by a shift from realization to a mark-to-market system, what should be the capital loss rules under the Wyden proposal? On one hand, the anti-cherry-picking rationale for limiting capital loss deductions against ordinary income does not apply under a mark-to-market regime.35 On the other hand, revenue loss concerns not only remain but are enhanced because all losses in tradable assets are automatically realized every year; taxpayers do not have to go to the trouble of selling assets to realize their losses.36

Start by considering a possible rule — motivated by the disappearance of the cherry-picking concern — that mark-to-market losses are fully deductible against all types of current-year income. One can imagine a sharply down-market year in which mark-to-market net losses of taxpayers subject to the anti-deferral rules would be great enough to eliminate all the tax those taxpayers would otherwise owe on their ordinary income. If that specter loomed large in the legislative mind, Congress could subject mark-to-market losses to limitations analogous to the charitable deduction limitations.37 Mark-to-market losses might be allowed to offset, for example, no more than 50 percent of a taxpayer’s ordinary income, with multiyear carryforward of losses disallowed under that rule.

However, revenue concerns standing alone — without reinforcement by anti-cherry-picking concerns — might not be sufficient justification for limiting the deductibility of mark-to-market losses against ordinary income. If the mark-to-market system applied only to the fabulously wealthy (rather than also to the merely wealthy, as Wyden proposes), little revenue would be at stake in the ordinary income offset question. ProPublica recently revealed that wages constituted a mere 1.1 percent of the income reported on the 2018 tax returns of the 25 wealthiest Americans.38 The smaller the sliver of the wealth distribution to which the mark-to-market regime applies, the weaker the revenue-based argument for limiting the deductibility of mark-to-market losses against ordinary income. But even if the coverage of the anti-deferral regime were broad enough to threaten major revenue losses from the offsetting of ordinary income in down-market years, the government could and should deal with the down-year revenue loss by basing spending on average annual revenues (reflecting both high tax collections in up-market years and low collections in down-market years), rather than by enacting loss limitations unsupported by any anti-cherry-picking rationale.

What about carryovers of mark-to-market losses in excess of current-year income?39 Unlimited carryforward of realized capital losses has been the rule since 1964, and there is no reason the rule should be different for mark-to-market losses. Thus, mark-to-market losses should be allowed without limitation against both mark-to-market and traditionally realized capital gains of later years, and against ordinary income of later years to the same extent those losses are allowed against current-year ordinary income.40

The trickier question is whether mark-to-market losses should be available for carryback to earlier years. Capital loss carrybacks for individuals are conspicuously absent from the federal income tax’s history of capital loss rules, even though corporations have long been permitted a three-year capital loss carryback,41 and net operating loss carrybacks have often been permitted for both individual and corporate taxpayers (with the original NOL carryback going back all the way to 1918).42

Under a realization-based income tax, the timing option for capital gains and losses is a reasonably satisfactory substitute for loss carrybacks. Consider a taxpayer who, in a realization-based tax regime with loss carryforwards but not carrybacks (in other words, the current system), has a $1 million capital loss in year 1 and a $1 million capital gain in year 2. The year 1 loss carries forward to year 2, and the taxpayer (appropriately) pays no tax in either year. But if the timing is reversed — the gain in year 1, and the loss in year 2 — carryback is not permitted, and the taxpayer (inappropriately) owes tax on the $1 million gain.

Why have taxpayers not strenuously objected to this state of affairs? Part of the explanation may be optimism that sooner or later our taxpayer will realize another $1 million gain, which the disallowed loss will then offset. But the timing option is also an important part of the explanation. Because taxpayers control the timing of their gain and loss realizations, they can avoid the unfair result in the second hypothetical by realizing the loss in an earlier year than the gain, or in the same year as the gain. But now suppose the second hypothetical, gain followed by loss, occurs under a mark-to-market regime permitting loss carryforwards but not loss carrybacks. The taxpayer’s complaint that she is being overtaxed cannot then be answered by an invocation of the timing option, because the very point of mark-to-market taxation is to deny her the timing option. In short, the shift to mark-to-market taxation strengthens the case for carrybacks of mark-to-market losses.

Then again, allowing loss carrybacks could greatly increase the volatility of tax revenues under a mark-to-market system. Allowing mark-to-market losses to offset all current-year income, with unlimited carryforward, may zero out the liabilities of wealthy taxpayers in down-market years, and perhaps in carryforward years as well, but it will never result in a year in which wealthy taxpayers have negative tax liabilities. Carrybacks, by contrast, will predictably result in years in which Treasury, instead of receiving tax payments from the wealthiest Americans, sends millions or billions of dollars to billionaires. Imagine a billionaire who has $1 billion of mark-to-market gain in year 1, $1 billion of mark-to-market loss in year 2, and $1 billion of mark-to-market gain in year 3. Assuming a 40 percent marginal tax rate in all years, and unlimited carryforwards but no carrybacks, tax payments for the three years will be $400 million, zero, and zero. On the same facts, but with loss carrybacks permitted, tax payments for the three years will be $400 million, negative $400 million, and $400 million. Setting aside time-value-of-money concerns, the net results over the three years are the same under either approach, and under either approach a well-run federal government could base annual spending on average anticipated revenues without being distracted by dramatic year-to-year fluctuations between positive and negative tax receipts from wealthy taxpayers.

Still, there are two powerful reasons to prefer the schedule of annual tax payments under the carryforward-only approach to the schedule under the carryback approach. First, with respect to the overall effects of the mark-to-market regime, it seems unlikely that the political system will fully recognize the multiyear equivalence of the two revenue schedules and so calmly accept the negative tax revenues produced by the mark-to-market regime in down-market years.43 Rather, there will be a tendency to overspend (from a sustainable long-term perspective) in up-market years and to panic unnecessarily in down-market years.44 Second, in terms of public perceptions of the tax treatment of particular high-profile wealthy taxpayers, carrybacks are likely to produce public relations disasters for Congress, the IRS, and the taxpayers themselves. With carrybacks, the wealthiest Americans may have nine- or 10-figure negative income tax liabilities in down-market years; they will pay no taxes in those years, and they will receive refunds of taxes paid in previous years. It is not difficult to imagine how the media would cover, and how the public would react to, say, Jeff Bezos’s receiving a check from Treasury for hundreds of millions of dollars (or more) for a year in which Amazon stock plummeted.45

Despite the strong theoretical case for carryback of mark-to-market losses, these practical objections should probably carry the day. As long as markets do not remain down for multiple years, wealthy taxpayers denied the ability to carry back their mark-to-market losses should be sufficiently mollified by the prospect of tax-free income (whether mark-to-market capital gains, traditionally realized capital gains, or ordinary income) in the not-too-distant future.


The Wyden proposal raises, but does not discuss, the question of the extent to which mark-to-market losses should offset ordinary income; it does not raise (let alone discuss) the closely related question of carryovers of mark-to-market losses. This article has offered a historically informed analysis of the extent to which capital losses should be deductible under a mark-to-market regime. Although the history of capital loss treatments under our realization-based income tax is instructive, taxpayers’ inability to cherry-pick loss realization under a mark-to-market regime has major implications for the loss deductibility analysis. Thus the recommendations of this article are (1) that mark-to-market losses should be deductible against all types of current-year income, including ordinary income, and (2) that mark-to-market losses in excess of current-year income should carry forward indefinitely (to be deducted against all types of later-year income), but should not carry back.


2 Id. at 11.

3 Id. at 10.

4 Id. The paper notes the existence of several different approaches to designing a deferral charge. Rather than advocating any particular approach, it solicits comments on the advantages and disadvantages of the various lookback rule options. Id. at 23-24.

5 Id. at 9.

6 Id. at 29-32.

7 Id. at 15.

8 Id.

9 Arguably the request for comments on “what limitations should be placed on mark-to-market losses” (id. at 15) is broad enough to include the carryover question, but it certainly does not specify the issue.

10 Offset of traditionally realized gains should be permitted regardless of whether the gains are subject to a deferral charge under the lookback rules, but offset of any deferral charge should not be permitted.

11 Section 1211(b).

12 Section 1212(b).

13 Revenue Act of 1942, P.L. 77-753, section 150, 56 Stat. 798, 843.

14 Revenue Act of 1964, P.L. 88-272, section 230, 78 Stat. 19, 99.

15 For a considerably more expansive historical survey of the first three decades of capital loss limitations, see Lawrence Zelenak, Figuring Out the Tax: Congress, Treasury, and the Design of the Early Modern Income Tax 135-165 (2018).

16 Revenue Act of 1913, P.L. 63-16, section II.B.2, 38 Stat. 114, 167 (specifying all allowable deductions, and not mentioning capital losses).

17 T.D. 2005, 16 Treas. Dec. 111 (1914).

18 Revenue Act of 1916, P.L. 64-271, section 5(a), 38 Stat. 756, 759.

19 Revenue Act of 1918, P.L. 65-254, section 214(a)(5), 40 Stat. 1057, 1067.

20 Revenue Act of 1921, P.L. 67-98, section 206(b), 42 Stat. 227, 233.

21 Under the rate-symmetry approach, the net result of the $100,000 of ordinary income and the $100,000 capital loss would be a tax of $58,000 - $12,500 = $45,500.

22 Revenue Act of 1921, section 214(a)(5), 42 Stat. at 240.

23 Revenue Act of 1924, P.L. 68-176, section 208(c), 43 Stat. 253, 263.

24 Revenue Act of 1932, P.L. 72-154, section 24(r), 47 Stat. 169, 183 (limiting losses on securities held for two years or less); and id. at section 101(b), 47 Stat. at 191 (long-term losses deductible against ordinary income).

25 National Industrial Recovery Act of 1933, P.L. 73-67, section 218(b), 48 Stat. 195, 209.

26 Statement of Morgenthau regarding the preliminary report of a House Ways and Means subcommittee, at 6 (1933).

27 Id.

28 Revenue Act of 1934, P.L. 73-216, sections 23(j), 117(d), 48 Stat. 680, 689, 715.

29 Revenue Act of 1938, P.L. 75-554, section 117(c), 52 Stat. 447, 501.

30 Id. at section 117(d)(2) and (e), 52 Stat. at 502.

31 Revenue Act of 1942: Hearings Before the Ways and Means Committee, 77th Cong., 2d Sess., at 85 (1942).

32 Revenue Act of 1942, section 150, 56 Stat. at 843.

33 See, e.g., Alvin C. Warren, “The Deductibility by Individuals of Capital Losses Under the Federal Income Tax,” 40 U. Chi. L. Rev. 291, 310 (1972) (explaining that “tax avoidance opportunities” arising from a taxpayer having “control over the timing of realization for his capital losses” justify capital loss limitations in a realization-based system).

34 For the dramatic market drop in 1917, see William N. Goetzmann, Roger G. Ibbotson, and Liang Peng, “A New Historical Database for the New York Stock Exchange 1815 to 1925: Performance and Predictability,” 4 J. Fin. Mark. 1, 30 (2001).

35 Under the wash sale rules of section 1091, if a taxpayer realizes a loss on a sale of stock or securities, and purchases “substantially identical” replacement property within 30 days of the date of the sale (before or after), no deduction for the loss is permitted. This can be understood as a limited anti-cherry-picking rule; taxpayers can cherry-pick losses despite section 1091, but only if they do not quickly replace the loss property. Section 1091 should not apply to mark-to-market losses. If it did apply, the effect would be a deeply asymmetrical system with mark-to-market taxation of gains, but the continuation of realization-based treatment of losses in tradable property.

36 In years in which stock market performance is mixed, mark-to-market losses will mostly offset mark-to-market gains, which is not a troubling prospect from a revenue perspective. But in years in which the overall market is decidedly down, mark-to-market losses will mostly offset ordinary income (to the extent deductions against ordinary income are allowed), with what may be very significant revenue consequences.

37 Section 170(b).

38 Jesse Eisinger, Jeff Ernsthausen, and Paul Kiel, “The Secret IRS Files: Trove of Never-Before-Seen Records Reveal How the Wealthiest Avoid Income Tax,” ProPublica, June 8, 2021.

39 Assuming mark-to-market losses are allowed against all types of current-year income (ordinary income, mark-to-market gains, and realized gains on non-tradable assets), the carryover question will arise only when mark-to-market losses exceed the sum of all types of current-year income.

40 Under the ordinary income approach urged here, this would mean no limitation on deducting mark-to-market losses against ordinary income in later years. But if, for example, a 50-percent-of-ordinary-income limitation applied in the current year, the same limitation should apply in all later years to which the loss is carried.

41 Section 1212(a).

42 Revenue Act of 1918, section 204, 40 Stat. at 1061-1062 (permitting NOLs to be carried back one year and forward one year).

43 Of course, even in the worst down-market years a mark-to-market regime would not result in negative total federal income tax revenues. It would, however, result in dramatically negative tax liabilities for many of the wealthy taxpayers subject to the regime.

44 As noted earlier, allowing mark-to-market losses to offset same-year ordinary income would also result in revenue volatility, the management of which would pose a challenge for the government; the revenue volatility and the resulting management challenge would be much greater, however, if carrybacks were permitted.

45 A good estimate of the amount of the check could be based just on the price history of Amazon stock and on publicly available information on Bezos’s Amazon stock ownership; it would not require another disclosure of confidential tax return information to ProPublica or any other news organization.


Copy RID