(iii) The market's evaluation of securities is essentially a process of capitalizing current and prospective earnings and dividends. A change in a company's earnings prospects arising out of factors altogether unrelated to the company's investments, such as a new and important contract, is a common cause of substantial changes in the market value of a company's securities. A general improvement in the business outlook will frequently send up the prices of virtually all stocks; a change for the worse in this outlook will send down the prices of virtually all stocks; a decline in the level of interest rates tends to enhance the market value of virtually all securities.

(iv) In the treatment of capital gains and losses for tax purposes, we do not deal with aggregates or with averages, but with the experience of individuals. Individuals differ considerably in their ability to appraise the securities of particular companies and the general outlook for business. One group of individuals may enjoy substantial net capital gains during the very same year that another group of individuals suffers heavy capital losses. Hence the taxation of net capital gains, because of the uneven distribution of capital gains and losses among reporting individuals, is capable of producing substantial revenues even for a period in which the aggregate of reported net capital losses exceeds the aggregate of reported net capital gains.

The undistributed profits tax will tend to reduce only one of several important sources of capital gains. The other sources would remain of considerable importance. Structural changes in particular industrial fields give rise to the early capitalization by the market of unrealized large earnings prospects for particular companies well in advance of any commensurate increase in the book value of the securities arising out of the reinvestment of earnings. Likewise, capital gains arising out of the exploitation, by individuals of cyclical fluctuations would not be affected. The superior ability of some individuals to appraise investment situations, even under static conditions of general business, and the weaknesses of other individuals for unprofitable commitments, would remain as sources of capital gains and losses. The changes in the rate of interest end in the speculative temper of the moment, which powerfully affect the capitalization ratio by which the earnings of any capital instrument are translated into a capital value, will likewise remain as sources of capital gains and losses. Finally, mention need hardly be made of the unaffected possibilities of such speculative gains as flow from discoveries of oil wells and metal mines, the exploitation of new patents or secret processes, appreciation in the value of improved and unimproved real estate, et cetera.

2. Complete Segregation Of Tax Treatment Of Capital Gains And Losses

It has also been suggested by Mr. Parker that capital gains and losses be completely segregated from other sources of income for tax purposes. In the Tax Magazine for October 1936, Mr. Parker wrote:

" . . . . . suppose we do not add capital gains to other income or subtract capital losses therefrom. Conceive of capital gains and losses going into a separate basket by themselves. Now, if you will imagine a capital gains tax with graduated rates approximating one-half our present tax rates, and imagine capital losses as an offset against gains with a provision whereby the capital net losses my be carried forward for a period of six years, it will be found that a fairly equitable result will be obtained."

Mr. Parker's suggestion differs from that of some others in that he would graduate the separate taxes applied to capital gains, whereas others would apply only a low flat tax. Neither of these types of separate taxation of capital gains would eliminate the legal and administrative difficulties that have been encountered in taxing this type of income. The great sacrifice of equity involved in divorcing the tax treatment of capital gains from that of other types of income is very inadequately reduced by graduating the separate taxes on capital gains. Such graduation would have no reference to the amount of income from other sources, and, hence, the progressive character of our income-tax system would be grossly impaired.

The revenue raised under any rate scale that is applied separately to different parts of income must inevitably fall far below the amount of revenue that could be raised if the same rate scale were applied to the entire income. Substantially higher, rather than lower rates on capital gains would be necessary to raise the same amount of revenue if capital gains were taxed separately. Mr. Parker, on the contrary, suggests rates equal to one-half those applied to ordinary income. Although the contention is often made that low rate on capital gains would allow more rather than less revenue, because of the increase in the volume of capital transactions, this contention has yet to be supported by convincing evidence. Revenues lost through reduced taxation of capital gains must be raised from other sources. The existing tax treatment of capital gains, we have previously demonstrated, already gives excessive preference to recipients of this source of income.

3. Segregation Of The Type In Force Between 1922 And 1933

This type of segregation was highly inequitable. Its effect was to give no benefit whatever to taxpayers with small or moderate incomes, and to confer highly preferential tax treatment to the capital gains of individuals with large incomes.

4. Segregation With Respect To Capital Losses Only

The same considerations that argue in favor of the inclusion of capital gains in ordinary taxable income also argue in favor of the deductibility of capital losses from such income -- though special treatment for capital losses sustained on assets held for more than one year is supported by the same considerations that support such treatment for capital gains on assets held for more than one year.

While equitable considerations clearly call for greater liberality in the deductibility of capital losses, it is extremely doubtful that a substitute treatment based upon equitable considerations alone can long endure. It is precisely in periods when large capital losses are, or may be, realized, that the Government's revenues tend to shrink, and its expenditures to rise. The record of the past clearly indicates that if the deductibility of capital losses is too freely allowed, such liberality would be unlikely to withstand the pressure for the additional revenues that will arise in some future period of business decline. It thus appears altogether probable that some measure of segregated tax treatment of capital losses is expedient to protect revenues in periods of declining business activity and declining markets.

Apart from revenue requirements as such, there are certain practicable and even equitable considerations that have influenced Congress to restrict the deductibility of capital losses. When capital losses are liberally deductible, an incentive is provided many taxpayers to design artificial transactions for the sole purpose of wiping out or greatly reducing tax liability on ordinary income. Further, since the tax treatment of capital gains and losses has been confined to realized gains and losses, the taxpayer has been given the valuable privilege of choosing the time for realizing both gains and losses; and this choice, other things being equal, can be expected to operate regularly to the advantage of the taxpayer and to the disadvantage of the Government. The mere power of a taxpayer to postpone the realization of capital gains is itself a great advantage. Not only can he choose the timing of realization to minimize his tax liability, other things being equal, but he also enjoys, interest-free, the funds that would be payable in taxes if he realized his gains as they accrued. In addition, there has been a vague feeling, only partly warranted, that capital losses in general are suffered by substantially the same persons as, at other times, enjoy capital gains; and, hence, that any preferential treatment accorded capital gains justifies some measure of harshness in the tax treatment of capital losses.

5. Complete Aggregation Of Capital Gains With Other Income But Only Partial Association For Purpose Of Tax Computation

Of the four variants of this proposal that appear to be current, the first three, cited below, were outlined by Professor Maguire in an undated memorandum filed in the Treasury, entitled "Taxation of Capital Gains and Deduction of Capital Losses."

(i) Apply the regular rate scale to ordinary income and apply the top rate reached by ordinary income to the capital gains.

(ii) Begin the aggregation of income with capital gains, apply perhaps one-half of the regular rates to these gains, and the aggregate thereon to ordinary income, to which the regular rates will apply.

(iii) Apply the regular rate scale to ordinary income, compute the effective totality rate of tax on ordinary income, and apply this rate to the capital gains.

(iv) Aggregate capital gains with other income, and apply the regular income-tax rates until the segments of income representing capital gains are reached; at this point, reduce the rates otherwise applicable by perhaps one-half.

The first three of these plans appear to possess no special merit whatever. They are merely alternative means of reducing the effective taxes on capital gains, with no especially logical basis for the reduction in each base. The first plan breaks down in all instances where all other sources of income result in a deficit. A special flat rate would obviously have to be applied to the capital gains of the taxpayers who reported no other income. Further, the method might produce exceedingly whimsical results. Thus, an individual with $10,000 of ordinary surtax net income, but $1 million in capital gains, would be subject to $100,000 of taxes on his capital gains. Another individual who also reported $1 million of capital gains, but had $100,000 of ordinary surtax net income, would be subject to $590,000 of taxes on his capital gains. In other words, $90,000 of difference in ordinary surtax net income would create a difference of $490,000 in the tax applicable to the same amount of capital gains.

The second variant appears to be nothing more than a more circuitous and less defensible method for reducing the effective rates on capital gains than is presented in the fourth variant.

The third variant would yield results just as whimsical as those of the first variant. Thus, because the effective totality rate of $10,000 of ordinary surtax net income is 7 percent, and on $100,000 34 percent, the taxes on $1 million of capital gains would be $70,000 for an individual reporting $10,000 of ordinary surtax net income and $340,000 for an individual reporting $100,000 of ordinary surtax net income. A difference of $90,000 in ordinary surtax net income occasions a difference of $270,000 in tax liability on the identical amount of capital gains.

The fourth variant, whereunder capital gains would be aggregated with ordinary income but subjected to perhaps half the surtax rates that would otherwise be applicable, is worthy of some consideration. It recognizes both the size of ordinary income and the size of the capital gains in determining the taxes on the latter. On the other hand, it substitutes for the present step-down provisions, according to the number of years during which an asset was held, an all-over tax reduction on capital gains irrespective of the length of the period during which such gains accrued. Careful analysis of the effect of applying this proposal to selected cases demonstrates that the proposal would be less meritorious on equitable grounds, and more whimsical in its effects, than the existing step-down provisions, and far more so than the average accrual method which is recommended herein.

The effects of applying this proposal, as compared with the present treatment and as compared with the "accrual basis" which we have previously used as a standard of equity, are illustrated in selected cases in Tables XIV and XV.

B. Proposals For Complete Association Of Capital Gains With Other Income

1. Procedure Followed Under The Revenue Act Of 1918

Under this Act, which was applicable to the taxable years 1918 to 1921, inclusive, gains and losses from the sale of capital assets were included with income and deductions from other sources, without any special treatment whatever for either. We have already cited the fact that this procedure is inequitable because capital gains realized in a single year often represent the gains accrued over a number of years; and the full taxation of such gains in the year of realization, because of our system of progressive rates, would result in substantially larger taxes than would have been payable had the gains been taxed as they accrued. For this reason the reintroduction of this procedure is not advocated.

2. Inventory Method

The inclusion in taxable income of accrued though unrealized gains and losses, as well as realized gains and losses, has recently been recommended, with some reservations, in a volume published by the Twentieth Century Fund, Inc. (Facing the Tax Problem). Such a treatment would obviate the necessity for rate differentials based upon duration of ownership, and would automatically eliminate the possibility of avoiding capital gains taxation by delaying transfers of assets until death. It would also permit an accurate application of the progressive income-tax schedule to the annual increases or declines in the economic power of individual taxpayers. Because it would do these things, and because, at the same time it might accomplish some of the essential purposes of the tax on undistributed corporate earnings, this possible method possesses considerable theoretical appeal -- granted that its constitutional difficulties could be overcome.

There are, however, grave and decisive practical objections to this method. In the first place, it may be noted that this method of determining income, while widely employed among business enterprises, is entirely alien to the practices of the majority of individuals; and would strike many of them as being grossly unfair. An individual who had no intention of selling his house would be asked to pay an additional income tax because the tax assessor or someone else asserts that the market value of his house has risen. Or an individual deriving an unchanged income of 5 percent from certain listed bonds would be asked to pay an additional income tax because, interest rates having fallen, the market prices of his bonds, which he has no intention of selling, have risen. Every time that there is a speculative rise or fall in the stock market or in local real estate markets, individuals who did not participate at all actively or consciously in such a speculative movement, would be told that their taxable incomes must be computed with reference to these paper changes in value.

The contention that this type of tax treatment of capital gains and losses would be utterly neutral in its effect upon the securities markets by no means holds. Taxpayers who might otherwise take little notice of changes in the market values of their securities and other assets would be put on notice to take active heed of such changes; would be warned in many cases that they must sell securities that they would otherwise keep, in order merely to obtain funds with which to pay their income taxes. The practical role of speculative changes in the market values attached to capital assets would be vastly increased.

For unlisted securities and physical properties, the practical difficulties involved in obtaining accurate appraisals of changes in value are admittedly great. We have already called attention to the striking discrepancies among listed securities between book and market values. Such discrepancies are presumably no less in magnitude among unlisted securities. For physical properties, the unreliability of assessment figures, particularly as between one jurisdiction and another, are notorious.

Finally, the effect upon income-tax revenues in periods of declining capital values would be catastrophic, unless the income tax base were averaged for several years. Between December 1930 and December 1931, the aggregate decline in the market value of securities listed on the New York Stock Exchange alone amounted to more than $30 billions.

C. Modification In Present Step-Down Provisions

If the recommended plan for the tax treatment of capital gains and losses is for any reason deemed unsuitable, the most superior alternative would be a system similar to that now in effect, with liberalization of the limitations upon the deductibility of capital losses. The major inequities of the present method could be greatly reduced if the reductions in the taxable proportions of capital gains were made less pronounced; and if the loss limitation provisions were altered in the manner recommended below. (Tables IX and X compare this improved step-down schedule with the present step-down schedule and with the standard of equity described in Section IV.)

(1) Taxable proportions of capital gains under improved step-down schedule compared with those under present schedule.

          Asset held           Proposed        Present

          Under 1 year           100             100
          1 - 2 years             80              80
          2 - 3 years             75              60
          3 - 4 years             70              60
          4 - 5 years             65              60
          5 - 10 years            60              40
          Over 10 years           55              30

(2) In addition to the capital losses allowed in the current year through operation of the step-down percentages to the extent of the included capital gains, the excess of such included losses should be carried forward for use in the next two years as an offset against the included capital gains in those years.

D. Proposals To Tax Capital Gains On An Average Basis

To obviate the more important legal and other difficulties involved in the inventory method, but nevertheless to approximate the equitable advantages of such a method, various types of averaging devices for capital gains and losses have been proposed. None of these plans would necessarily diminish the deterrent influence of capital gains taxes upon transactions in capital assets for they are all based upon the tax treatment of realized gains and losses.

1. Proposal To Average Total Income Including Capital Gains And Losses In Full

The most usual form taken by the averaging proposals is that in which all income of the taxpayer including capital gains and losses is averaged for a period of years to determine the tax base. Great Britain formerly employed a 3-year average of income for income-tax purposes generally (though gains and losses from casual sale of property were excluded); but objections from taxpayers who were called upon to pay relatively heavier taxes in periods of declining incomes led to the abandonment of the averaging system in 1927.

Professor Henry C. Simons has suggested that capital gains and losses be reported fully with other income in the year realized, but that every five years or so the Treasury make rebates to those individuals whose actual tax payments exceeded by more than five percent the aggregate taxes that would have been payable if their taxable incomes each year had equalled their average income for the period. This proposal is attractive in many respects; but it would involve an enormous administrative burden upon the Bureau of Internal Revenue, which would be required to compute the putative tax liability for each taxpayer who reported gains or losses from capital transactions in the average period fixed upon.

2. Proposal To Average Capital Gains And Losses Separately From Other Sources Of Income

The proposal to average capital gains and losses separately from other sources of income has not been incorporated in the income tax law of any major country so far as known. This proposal has the merit of ironing out severe fluctuations in capital gains and losses, but is subject to somewhat the same limitations as the plan for averaging capital gains and losses together with other income.

3. The Average Accrual Method (Recommended Plan)

Among the various proposed methods of averaging capital gains and losses and associating them with ordinary income, the method that appears to be freest from anomalies is the average accrual method, recommended in Section VII.

The essential merit of the average accrual approach to the tax treatment of capital gains is that it places the treatment on a definitely uniform basis for all classes of taxpayers, and that it appears to represent the nearest practicable approach to perfect equity in such treatment. The treatment accorded net capital losses may be debatable, because of the limitations placed upon their deductibility; but, as previously noted, it is virtually impossible to design an equitable tax treatment for capital losses that can be expected to override revenue requirements in years when opportunities for loss realization are great. The computation of the tax may conceivably be deemed unduly complicated for the average taxpayer; but this difficulty, if it be a real one, is worth facing if the primary object is to obtain the most equitable and continuing tax treatment of capital gains that seems practicable.

The most serious objection that may be raised to this method, in the minds of many, is to be found in its results upon the effective rates of taxes applied to capital gains. While the essential effect of this method would be merely to subject capital gains to substantially the same rates of tax that apply to ether sources of income, a measure of the extent to which the method would raise the taxes now applicable to capital gains is presented in Table XI. The claim will be made that these increases will seriously constrict the volume of transactions in capital assets. On this point, the available evidence does not permit a conclusive position.

It is important to emphasize two considerations in this connection, however. In the first place, if the conviction became widespread that the application of this method marked the end of our long period of experimentation with the taxation of capital gains, and if the existing opportunity for the complete avoidance or substantial reductions in such taxes by transfers at dearth and through inter vivos gifts were eliminated, the incentive for delayed realization of capital gains would be greatly reduced. In the second place, and because of the same factors, the contingent tax liability involved in unrealized capital gains could be expected to exercise a far more powerful influence in counteracting any existing tendency to delay liquidation than it does today.

There are several important reasons for the latter expectation. An investor who today holds assets incorporating unrealized capital gains is offered a definite tax reduction, assuming a constant future income, to delay liquidation for a period up to just beyond ten years in any event; and even until death. This tax inducement would be eliminated if the recommendations were adopted. The investor might still obtain a tax advantage by delaying liquidation if he expected his ordinary income in some future year to be substantially below his present income, or his capital losses greater, provided that he could look forward confidently to realizing much of his present possible gain in such a year. It should be noted that the conditions which are unfavorable for large amounts of other income, and favorable for the realization of capital losses, tend to be unfavorable for the realization of capital gains. Hence, it can be said that the subject proposal would greatly diminish this deterrent influence to the liquidation of capital assets.

Another consideration which is important in the minds of large investors is the possibility that our long period of experimentation with capital gains taxation will be indefinitely continued, with recurring opportunities, by reason of favorable alterations in tax treatment, to liquidate capital assets with far lesser tax liability than would be incurred today. Thus, many persons at this moment are doubtless contemplating the possibility of the return to the 12 1/2 percent maximum rate that was in force between 1922 and 1933, inclusive, or one even lower. If it appeared that the adoption off the average accrual plan, because of its preeminent equity, as compared with practicable alternatives, and because of its consistency with the remainder of our income-tax structure, definitely settled the problem of capital gains taxation, a very considerable part of the existing tax influence upon capital transactions would be removed.