NYSBA Members Submit Report on Ownership Changes in Loss Corporations

Dear Sirs:

We are pleased to submit a report on Section 382(l)(3)(C), which addresses the effect of certain value fluctuations on the determination of whether a loss corporation has an ownership change under Section 382(g). This Report is intended to inform the difficult technical and policy choices that the government must make in formulating final guidance with respect to Section 382(l)(3)(C).

Section 382(g)(1) generally provides that an ownership change occurs if, on a testing date, a "5-percent shareholder" of a corporation has increased its ownership in the stock of a corporation by more than 50 percentage points during the preceding three-year period. Section 382(b(3)(C) modifies Section 382(g) by providing, "Except as provided in regulations, any change in proportionate ownership which is attributable solely to fluctuations in the relative fair market values of different classes of stock shall not be taken into account." Temporary regulations issued in 1987 "reserved" with respect to the treatment of changes in proportionate ownership attributable to fluctuations in value. Because of the lack of guidance, questions surrounding fluctuations of value under Section 382(b(3)(C) thus far have been resolved on a case-by-case basis through the Internal Revenue Service's private letter ruling program.

The threshold issue under Section 382(l)(3)(C) is whether the provision should be limited only to prohibiting fluctuations in the relative values of two or more classes of stock from causing a testing date, or whether it should more broadly factor out such value fluctuations in the owner shift analysis. We support an approach to Section 382(l)(3)(C) that does not limit its application to the mere avoidance of a testing date when relative values change and would allow taxpayers to factor out value fluctuations in at least certain circumstances.

The next issue is whether the government should craft a broad rule generally factoring out value fluctuations or prescribe a rule that more narrowly targets specified circumstances. A majority of the Tax Section's Executive Committee supports some formulation of a broad approach that generally factors out value fluctuations in determining owner shifts. We envision that such an approach would be similar to the methodology that the IRS has used in a number of private letter rulings (the "PLR Methodology"). However, we did not reach an unequivocal consensus on this point, as a significant minority of the Tax Section's Executive Committee supports a narrower approach.

The majority's viewpoint is fundamentally grounded in the belief that value fluctuations are not indicative of loss trafficking and thus generally should not cause the imposition of a limitation on the use of corporate tax attributes. As a matter of general principles, the majority would distinguish between (i) situations in which a shareholder's proportionate interest increases solely from value fluctuation -- that is, where the shareholder has not acquired additional shares, and (ii) those in which the shareholder's proportionate interest increases by virtue of a combination of value fluctuation and acquisition. The majority would factor out the impact of value fluctuation in the former case, but not in the latter case.

The report recommends that additional guidance be issued in order to improve administrability of the PLR Methodology and prevent abusive results, some of which are illustrated in the examples this Report. In particular, the report recommends that if the PLR Methodology is adopted as the framework for final guidance, regulations should (i) address the impact of fluctuations in value prior to the testing period, (ii) provide one or more conventions for a shareholder that redeems or disposes of stock, where the shareholder has acquired multiple blocks of that stock at different times, (iii) clarify the interaction of the PLR Methodology with the small issuance and cash issuance exceptions, and (iv) address the effect of recapitalizations and other exchanges of stock.

The minority's viewpoint is based on the belief that Section 3 82's owner shift analysis was only intended to be a rough-cut method of identifying circumstances that have the potential for loss trafficking or inconsistency with the income averaging function of the carryover regime. In the minority's view, a broad approach to factoring out fluctuations in value would necessarily involve substantial complexity, could create unintended opportunities for abuse, or both. Accordingly, in this view, fluctuations in relative value among multiple classes of stock should be factored out in certain limited circumstances: where shifts resulting from actual share movements are de minimis, and potentially where relatively small shifts occur shortly before or during a bankruptcy proceeding. The approach of adopting a few limited exceptions may alleviate a considerable amount of complexity, but may be criticized as providing relief in too narrow a range of cases, lacking a persuasive conceptual framework, and raising line-drawing issues.

In the end, there is no one perfect solution for whether, the extent to which, and how fluctuations in relative value should be disregarded in calculating owner shifts under Section 382. It is clear that trade-offs will be required to be made (e.g., between complexity and precision). But there can be no doubt that it is important for the Service and the Treasury to bring clarity to this uncertain area of practice.

We appreciate your consideration of our comments. Please let us know if you would like to discuss these matters further or if we can assist you in any other way.

This report of an ad hoc committee of the Tax Section of the New York State Bar Association (the "Report") discusses the effect of Section 382(l)(3)(C) on the determination of whether an ownership change has occurred under Section 382(g). Section 382(l)(3)(C) provides that changes in proportionate ownership attributable solely to fluctuations in the relative fair market values of different classes of stock are not to be taken into account in determining whether a loss corporation has an ownership change and is thus subject to limitation on the utilization of certain of its attributes. The question of whether and how to take change in relative stock valuation into account is of particular concern for companies that have issued one or more classes of preferred or other stock that is privately held or does not regularly trade, for which valuation may be difficult.

The circumstances under which Section 382(l)(3)(C) is applied and the methodology for doing so are uncertain. Since 2004, the Service has issued several private letter rulings ("PLRs") that have permitted the subject taxpayers to factor out fluctuations in value under Section 382(l)(3)(C). However, it remains unclear whether a taxpayer can factor out value fluctuations in absence of a private letter ruling and the principles described in the PLRs leave unanswered important technical questions as to how fluctuations should be dealt with in numerous circumstances not addressed in the PLRs. Tax practitioners currently take different approaches to Section 382(l)(3)(C), resulting in variances in treatment across taxpayers, even where they are similarly situated.²

The Service and the Treasury have indicated that they intend to release interim guidance while consideration of final guidance proceeds. We support the issuance of interim guidance as necessary to provide clarity pending the issuance of final guidance. While we express no views as to the proper interpretation of Section 382(l)(3)(C) in the absence of the issuance of regulations, clarification of the rules on an interim basis would be helpful for taxpayers and the government alike; for example, it would diminish uncertainty attending the preparation of financial statements.

The objective of this Report is to inform the government's decision-making regarding final guidance, by illuminating policy choices that must be addressed and analyzing the principal alternatives. The Report will do so by considering whether, the extent to which, and how fluctuations in relative stock value should be disregarded in calculating owner shifts under Section 382. Part II of the Report summarizes our recommendations for final guidance under Section 382(l)(3)(C). Part III of the Report provides some background on the issue before turning in Part IV of the Report to whether fluctuations in relative stock values should be disregarded at all under Section 382. After looking at the tax policies for and against disregarding fluctuations in relative stock values, the Report examines the extent to which and the methodology by which these fluctuations should be factored out. In addition to examining certain approaches to neutralizing fluctuations in value that have gained some visibility among tax professionals, the Report considers other alternatives also intended to strike a balance between the policies generally underlying Section 382, the policies underlying Section 382(l)(3)(C) in particular, and administrability concerns.

II. SUMMARY OF RECOMMENDATIONS

1. We support an approach to Section 382(l)(3)(C) that does not limit its application to the mere avoidance of a testing date when relative values change and would allow taxpayers to factor out fluctuations in the relative value of various classes of stock in calculating shifts in ownership in at least certain circumstances.

2. A majority of the Tax Section's Executive Committee supported some formulation of a broad approach that generally factors out value fluctuations in determining owner shifts. This viewpoint is fundamentally grounded in the belief that value fluctuations are not indicative of loss trafficking and thus generally should not cause the imposition of a limitation on the use of corporate tax attributes. As a matter of general principles, the majority would distinguish between (i) situations in which a shareholder's proportionate interest increases from value fluctuation, but the shareholder has not acquired additional shares, and (ii) those in which the shareholder's proportionate interest increases by virtue of a combination of value fluctuation and acquisition. The majority would factor out the impact of value fluctuation in the former case, but not in the latter case.

Nevertheless, we did not reach an unequivocal consensus in support of such approach or on the particular formulation of a broad approach to factoring out value fluctuation. We acknowledge that, in practice, the PLR Methodology in its current form can be difficult to apply and administer and raises anti-abuse issues.³ Additional guidance is required to improve administrability of the PLR Methodology and prevent abusive results, some of which are illustrated in the examples this Report. In any event, the PLR Methodology is preferable to a methodology that simply converts shares of different classes of stock into a number of common shares based on their relative values at the same time both classes are first outstanding (the "Common Share Equivalent Methodology").

3. A significant minority of the Tax Section's Executive Committee supports a more limited approach to value fluctuation. This viewpoint is based on the belief that Section 382's owner shift analysis was only intended to be a rough-cut method of identifying circumstances that have the potential for loss trafficking or inconsistency with the income averaging function of the carryover regime. In the minority's view, a broad approach to factoring out fluctuations in value would necessarily involve substantial complexity, create unintended opportunities for abuse, or both. Accordingly, in this view, fluctuations in relative value among multiple classes of stock should be factored out in certain limited circumstances. A de minimis exception would balance the multiple concerns of Section 382(g) by providing relief in the most sympathetic circumstances, while being easier to implement than the PLR Methodology. In addition, while recognizing that reasonable people may differ as to whether there ought to be a pre-bankruptcy exception, this Report develops arguments that the Service and Treasury should consider in determining whether additional protection against value fluctuations ought to be accorded to a loss corporation in the periods immediately before a bankruptcy filing and during the pendency of the bankruptcy case. The approach of adopting a few limited exceptions to a general narrow rule may alleviate a considerable amount of complexity. However, it may be criticized as providing relief in too narrow a range of cases, lacking a persuasive conceptual framework, and raising line-drawing issues.

III. BACKGROUND

Under Section 382(g)(1), an ownership change occurs if, immediately after any owner shift involving a 5-percent shareholder or any equity structure shift, one or more 5-percent shareholders increase their ownership in the loss corporation stock, in the aggregate, by more than 50 percentage points during a rolling three-year period (the "testing period").⁵ A 5-percent shareholder is a person that owns (directly or indirectly) 5 percent or more of the loss corporation's stock measured by value.⁶ An owner shift is any change in the ownership of the stock of a loss corporation that affects the percentage of such stock owned by any 5-percent shareholder.⁷ An equity structure shift is any reorganization within the meaning of Section 368, except divisive "D" or "G" reorganizations and "F" reorganizations.⁸

A loss corporation is required to determine whether an ownership change has occurred on each testing date, which is any date on which there is any change in the ownership of the stock of the loss corporation that affects the percentage of stock owned by any 5-percent shareholder.⁹ Reg. § 1.382-2T(c) provides the following mechanical "snapshot" test to determine whether an ownership change has occurred:

The general mechanics of Section 382(g) can be demonstrated by the following example.

The preceding example is a simple illustration of how Section 382(g) operates. In the example, P has increased its interest in LCorp by more than 50 percentage points entirely through the actual acquisition of the requisite amount of LCorp stock. That P's actions should trigger an ownership change is intuitively correct. However, the application of the value-based approach of Reg. § 1.382-2T may cause an ownership change in circumstances that raise difficult questions going to the core of the conceptual underpinnings of Section 382. Specifically, fluctuations in value between different classes of stock may trigger an ownership change despite the fact that neither the 5-percent shareholders nor the loss corporation have engaged in a substantial acquisition, disposition, issuance, or redemption of stock of the loss corporation.

The following example illustrates a situation in which a strict application of the value-based rule in Reg. § 1.382-2T triggers an ownership change that is principally attributable to value fluctuations between two classes of stock:

The calculation above does not take into account the fact that 60 points of the 80 percentage point shift is attributable to fluctuations in the value of P's preferred stock relative to the value of LCorp's common stock and not from the acquisition or disposition of stock by P. We will examine later in this Report to what extent it would be appropriate to factor out fluctuations in this situation and, if so, how to do so.

In PLR 200411012, the loss corporation ("Taxpayer") underwent an ownership change on Date 1. On Date 1, Taxpayer, which already had Common Stock outstanding, issued Class 1 Preferred Stock. After the issuance of the Class 1 Preferred Stock, Shareholder A owned shares of both Taxpayer Common Stock and Class 1 Preferred Stock. As of Date 2, the Common Stock had declined in value. On Date 2, Taxpayer issued two additional classes of preferred stock, Class 2 Preferred Stock and Class 3 Preferred Stock. The Class 3 Preferred Stock was issued to Shareholder A. The Class 2 Preferred Stock was issued to a third party.

On Date 5, which fell in between Date 1 and Date 2, Shareholder A bought additional shares of Common Stock on the open market. In addition, during the period in between Date 1 and Taxpayer's proposed transaction, there were two issuances of Common Stock by Taxpayer as a result of the exercise of stock options and six sales of Common Stock by 5-percent shareholders. Taxpayer requested a ruling because it was proposing to issue additional stock.

Based on the facts and representations, the Service ruled that:

Applying this "PLR Methodology" would eliminate the ownership change in Example 2. P's percentage ownership in LCorp on January 1, 2010, the testing date, would remain constant since the date on which P acquired the stock, 20 percent. O's percentage ownership of the loss corporation on January 1, 2010 would be 20 percent, representing the percentage ownership of O on the date that O acquired the stock. The total cumulative shift on January 1, 2010 would, therefore be 20 percentage points, representing the 20 percent interest acquired by O minus zero, O's lowest percentage ownership during the testing period.

IV. WHETHER, HOW, AND UNDER WHAT CIRCUMSTANCES SHOULD FLUCTUATIONS IN RELATIVE VALUE BE DISREGARDED?

The value fluctuation rule of Section 382(l)(3)(C) applies "except to the extent provided in regulations." Congress thus has given Treasury and the IRS wide latitude to decide how and under what circumstances fluctuations in relative stock value should be disregarded in calculating owner shifts under Section 382(g). In fashioning the mechanics of a regulatory rule under Section 382(l)(3)(C), the Service and Treasury should be guided by the policies underlying Section 382.

As discussed below, we believe that the policies of Section 382 do not compel the adoption of any single approach to factoring out fluctuations in value. An approach to Section 382(l)(3)(C) that furthers one goal underlying Section 382 may simultaneously obstruct another. Therefore, in formulating an appropriate methodology for Section 382(l)(3)(C), the government is likely to find itself in the unenviable position of having to choose between conflicting goals (e.g., precision in factoring out value fluctuations and avoiding the creation of opportunities for abuse vs. administrability). In the section immediately below, we will review the policy goals of Section 382 so to provide the analytical framework for fashioning a methodology under Section 382(l)(3)(C).

The House Committee Report for the Tax Reform Act of 1986 repeats the loss trafficking concern articulated in the 1983 Report, i.e., the concern that tax considerations should neither incentivize nor discourage corporate acquisitions. Accordingly, Congress attempted to revise Section 382 "to reduce the number of circumstances in which NOL carryforwards can be used as a device for transferring tax benefits."¹⁷ To achieve that end, Congress decided to retain the objective triggering event rules of old Section 382, concluding that "changes in a loss corporation's stock ownership continues to be the best indicator of a potentially abusive transaction."¹⁸

Though Congress did not adopt a statutory pool-of-capital rule, the second principle that only the owners who funded the losses should benefit from the tax benefit therefrom remains evident in the legislative history. Regarding the general operation of Section 382(g), the Conference Committee explained that "the special limitations generally apply when shareholders who bore the economic burden of a corporation's NOLs no longer hold a controlling interest in the corporation. In such a case, the possibility arises that new shareholders will contribute income-producing assets (or divert income opportunities) to the loss corporation, and the corporation will obtain greater utilization of carryforwards than it could have had there been no change in ownership."¹⁹ The concern reflected in the above language is clearly independent of the principle that the transfer of tax benefits should not incentivize acquisitions and seems to be rooted in the emphasis on the averaging function for loss carryovers. Congress believed that the averaging function of the loss carryover rules should be inoperative where there was a substantial change in share ownership in the interim between when the loss is generated and the income is earned, regardless of whether the acquiror purchased the stock because of the tax benefit.

The final goal of Section 382, to provide taxpayers with administrable rules for loss limitation, is reflected in a statutory scheme that rejects a subjective intent-based test of whether a transaction is undertaken, in whole or in part, for trafficking reasons in favor of an objective test based on changes in percentage ownership of stock. Congress originally enacted Section 382 in 1954, in part, because Section 269 had proven ineffectual as a weapon against loss trafficking. Section 269 requires a showing that the taxpayer acquired control of a corporation for the primary purpose of tax avoidance. This intent-based standard proved too difficult to administer and enforce as the primary tool for policing avoidance. In crafting Section 382, Congress chose instead to trigger the limitation on the use of losses upon an objective event, namely a cumulative 50-percent change in the ownership of the loss corporation's stock.

The question then is how Section 382(l)(3)(C) meshes with the stated policy concerns of Section 382. Under one view, shifts in percentage ownership attributable to value fluctuations do not implicate "loss trafficking" concerns. To the extent that "loss trafficking" is defined as tax-motivated corporate acquisitions, a shift in ownership attributable to value fluctuations does not offend the anti-loss trafficking goal of Section 382. Value fluctuations, by definition, involve no actual shift in ownership of shares and thus there could not be a tax-motivated acquisition of shares. As one commentator stated, factoring out fluctuations in value "makes sense because the holders of loss corporation stock whose proportionate interest increases without them taking affirmative action should not be treated as engaged in loss trafficking, the prevention of which is the objective of the statutory regime."²⁰

Nevertheless, the policy implications of value fluctuations are less clear if the owner shift calculations under Section 382 are viewed merely as identifying circumstances creating the potential for loss trafficking more broadly defined -- i.e., the ability of any person whose ownership interest in a loss corporation has increased to achieve enhanced utilization of corporate losses. Under this view, tax-motivated changes in ownership of loss corporations are not the only object of Congressional concern. Instead, changes in ownership percentages are indicative of the potential for persons whose ownership interests have increased to cause the loss corporation to engage in transactions enhancing use of its tax attributes. It could be argued that, regardless of whether an increase in ownership is attributable to acquisition or fluctuation in value, the shareholder whose ownership interest has increased has the same incentive to act in a manner resulting in enhanced loss utilization (e.g., "stuffing" or "income acceleration"). It is not clear that Congress' conception of loss trafficking was limited to the more narrow view. Indeed, the legislative history of Section 382, discussed above, seems to associate shifts in ownership alone as an indicator of loss trafficking potential. Moreover, as discussed more fully below, as a practical matter, enacting a rule generally disregarding fluctuations in relative stock value likely leaves open the possibilities of loss trafficking.

The principle that fluctuations in relative stock value should be disregarded also may be viewed as being at odds with the income averaging policy of the carryover rules and the correlative policy that only the persons who bore the economic losses should receive tax benefits resulting therefrom. In Example 2, P ultimately comes to own 80 percent of the value of LCorp as a result of the depressed value of C's common stock. Because C provided 80 percent of the funds into LCorp at formation in the form of common equity, C suffered all of the economic losses that are reflected in the loss carryforwards of LCorp. Nonetheless, because of the fluctuation in value, on a going-forward basis P may benefit from the use of the NOL carryforwards to shelter income streams funding its preferred returns.

Finally, Treasury and IRS should consider the effect of their interpretation of Section 382(l)(3)(C) on Congress' goal to provide simple, objective, and administrable rules for limiting losses. As more fully explored below, disregarding fluctuations in relative stock value necessarily imports an additional layer of complexity into the already difficult ownership change calculations of Section 382(g). Crafting regulations that get to the "right answer" under every fact pattern or even most fact patterns in which fluctuations in value are at play is very likely to involve considerable complexity, and this factor raises important questions as to whether, on balance, such an approach is appropriate in the context of the "rough-cut" approach of Section 382(g).

The above analysis suggests that the policies underlying Section 382 do not compel the creation or maintenance of a broad rule factoring out fluctuations in value in measuring owner shifts. But the inclusion of Section 382(l)(3)(C) itself suggests that Congress believed that shifts in ownership attributable to value fluctuation represented a lesser concern than shifts in ownership attributable to acquisition. While the Service and the Treasury certainly have authority to cut back the scope of any value fluctuation rule, it seems unlikely that Congress would have intended that the rule be eliminated in its entirety. As one example, little purpose would seem to be served by requiring loss corporations to revalue classes of stock on each date, and thus limiting testing dates to days on which shares actually move makes clear sense in terms of administrability.

In the following section, we will consider how the stated policy goals of Section 382 mesh with potential approaches to factoring out fluctuations in value.

A narrow reading of Section 382(l)(3)(C) has certain advantages. First, the narrow approach can be viewed as being most consistent with the language of Section 382(l)(3)(C) itself. The statute allows taxpayers to disregard "any change in proportionate ownership which is attributable solely to fluctuations in the relative fair market values of different classes of stock."²³ The use of the word "solely" suggests that fluctuations in relative value could, in combination with other factors, impact ownership percentages. While more expansive views of Section 382(l)(3)(C) have widespread consequences based on temporal distinctions, the narrow view is more constrained because fluctuations in relative share class values are taken into account when proportionate ownership otherwise shifts.

Second, the narrow view may be viewed as consistent with the current regulations and thus familiar to taxpayers. As drafted, the Regulations base the percentage ownership on the "relative fair market values" of the shareholder's stock and the corporation's outstanding stock at "any time during the testing period," essentially contemplating continuous revaluation.²⁴ Proponents of the narrow approach interpret the Regulations to require revaluation at each testing date. Because this method measures relative fluctuations in value between classes of shares, it matches Section 382 ownership changes with actual shifts in economic ownership.

Third, while the narrow approach requires revaluation of classes of stock on each testing date and the valuation of stock that is not publicly traded is an endemic problem in Section 382 owner shift analyses, the narrow approach is, on the whole, simpler to apply than more expansive approaches. A narrow interpretation of Section 382(l)(3)(C) does not require an algorithm based cumulatively on a register of the value of every share relative transferred relative to value of every other share at the time of transfer. As a result, the formula is easier to derive and compute. Furthermore, the sum of shareholders' total ownership will equal 100 percent, leading to more intuitive results.

On the other hand, a narrow approach to Section 382(l)(3)(C) would accord very little deference to Congress' apparent belief that shifts in ownership attributable to fluctuations in value are less worrisome from a policy perspective. In a sense, it would render Section 382(l)(3)(C) to be a nullity. The definition of "owner shift" under Section 382(g)(2) requires a "change in the respective ownership of stock of a corporation," which presumably requires an actual movement of shares, not merely a fluctuation in relative value between two or more classes of stock. Accordingly, if Section 382(l)(3)(C) were limited to precluding value fluctuations from causing a testing date, it would simply confirm the result already achieved by the "owner shift" definition. As noted in the NYSBA Supplemental Report, "[t]he effect of [the precursor to Section 382(l)(3)(C)] appears to be only to confirm that something more than a fluctuation in value is needed in order to have an ownership shift."²⁵

Further, the narrow interpretation would expose a loss corporation to a harsh limitation on loss utilization where the shift attributable to an actual acquisition is minimal, or where the shift in ownership is disproportionately attributable to value fluctuation and occurs close in time to a bankruptcy filing. For example, the narrow approach can lead to inequitable results where a 5-percent shareholder's 50 percentage point increase is due predominantly, not solely, to fluctuations in share value.

If LCorp were to file for bankruptcy soon after the sale to O, it would effectively be precluded from accessing the benefits of Section 382(l)(5) or Section 382(l)(6).²⁶ While one of these special bankruptcy rules could apply to a subsequent ownership change occurring pursuant to the bankruptcy reorganization, LCorp's attributes generally would already be subject to the lower limitation from the first ownership change pursuant to the operation of Reg. § 1.382-5(d).

In sum, while the narrow approach to Section 382(l)(3)(C) advances certain policies of Section 382, especially the goal of providing rules that are administrable for both the Service and taxpayers, it is too constricted in our view and does not sufficiently carry out the Congressional purposes underlying Section 382(l)(3)(C). Accordingly, we do not recommend the adoption of a narrow approach to Section 382(l)(3)(C).

The intermediate approach would provide special rules for specific circumstances where under a pure value-based rule, fluctuations in value would trigger results that clearly seem unfair or clearly conflict with competing policies. It thus would alleviate some of the harshness involved in adopting a narrow rule, without giving rise to all of the complications that a broad rule might. However, the intermediate approach could be criticized on the grounds that the considerations leading to the special rules are present in other change of ownership fact patterns, albeit in less extreme form, and that there is therefore no persuasive conceptual framework underpinning to the special rules. Moreover, as with other de mimimis rules or targeted exceptions, any special rule would by its nature be arbitrary in terms of where the line was drawn and would potentially give rise to cliff effects. As a result, it could difficult to determine the "right" answer for fact patterns not directly addressed by final guidance, and could lead to pressure to expand the special rules to very similar fact patterns. As some of the discussion below illustrates, the result can be a "creeping" broad approach, which could become less stable and more difficult to administer over time than a more comprehensive approach. Notwithstanding that, we believe that a de minimis rule of some kind would be appropriate.

Over twenty years ago, the NYSBA Supplemental Report proposed a de minimis exception to the normal owner shift rules. Under this exception, a loss corporation would not have an ownership change unless the percentage point shift resulting from the particular transfer of shares that triggers the testing date (together with all prior share transfers subject to the exception) exceeds a certain de minimis amount.²⁷ The NYSBA Supplemental Report suggested a flat 2-percent threshold. The NYSBA Supplemental Report's proposal was set forth in summary fashion, and its mechanics are not clear. In fact, the proposal is described in two sentences:

As a result, the protection of the de minimis rule envisioned by the NYSBA Supplemental Report appears to be quite narrow. One greater-than-2-percent acquisition precludes qualification under the de minimis exception from that point forward. It seems likely that qualification for the de minimis exception would be limited to a small minority of loss corporations in practice (such as a fact pattern similar to Example 3 above).

Perhaps a better formulation of a de minimis exception (and probably a better interpretation of the intent of the NYSBA Supplemental Report) is to have a rolling analysis. Under this approach, share valuations would be held constant so long as shifts attributable to actual share movements are de minimis since the last testing date that was a revaluation date. Not each testing date would be a revaluation date. Revaluation of all shares would not be required on any testing date unless the cumulative shift since the immediately preceding revaluation date exceeded the de minimis threshold (say, two percentage points), determined using the values that existed on the prior revaluation date. This version avoids a single greater-than-two-percent acquisition from precluding access to the de minimis exception for the entire testing period, and we recommend that the government consider it.

We recommend that the government consider adopting a higher threshold for a de minimis exception than the previously-recommended 2-percent threshold. We make this recommendation in light of the greater volatility in trading patterns with respect to distressed companies that has developed since the NYSBA Supplement Report was issued. A 2-percent threshold may provide relatively little relief given the degree to which there is trading in the stock of loss corporations in the current marketplace that may not be related to any actual loss trafficking. Accordingly, a threshold of at least 5 percent is appropriate and a strong argument can be made for an even higher threshold (e.g., 10 percent).

A more generous application of the de minimis exception in the context of an impending or current bankruptcy would provide relief for a loss corporation during the period when such fluctuations in relative value are most likely to occur or would be most likely to deny the loss corporation access to the special bankruptcy rules of Section 382(l)(5) and (l)(6).

A more liberal approach to value fluctuation may be appropriate in these circumstances in light of the competing policies underlying these special rules. Triggering an ownership change at a time when the value of the loss corporation's stock is, or is almost, worthless as a result of value fluctuations may effectively result in a very low Section 382 limitation, thereby preventing the loss corporation from benefiting from Section 382(l)(5) or Section 382(l)(6). Such a result conflicts with the bankruptcy policies favoring reorganization of troubled businesses. It also runs counter to the conceptual underpinning of the special bankruptcy rules -- that is, that the creditors frequently have economically borne a portion of the loss corporation's NOL -- and thus shifts in ownership pursuant to a bankruptcy workout with the creditors are entitled to more liberal treatment under Section 382 because the creditors financed some of the losses. Resolving the conflict in favor of these competing policies, it can be argued, seems appropriate given that Congress appears to have had a lesser concern with ownership changes attributable to value fluctuations.

On the other hand, it can be argued in the alternative that no special fluctuation in value rules ought to be created for bankrupt loss corporations. Providing a special bankruptcy rule would, for the first time, create a difference in the owner shift calculation for loss corporations reorganizing inside or outside of bankruptcy. While Sections 382(l)(5) and 382(l)(6) do create differences in the impact of having an ownership change for bankrupt and non-bankrupt loss corporations, the rules for determining whether a loss corporation has an ownership change are the same whether or not the loss corporation is in bankruptcy. Other arguments that weigh against the adoption of a title 11 exception are: (1) such an exception could push a loss corporation to file bankruptcy or, at least, file sooner than necessary; (2) the exception necessarily characterizes the impact of shifts in ownership based on "after-the-fact" events (i.e., the existence of a later bankruptcy filing); (3) the presence of Section 382(l)(5) and (l)(6) equally supports the underlying assumption that Congress believed that fluctuations in value should not be the cause of an ownership change, regardless of whether the loss corporation ultimately files for bankruptcy; and (4) the language of Section 382(l)(3)(C) suggests no special importance for fluctuations in relative value leading up to filing for bankruptcy.

The basic operation of the PLR Methodology can be illustrated by applying it to the facts of Example 2. In that example, after P's preferred stock has increased from 20 percent of the total value of LCorp's stock to 80 percent of such value as a result of fluctuations in the relative of the common and preferred stock, C sells all of the common stock to O, which represents 20 percent of the value of LCorp at the time of the sale. The PLR Methodology requires that the value of P's preferred stock relative to all the stock of LCorp outstanding on the date of P's acquisition of such stock remain constant. However, the common stock acquired by O from C on the testing date is revalued. Therefore, because the preferred stock has not moved since P's acquisition, P's interest in LCorp remains "frozen" at 20 percent for purposes of Section 382(g), despite the fact that at the time of the sale to O, P's preferred stock actually represents 80 percent of the value of LCorp. Accordingly, the cumulative shift of 5-percent shareholders on the testing date is a mere 20 percentage points, which is entirely as a result of O's increase in its ownership of LCorp stock.

While there may be many methods of effectuating the principles discussed above, we examine two alternatives for applying the PLR Methodology.

Alternative 2 employs the following procedure to factor out fluctuations in value. The value relationship of a share of loss corporation stock (i.e., the percentage ownership represented by a share of loss corporation stock) is initially fixed upon receipt, based on relative fair market values of all classes of loss corporation stock outstanding at that time.

The percentage ownership represented by such share of loss corporation stock is adjusted only (a) if that share is acquired by a different 5-percent shareholder, or (b) in the event of a subsequent issuance or redemption of loss corporation stock.

When a share of loss corporation stock is acquired by a different 5-percent shareholder, the percentage ownership of the acquired share in the hands of the acquiring shareholder is determined by reference to relative fair market values of all classes of loss corporation stock outstanding on the date the share is acquired, and the reduction in the transferor's percentage ownership is based on the percentage ownership of the transferred share in the hands of the transferor immediately prior to the transfer.

Using Alternative 2, upon an issuance or redemption of loss corporation stock, each 5-percent shareholder's percentage ownership is adjusted by an issuance or redemption factor. The following example illustrates a difference in implementing the PLR Methodology using Alternative 1 and Alternative 2.

Under Alternative 1, the percentage ownership reflected in Investor's stock equals $10, the actual value of the Series A Preferred Stock, divided by the sum of

As a matter of general principles, the majority would modify the PLR Methodology by distinguishing between (i) situations in which a shareholder's proportionate interest increases from value fluctuation, but the shareholder has not acquired additional shares, and (ii) those in which the shareholder's proportionate interest increases by virtue of a combination of value fluctuation and acquisition. The majority would factor out the impact of value fluctuation in the former case, but not in the latter case.³⁵

As a group, we do have certain remaining concerns about aspects of the PLR Methodology, which are discussed more fully below and are summarized here for purposes of illustrating issues that may need to be addressed in final guidance. First, the PLR Methodology does not match owner shift calculations with actual economic shifts in ownership and, as a result, inevitably leads to results that are under and over inclusive -- sometimes favorable to taxpayers and sometimes adverse to taxpayers. Second, results under the PLR Methodology are counter-intuitive in that ownership percentages do not necessarily add up to 100 percent, making it difficult to estimate whether the methodology has been correctly applied. Third, the PLR Methodology (whether Alternative 1 or Alternative 2) would be complicated and difficult to apply in some circumstances.

In the view of a significant minority of the Committee, these concerns illuminate a fundamental policy decision facing the government; while an expansive approach intuitively more closely aligns the owner shift calculation with anti-loss trafficking policies (i.e., deterrence of tax-motivated acquisitions) and avoids the conceptual discontinuities of the intermediate approach, it inherently is complicated in its application, particularly to the extent that (i) it attempts to distinguish precisely between the portion of the shift attributable to share movements and the portion attributable to value fluctuation, and (ii) it has needed safeguards to prevent abuse. The majority believes that the complications are manageable.

The remainder of the discussion below is designed to inform the government's view of matters that should be addressed in developing a workable, broad approach to factoring out fluctuations in value.

The fact that the PLR Methodology does not result in the application of the Section 382 limitations when significant economic changes in ownership occur is inconsistent with the purpose of Section 382. Investors make present purchasing decisions based on current ownership percentages rather than their hypothetical ownership percentage had there been no market fluctuation. Thus, basing owner shift analysis on hypothetical ownership percentages may run counter to the purpose of Section 382 in certain circumstances.

It is possible that concerns about under-inclusiveness could be addressed by adopting an anti-abuse rule under which appropriate adjustments would be made to the general rule factoring out value fluctuations if shareholders engage in transactions with a principal purpose to avoid the purposes of Section 382(l)(3)(C) (i.e., a series of transactions that, in their entirety, cause an economic shift in share ownership significantly greater than the shift measured under the general value fluctuation rule). Adoption of a subjective anti-abuse rule, however, would be contrary to the Congressional purpose of providing objective rules for determinations under Section 382 and would add further uncertainty to the owner shift analysis, which currently is generally done based on objective rules.³⁷ Accordingly, we believe that the government should proceed cautiously in regard to adopting subjective anti-abuse rules in providing rules for value fluctuations and adopt such rules only if abusive results cannot be weeded out with objective exceptions.

As the following example demonstrates, the PLR Methodology can also be over-inclusive in certain cases and would trigger an ownership change when otherwise a change would not have occurred under the general operation of Section 382(g).

On the one hand, as a conceptual matter, it may be appropriate to factor out fluctuations in value that occurred prior to the testing period. Factoring out such fluctuations would be consistent with the fact that the testing period is a rolling three-year period. Accordingly, as a general matter, owner shift calculations must go back to the beginning of the corporation's existence as a loss corporation (but not earlier than 1986) to determine whether an ownership change has occurred in any particular testing period. Applying a rolling shift calculation would be inconsistent with applying a different convention for factoring out fluctuations in value. For instance, failing to factor out fluctuations in value that occurred prior to the testing period in question arbitrarily may cause an ownership change to occur during the current testing period that is, in fact, attributable to such fluctuations; as one example, there may have been significant value fluctuations during the period between (a) the last testing date that occurred prior to the beginning of the testing period, and (b) the beginning of the testing period. Moreover, factoring out value fluctuations only from the beginning of the testing period would necessarily require that relative valuations of classes be determined on every date because the beginning of the testing period constantly changes.

On the other hand, factoring out fluctuations in value for periods prior to the testing period would be complicated as a practical matter. A loss corporation potentially would be required to determine relative valuations and perform complicated calculations demanded by the PLR Methodology on a retroactive basis going back to as early as when it became a loss corporation (or perhaps 1986). This would impose a substantial fact-gathering and recordkeeping burden on taxpayers and inject additional uncertainty into the determination of whether a loss corporation has, in fact, experienced an ownership change. At a minimum, this seems inconsistent with other rules in Section 382 that were designed to limit the amount of information that needs to be gathered and documented in order to calculate the cumulative owner shift.³⁹

The following example illustrates the issue.

The following example illustrates the application of the Common Share Equivalent Methodology.

Because the common share equivalent methodology "fixes" the exchange ratio of preferred stock to common at the former's issuance, this methodology does not reflect relative changes in the value of share classes and does not match Section 382 ownership changes with changes in economic ownership. As shown by Example 12a, this mismatch occurs when the exchange ratio of share classes is set upon initial issuance, and subsequently one class increases in value significantly relative to the other. This mismatch may favor a taxpayer or the government in particular circumstances.

Moreover, because of the mismatch between assumed and actual valuations, the Common Share Equivalent Methodology is subject to abuse. Without revaluation, a company whose common share value has increased significantly over its preferred shares can issue new common at the former common-to-preferred ratio, thus obtaining significant capital infusions without triggering a Section 382 ownership change, as demonstrated by Example 12c.

Periodic revaluations would be more accurate than the basic Common Share Equivalent Methodology and thus less subject to abuse. However, it would be subject to abuse within the period between revaluations. For example, in the current marketplace, values could fluctuate substantially within a one-year period and the Periodic Revaluation Methodology would allow large issuances of shares of a pre-existing class within the year that would not be captured unless there were a testing date following the next revaluation date.

At the same time, the Periodic Revaluation Methodology would not factor out fluctuations in value as effectively as the PLR Methodology. In many cases, the Periodic Revaluation Methodology would only defer value fluctuations from causing an ownership until the occurrence of a testing date following the next revaluation date.

For these reasons, we believe that the Periodic Revaluation Methodology should not be adopted as a general solution to the fluctuation in value problem, at least at the current time. However, adopting a reasonably narrow version of the Periodic Revaluation Methodology may be appropriate to ease the administrative burden of determining the value of separate classes of loss corporation stock where testing dates on which small amounts of stock are transferred occur frequently, as in the case of small issuances of corporate stock pursuant to the exercise of employee stock options. As an administrative matter, it may be appropriate to assume that the value relationship between two classes of stock that are not publicly traded, or between one publicly traded class and one class that is not publicly traded, remains constant for all testing dates during the same month, assuming that the only transactions that occur during that period are transactions that would qualify as small issuances under the small issuance exception of Reg. § 1.382-2T(j)(2)(iii).

V. CONCLUSION

In this Report, we have principally attempted to elucidate the difficult decisions that the Service and the Treasury will confront in issuing final guidance under Section 382(l)(3)(C) through analyzing the policy considerations underpinning Section 382 and by highlighting certain conceptual and practical issues that arise under the various approaches to factoring out fluctuations in value. In the end, however, there is no one perfect solution for whether, the extent to which, and how fluctuations in relative value should be disregarded in calculating owner shifts under Section 382. It is clear that trade-offs will be required to be made (e.g., between complexity and precision). But there can be no doubt that it is important for the Service and the Treasury to bring clarity to this uncertain area of practice.

² In our experience, some practitioners follow the principles expressed in the PLRs (the "PLR Methodology"). Even within this population, different variations are often employed. Other practitioners believe that the principles of the PLRs can be relied on only if a ruling is received by the particular taxpayer. These practitioners generally do not factor out fluctuations in value in measuring owner shifts. To complicate matters even more, for certain taxpayers, their tax advisor may take a different view of the issue than their financial auditor.

³ While not recommending that the PLR Methodology in its current form be adopted in final guidance, this Report does not question the validity of rulings in which the Service has applied this methodology on a case-by-case basis. The PLR Methodology represents a reasonable interpretation of current law as applied to the facts of each PLR. If the Service chooses to issue final guidance similar to the PLR Methodology, we believe that the technical issues and antiabuse concerns outlined in this Report ought to be addressed in order to assist taxpayers in applying the PLR Methodology appropriately.

⁴ Section 382(b)(1). An annual limitation is also imposed by Section 383 on the utilization of certain excess credits and net capital loss carryovers following an ownership change. Further, Section 384 prohibits the utilization of pre-acquisition loss of one entity from offsetting built-in gain of another entity. Still other limitations exist, such as Sections 269 and 482, which impact the availability of tax attributes. However, a thorough discussion of the nature of the limitations under Sections 269, 382, 383, 384, and 482 is beyond the scope of the Report.

⁵ Section 382(g)(1), (i)(1). The testing period will be less than three years, however, if the loss corporation had an ownership change during that period, in which case the testing period for a subsequent ownership change will not begin before the first day following the date of the most recent ownership change. Section 382(i)(2) and Reg. § 1.382-2T(d)(2).

⁶ Section 382(k)(6)(C), (7); Reg. § 1.382-2T(g). For this purpose, shareholders who directly own less than 5 percent of the loss corporation stock are aggregated and treated as a single 5-percent shareholder, referred to as a "public group." Reg. § 1.382-2T(g)(1)(ii), (j)(1), (j)(2)(ii). Similarly, if an entity has a direct or indirect ownership of 5 percent or more of the loss corporation stock, owners of that entity who (by reason of such ownership) are attributed a less-than-five percent interest are aggregated into a separate public group. Reg. § 1.382-2T(g)(1)(iii), (j)(1)(iv)(C). Additional public groups may be created by certain types of transactions. See Reg. § 1.382-2T(j)(2)(iii)(B), (D).

⁷ Reg. § 1.382-2T(e)(1)(i). The temporary regulation provides a non-exclusive list of five transactions that constitute owner shifts: (1) a purchase or disposition of loss corporation stock by a 5-percent shareholder; (2) a Section 351 exchange that affects the percentage of stock owned by a 5-percent shareholder; (3) a redemption or a recapitalization that affects the percentage of stock owned by a 5-percent shareholder; (4) an issuance of loss corporation stock that affects the percentage of stock owned by a 5-percent shareholder, and (5) an equity structure shift that affects the percentage of stock owned by a 5-percent shareholder.

⁸ Section 382(g)(3).

⁹ Reg. §§ 1.382-2(a)(4), -2T(e). A loss corporation is any corporation that: (i) is entitled to use a net operating loss carryover, a capital loss carryover, a carryover of excess foreign taxes under Section 904(c), a carryover of general business credit under Section 39, or a carryover of minimum tax credit under Section 53 (each, a "tax attribute"); (ii) has a tax attribute for the taxable year that includes a testing date; or (iii) for the taxable year that includes a testing date has a net unrealized built-in loss. Reg. § 1.382-2(a)(1).

¹⁰ Reg. § 1.382-2(a)(3)(i).

¹¹ Reg. § 1.382-2(a)(3)(i). Section 1504(a)(4) provides that the term "stock" does not include any stock which (A) is not entitled to vote, (B) is limited and preferred as to dividends and does not participate in corporate growth to any significant extent, (C) has redemption and liquidation rights which do not exceed the issue price of such stock (except for a reasonable redemption or liquidation premium), and (D) is not convertible into another class of stock.

¹² Dec. 5, 2003.

¹³ PLR 200901001 (Sept. 30, 2008); PLR 200901003 (Sept. 12, 2008); PLR 200622011 (Feb. 2, 2006); PLR 200520011 (Feb. 18, 2005); PLR 200511008 (Dec. 6, 2004).

¹⁴ H.R. Rept. 99-426, pt. 3, at 256 (1985).

¹⁵ Bittker and Eustice, Federal Taxation of Corporations and Shareholders, ¶ 14.42[2] (7th Ed. 2006).

¹⁶Id.

¹⁷ H.R. Rept. 99-426, pt. 3, at 256 (1985).

¹⁸Id.

¹⁹Id.

²⁰ Mark Hoffenberg, Owner Shifts and Fluctuations in Value: A Theory of Relativity, 106 TNT 1446 (Mar. 21, 2005) (hereinafter referred to as "Hoffenberg").

²¹See Committee on Net Operating Losses, New York State Bar Association, Tax Section, Supplemental Report on Section 382, 88 TNT 42-37 (Feb. 22, 1988) (hereinafter "NYSBA Supplemental Report").

²²Id.

²³ Section 382(l)(3)(C).

²⁴ Reg. §§ 1.382-2(a)(3)(i), -2T(c)(1).

²⁵ NYSBA Supplemental Report at 159.

²⁶ Under Section 382(l)(5), if a loss corporation has an ownership change pursuant to a bankruptcy plan of reorganization, and if the loss corporation's pre-change shareholders and qualified creditors own at least 50 percent of the loss corporation immediately thereafter, no Section 382 limitation applies to the use of its pre-changes losses. However, the amount of such losses is reduced by redetermining its losses as if certain interest deductions were disallowed.

Under Section 382(l)(6), if a loss corporation has an ownership change pursuant to a bankruptcy plan of reorganization, and if Section 382(l)(5) does not apply to the ownership change, the Section 382 limitation applicable to the ownership change generally is determined by reference to the post-change equity value of the loss corporation, rather than its pre-change equity value. Reg. § 1.382-9(j).

²⁷ NYSBA Supplemental Report at 161.

²⁸Id.

²⁹ Taking the arguments made above one -- perhaps large -- step further, it can be argued that even if applied on a rolling basis and with a higher threshold, an overall de minimis exception would still be too narrow to effectuate the policy concerns of Section 382(l)(3)(C), and that de minimis rule could reasonably be applied on a shareholder-by-shareholder basis so that the rule takes into account only shareholders who seek to benefit from an increased ownership percentage through an affirmative act of acquiring additional shares. For instance, the de minimis transfer rule as formulated above would not "correct" the result reached in Example 2. In that example, P's preferred stock interest has increased to 80 percent of the value of LCorp when C sells all of the common stock to O. C's transfer of stock representing 20 percent of the value of LCorp would not be de minimis, and, therefore, would result in a testing date and a revaluation date. Under the overall de minimis rule described in the text, although the beneficiary of the value fluctuation (P) did not engage in a transaction to increase its ownership, the non-de minimis transfer by C would result in an aggregate 80 percentage point increase in the interests of LCorp 5-percent shareholders and, thus, an ownership change. Under a shareholder-by-shareholder version of the de minimis rule, O's interest in LCorp would be fully revalued as a result of purchasing 20 percent of LCorp, which far exceeds the rolling 5 percent de minimis threshold. On the other hand, P's percentage ownership in LCorp would not have increased by 5 percentage points or more as a result of actual share movements, so that P's ownership in LCorp would not be revalued. If, however, C sold its common stock to P rather than O, P's entire ownership position would be revalued. And if C sold no shares but P sold 50 percent of its preferred stock to O, it seems appropriate to take into account O's increase in ownership but not to revalue P's shares, since P has reduced its ownership of shares. This result seems consistent with the theory underlying the shareholder-by-shareholder approach -- i.e., that there is loss trafficking potential only where the shareholder benefiting from value fluctuation acquires more than a de minimis number of additional shares.

It should be noted, however, that a shareholder-by-shareholder de minimis approach is likely to suffer from a number of the complexities associated with a broad approach to Section 382(l)(3)(C), described below. Such a de minimis exception would result in different values being assigned to shares held across shareholders. Unlike the PLR Methodology, discussed below, all shares held by a single shareholder would have the same value. Nonetheless, the loss corporation would still have to track different share values for different shareholders. Moreover, share issuances and redemptions would have to be taken into account in determining the percentage shift represented by actual share movements (i.e., acquisitions) by each shareholder. Thus, qualification under the de minimis threshold would have to be retested and tracked even though the shareholder may not participate in the issuance or redemption resulting in a testing date. Further complexity would be introduced if actual share movements that increase a 5-percent shareholder's interest were netted against decreases. Finally, it may be necessary to provide an anti-abuse rule to prevent the purposes of the shareholder-by-shareholder de minimis rule from being avoided through the use of various mechanisms including the use of related parties. Cf. Reg. § 1.382-2T(k)(2) and (k)(4). Accordingly, while we offer the concept of a shareholder-by-shareholder de minimis rule for the government's consideration, we do not affirmatively recommend it.

³⁰ Dec. 5, 2003.

³¹ PLR 200901001 (Sept. 30, 2008); PLR 200901003 (Sept. 12, 2008); PLR 200622011 (Feb. 2, 2006); PLR 200520011 (Feb. 18, 2005); PLR 200511008 (Dec. 6, 2004).

³² Hoffenberg, 106 TNT 1446. Although the formula set forth in Hoffenberg is consistent with the PLR Methodology, it is not the only algorithm consistent with the PLRs.

³³Id.

³⁴ In addition, in certain circumstances, it appears that Alternative 1, by recalculating the percentage ownership on each testing date, may cause some shift to be associated with a shareholder who did not participate in any transaction occurring on the testing date.

³⁵ To protect the integrity of this approach, additional rules would be needed to take into account coordinated acquisitions made by other persons or acquisition structures implemented to avoid triggering redetermination of values. See, e.g., Treas. Reg. §§ 1.382-3(a)(1), 1.382-2T(k)(4).

³⁶ If O (rather than D) purchased the common stock from C, then there would be a 100 percentage point shift under the modification to the PLR Methodology proposed under the majority's view.

³⁷But see Reg. §§ 1.382-2T(k)(4), and 1.382-4.

³⁸ PLR 200901001 (Sept. 30, 2008) was the only ruling that did not contain the no opinion language related to fluctuations occurring before the testing period.

³⁹Cf. Reg. § 1.382-2T(k)(1) (permitting the loss corporation to rely on the existence or absence of certain securities law filings).

⁴⁰Cf. Reg. § 1.382-2T(j)(2)(vi) (providing for a pro ration of redeemed shares across public groups).

⁴¹ Reg. § 1.382-3(j)(3).

⁴² September 30, 2008.