NYSBA Tax Section Suggests Changes to Proposed Regs on Corporate Equity Reduction Transactions

We are pleased to submit the attached report of the Tax Section of the New York State Bar Association regarding the proposed corporate equity reduction transaction ("CERT") regulations issued on September 13, 2012 (the "Proposed Regulations").

The CERT rules were enacted to prevent the use of a net operating loss ("NOL") carrybacks to finance leveraged transactions. Under current law, if a corporation engages in a major stock acquisition ("MSA") or an excess distribution ("ED"), the portion of the NOL incurred in a "loss limitation year" ("LLY") attributable to interest related to the CERT (the corporate equity reduction interest loss or "CERIL") cannot be carried back to any taxable year prior to the year of the CERT. As summarized below, this Report provides a number of comments and recommendations on several technical and policy issues raised by the Proposed Regulations.

The Proposed Regulations consist of Prop. Reg. §§ 1.172(h)-0 through 1.172(h)-5, Prop. Reg. § 1.1502-72, and certain proposed amendments to existing Reg. § 1.1502-21. As summarized below, we generally support the Proposed Regulations, which provide much-needed guidance on the CERT provisions in the Code. The attached report provides the following comments and recommendations on several technical and policy issues raised by the Proposed Regulations:

1. The Proposed Regulations provide that a stock-for-stock acquisition is a CERT. Because such acquisitions generally do not reduce corporate equity, we recommend that final regulations exclude them from the CERT limitations. One way to do so would be to exclude the stock portion of the consideration as a CERT cost. As an alternative to excluding stock consideration as a CERT cost, final regulations could exclude stock-for-stock acquisitions from the definition of an MSA (except to the extent of any boot). If desired, to achieve parity among economically equivalent transactions, final regulations could also adopt correlative rules that: (i) decrease the pool of costs treated as arising from a CERT ("CERT costs") by the proceeds of any share issuance made in connection with a cash acquisition, and (ii) increase CERT costs by the costs of any share redemption made in connection with a stock-for-stock acquisition. To avoid the complexity associated with developing rules to establish when a share issuance or redemption is made "in connection with" a CERT, final regulations could provide that any such issuance or redemption will be integrated with a CERT only if it occurs within six months of the CERT.

2. The Code expressly exempts stock acquisitions treated as asset acquisitions under section 338 from the definition of a CERT, thereby clearly evidencing Congressional intent to exclude asset acquisitions. An asset acquisition may nevertheless be treated as a CERT in certain circumstances (e.g., a tax-free merger may be considered an ED of the target corporation succeeded to by the acquiring corporation). Given the economic similarity between tax-free asset and stock acquisitions, final regulations could include tax-free asset acquisitions in the definition of a CERT, but exclude the stock consideration as a CERT cost. Alternatively, if final regulations exclude tax-free asset acquisitions entirely, we recommend that they clarify that the actual or deemed distribution by the target corporation of the consideration received from the acquiring corporation is not an ED.

3. The Proposed Regulations provide that a CERT includes transactions qualifying under section 351. Because these transfers do not reduce the equity of the corporate transferor, we recommend that final regulations exclude them from the definition of a CERT.

4. The Proposed Regulations also treat a tax-free spin-off under section 355 as a CERT. Under one view of these transactions, a spin-off should not be treated as a CERT because it merely divides corporate equity into smaller pieces without reducing aggregate corporate equity (except to the extent of any boot). Under an alternative view of these transactions, a spin-off should be treated as a CERT because it separates the leverage (whether existing or newly incurred) from the historic income pool of the distributing and controlled businesses, which may generate NOL carrybacks in the more leveraged entity. If final regulations follow the Proposed Regulations, we recommend that they modify the apportionment methodology to allocate a larger percentage of the CERT costs to the more leveraged entity.

5. We recommend that final regulations clarify that a stock distribution under section 305 (whether taxable or tax-free) is not an ED. Such stock distributions merely divide corporate equity into more units and do not reduce aggregate corporate equity.

6. We support treating a distribution by a target corporation as part of an MSA where the distribution and the MSA are part of the same plan, regardless of whether such distribution also constitutes an ED if tested independently. We recommend that final regulations clarify that a redemption of acquiring corporation stock in connection with an MSA is not considered to be part of the same integrated plan as the MSA, so that such redemption is tested separately as an ED.

7. Under current law, preferred stock described under section 1504(a)(4) is disregarded for purposes of determining whether an ED has occurred (i.e., the redemption of such stock is not an ED and the issuance of such stock does not offset other distributions). We suggest that Treasury and the IRS consider expanding the definition of preferred stock to include voting preferred stock that is otherwise described in section 1504(a)(4). Alternatively, consideration could be given to abandoning the definition of preferred stock under section 1504(a)(4) in favor of the definition under section 351(g), which we believe more precisely captures the type of preferred stock that Congress intended to exclude from the ED rules.

8. The Proposed Regulations include a special rule treating borrowing costs on debt that facilitates an MSA or ED as CERT costs. We recommend that final regulations treat borrowing costs on all debt (whether or not it facilitates a CERT) as interest expense for this purpose.

9. The Code provides that indebtedness is allocated to a CERT in the manner prescribed under section 263A(f)(2)(A)(ii), but without regard to section 263A(f)(2)(A)(i), which provides for tracing of certain debt (the "Avoided Cost Method"). We generally support the Avoided Cost Method and do not recommend the adoption of a tracing methodology as a general alternative. To mitigate the occasional harshness of the Avoided Cost Method, however, we recommend that Treasury and the IRS explore whether an administrable rule could be devised that would permit tracing in limited circumstances (e.g., when a target corporation with debt outstanding joins the group in an MSA), or that would adopt another presumption (such as pro rata allocation of sources of funds to expenditures).

10. If a taxpayer experiences an unforeseeable extraordinary adverse event (an "Unforeseeable Event") following a CERT, debt is first allocated to costs associated with such event, rather than to the CERT. We recommend that any regulatory definition of Unforeseeable Event adopt an objective standard and be limited to events (i) with a low probability of occurrence (no more than one-in-ten) at the time of the CERT; (ii) that arise from circumstances beyond the taxpayer's control; and (iii) that create an unexpected need for excess expenditures or an unexpected loss in revenue.

11. Under current law, the amount of the CERIL for any LLY is capped (the "CERIL cap") at the excess (if any) of (i) the amount allowable as an interest deduction during the LLY, over (ii) the average of such amounts for the three taxable years preceding the taxable year of the CERT. The Proposed Regulations provide that in the event of a short LLY, the three-year average is prorated by the number of days in the short LLY. Furthermore, if a corporation was not in existence for three taxable years preceding the taxable year of the CERT (the "interest lookback period"), the corporation is deemed to have been in existence for periods during which it accrued no additional interest. To avoid understating the historic interest expense of such corporations, we recommend an adjustment comparable to the adjustment made for a short LLY.

12. The preamble to the Proposed Regulations requests comments on how final regulations should factor out the impact of fluctuations in interest rates following a CERT. We recommend that final regulations adjust the three year interest history (and correspondingly the CERIL cap) upwards or downwards based on changes in the applicable federal rates. In effect, the limitation would be multiplied by a fraction, the numerator of which would be the average of the federal mid-term rates for the LLY and the denominator of which would be the weighted average of the applicable mid-term federal rates for the interest lookback period. The CERIL cap would therefore be increased if the fraction is greater than one and reduced if the fraction is less than one.

13. The Proposed Regulations adopt single entity rules for consolidated groups. If a corporation joins a consolidated group, the group inherits the CERT attributes of that member. When a corporation departs from a consolidated group, the CERT attributes of the group are generally apportioned to the departing member based on relative net fair market values. We believe consideration should be given to other allocation methods, such as the relative gross fair market value of assets or the relative amount of indebtedness of the departing and remaining members. We also recommend that the group's distribution and share issuance histories be apportioned to the departing member based upon the relative net fair market values of the departing and remaining members.

14. The Proposed Regulations disregard intercompany transactions between group members unless they occur pursuant to a plan in which a party to the transaction becomes a non-member. We recommend that final regulations generally eliminate the "pursuant to a plan" exception.

15. A departing member may make a unilateral election to relinquish the carryback of all NOLs to taxable years of its former consolidated group and any preceding years. In such a case, none of the CERT history or attributes of the group are apportioned to the departing member. In the absence of such an election, the required apportionment of CERT attributes to the departing member will compel selling groups to provide substantially more information to the departing member. We nevertheless recommend that final regulations retain the default rule because it is consistent with the default rule for other carryback elections.

16. Finally, we recommend that taxpayers be permitted to apply the final regulations retroactively, providing that they do so consistently for all CERTs undertaken prior to the effective date of the final regulations.

We appreciate your consideration of our recommendations. If you have any questions regarding this report, please feel free to contact us and we will be glad to discuss or assist in any way.

cc:

This Report¹ provides comments to the Treasury Department ("Treasury") and the Internal Revenue Service (the "Service") regarding proposed regulations issued under the corporate equity reduction transaction ("CERT ") provisions of the Code on September 13, 2012.² The CERT provisions, which are set forth in section 172(b)(1)(E) and (h), limit the carryback of all or portion of a net operating loss ("NOL") in the taxable year of the CERT and certain subsequent years. The proposed regulations (REG-140668-07) consist of Prop. Reg. §§ 1.172(h)-0 through 1.172(h)-5, Prop. Reg. § 1.1502-72, and certain proposed amendments to existing Reg. § 1.1502-21 (collectively, the "Proposed Regulations"). As described in detail below, we generally support the Proposed Regulations, which provide important guidance on the CERT provisions of the Code.

II. Summary of Recommendations

The CERT rules were enacted to prevent the use of NOL carrybacks to finance leveraged transactions. Under current law, if a corporation engages in a major stock acquisition ("MSA") or an excess distribution ("ED") (i.e., a CERT), the portion of any NOL incurred in a "loss limitation year" ("LLY") attributable to interest related to the CERT (the corporate equity reduction interest loss or "CERIL") cannot be carried back to any taxable year prior to the year of the CERT.³ As summarized below, this Report provides a number of comments and recommendations on several technical and policy issues raised by the Proposed Regulations.

If there is a CERT and an "applicable corporation" ("Applicable Corporation") has a CERIL for any LLY, then the portion of an NOL for the LLY attributable to the CERIL cannot be carried back to any taxable year prior to the year of the CERT.¹¹ Any portion of an NOL that cannot be carried back due to the operation of the CERT rules may be carried forward.¹²

An MSA is the acquisition by a corporation pursuant to a plan of such corporation (or any group of persons acting in concert with such corporation) of stock in another corporation representing 50% or more (by vote or value) of the stock in such other corporation.¹⁴ All acquisitions during any 24-month period are treated as pursuant to one plan.¹⁵ An MSA does not include a qualified stock purchase to which an election under section 338 applies.¹⁶

An ED is the excess (if any) of the aggregate distributions (including redemptions) made during a taxable year by a corporation with respect to its stock over the greater of (i) 150% of the average (the "three-year distribution average") of such distributions during the three taxable years immediately preceding such taxable year (the "distribution lookback period"), or (ii) 10% of the fair market value of the stock of such corporation as of the beginning of such taxable year.¹⁷ The amount of aggregate distributions and the three-year distribution average are each reduced by the aggregate amount of stock issued by the corporation during the applicable period in exchange for money or property other than stock in the corporation.¹⁸ A similar rule is not provided for MSAs. For purposes of the ED rules, "plain vanilla" preferred stock described in section 1504(a)(4) and distributions (including redemptions) with respect to such stock are disregarded.¹⁹ A similar rule is not provided for MSAs.

Allocable Interest Deductions are deductions allowed for interest on the portion of any indebtedness allocable to a CERT.²³ In general, except as provided in Regulations, indebtedness is allocated to a CERT pursuant to the Avoided Cost Method.²⁴ The Avoided Cost Method generally allocates indebtedness to a CERT to the extent that a corporation's indebtedness would have been reduced if amounts expended for CERTs during the applicable measurement period instead were used to repay debt. Allocable Interest Deductions essentially are equal to the allocable debt times the weighted average interest of corporate indebtedness for the measurement period. This means that the debt of a corporation is allocated first to the CERT, regardless of whether the MSA or ED was actually financed with debt.

Allocable Interest Deductions for any LLY cannot exceed the excess (if any) of (i) the amount allowable as a deduction for interest paid or accrued by the taxpayer during the LLY over (ii) the three-year interest average.²⁵ That is, only the amount of interest that exceeds the taxpayer's "normal" interest deduction is potentially subject to the CERT limitation.²⁶

If an Unforeseeable Event occurs during an LLY but after the CERT, indebtedness is allocated in the manner described in section 172(h)(2)(B) to unreimbursed costs paid or incurred in connection with such event before being allocated under the Avoided Cost Method to the CERT.²⁷ In other words, corporate indebtedness (and interest deductions) are allocated first to expenditures associated with an Unforeseeable Event. The Current Rules do not define an Unforeseeable Event.

For a consolidated group, the CERT limitation applies to the consolidated net operating loss ("CNOL") and the amount of the CERIL is determined on a consolidated basis³² Any change in the composition of the group potentially affects the status of the group with respect to prior CERTs and a CERT occurring as a result of a corporation joining or leaving the consolidated group (e.g., the computation of the group's CERIL cap). Neither the statute nor the legislative history provides clear guidance regarding the precise impact of such changes.

The Proposed Regulations also provide guidance regarding the determination of an ED when the year of the potential ED is a short taxable year. In such a case, the distribution history with regard to any year of the taxpayer during a distribution lookback period equals (i) the amount of distributions made during any year of the taxpayer ("lookback period year") in the distribution lookback period, (ii) multiplied by a fraction the numerator of which equals the number of days in the short taxable year of the potential ED and the denominator of which equals the number of days in the distribution lookback period year.⁴⁰ Under this rule, the 150% "safe harbor" provided to taxpayers is also appropriately scaled so that taxpayers do not achieve an unfair advantage in comparing a 12-month average distribution history with a distribution made in a short period. In addition, for purposes of determining any three-year distribution average of a successor, the distributions made by a successor include distributions made by its predecessors.⁴¹

Before the Proposed Regulations were issued, there was debate among practitioners concerning whether the definitions of MSA and ED under the Current Rules apply (or should apply) to nonrecognition transactions.⁴² The consideration provided in a nonrecognition transaction typically consists of stock of the acquiring corporation and potentially some boot. Since there is little cost associated with issuing stock, the acquiring corporation is not required to incur additional debt to fund such acquisitions except as may be necessary to finance the payment of boot (if any). The Proposed Regulations nevertheless provide that such transactions may constitute a CERT and must be tested under the CERT rules regardless of whether gain or loss is recognized by any party.⁴³ Thus, a distribution that qualifies for tax-free treatment under section 355 is tested as a potential ED (or part of a potential ED), and the acquisition by a corporation of 50% or more of the stock of another corporation in a transaction meeting the requirements of sections 351, 368(a)(1)(A) and (a)(2)(E), or 368(a)(1)(B) constitutes an MSA.⁴⁴

The Preamble notes that, under the Proposed Regulations, Allocable Interest Deductions are computed by multiplying the "weighted average interest rate" by "average excess expenditures" (as defined in Reg. § 1.263A-9(c)(5)(ii) and (iii)).⁴⁸ The Preamble acknowledges that section 263A contemplates transactions that are very different in nature from CERTs and that it is therefore often difficult to identify the costs associated with a CERT that are analogous to average excess expenditures. Accordingly, the Proposed Regulations ameliorate this difficulty by providing a method for identifying CERT costs.⁴⁹ The interest allocable to those CERT costs is then computed under the Avoided Cost Method, which does not entail the tracing of debt to particular expenditures.⁵⁰

In the case of an MSA, CERT costs include the fair market value of the stock acquired, whether such stock is acquired in exchange for cash, for stock of the acquiror, or for other property.⁵¹ The Preamble justifies the treatment of stock as a CERT cost as necessary to ensure that stock-for-stock acquisitions are treated in the same manner as cash acquisitions funded by the issuance of stock. In addition, CERT costs include the fair market value of any distributions to shareholders that are treated as part of the MSA.⁵² In the case of an ED, CERT costs include the fair market value of any distributions to shareholders during the year of the CERT.⁵³

An important clarification in the Proposed Regulations is the treatment of expenses incurred to facilitate a CERT. Under the Proposed Regulations, CERT costs of an MSA and an ED include the sum of all amounts paid or incurred to facilitate any step of the MSA or ED, respectively, to the extent those amounts are required to be capitalized under section 263(a), and any amounts disallowed under section 162(k).⁵⁴ In addition, the Proposed Regulations provide that borrowing costs of debt incurred to facilitate a CERT are CERT costs.⁵⁵

With regard to the CERIL cap, the Proposed Regulations include new rules for short LLYs. For purposes of computing the three-year interest average, if the relevant LLY is less than a full 12 months, the interest paid or accrued with regard to any lookback period year in the interest lookback period equals (i) the amount of interest treated as paid or accrued multiplied by (ii) a fraction, the numerator of which equals the number of days in the short LLY, and the denominator of which equals the number of days in the lookback period year.⁵⁶ The value of the fraction may not exceed 100% (i.e., where a lookback period year is shorter than the short LLY).

The Current Rules do not address the impact of incomplete lookback periods. The Proposed Regulations provide that, if an Applicable Corporation was not in existence for the full lookback period, its lookback period is deemed to have additional 12-month periods as necessary to create a full look-back period.⁵⁷ However, the Applicable Corporation is treated as not having paid or accrued any interest during the years in which it was deemed to be existence.⁵⁸

The Proposed Regulations do not provide guidance on the definition of an Unforeseeable Event but rather "reserve" on this point.⁵⁹ In the Preamble, however, Treasury and the Service request comments regarding whether rules are necessary and, if so, what type of events should constitute Unforeseeable Events.

The Proposed Regulations also provide important guidance as to how the CERT rules operate when members join and depart a group. In general, when a new member joins a consolidated group, its CERT attributes fold into the group, and the new member ceases to be separately tracked as an Applicable Corporation. The CERT rules thereafter apply to the group, rather than to the new member, with regard to any CERT for which the member had been an Applicable Corporation, including an MSA in which the member was acquired by the group.⁶¹ When a member leaves the group, the departing member generally shares the group's CERT history and takes with it a portion of the group's CERT attributes.

For purposes of determining whether a group has made an ED during any consolidated return year ("potential ED year"), distributions by all members of the group to non-members during the potential ED year are aggregated.⁶³ Distributions between members of the same group are generally disregarded for purposes of applying the CERT rules unless a party to the transaction is deconsolidated pursuant to the same plan.⁶⁴

Another issue addressed by the Proposed Regulations is the impact of stock issuances by a member of the group. Stock issued by a member of a group is taken into account only if the stock is issued to a non-member (e.g., as an offset to distributions by the group). Intercompany stock issuances are disregarded except where a party to the transaction ceases to be a member as part of the same plan.⁶⁵

New members. The computation of the group's three year distribution average includes (i) distributions made to non-members by each member during the potential ED year, (ii) distributions made by predecessors of those members, and (iii) certain distributions made by corporations during separate return years.⁶⁶ If a corporation was a member of a prior group during a portion of a distribution lookback period, the distribution history of that corporation during taxable years of the prior group generally includes only distributions made by that corporation to non-members of the prior group.⁶⁷ If any member is included in the group for less than an entire potential ED year, then only a pro rata portion of the partial-year member's distribution history is included.⁶⁸

Departing members. If a corporation deconsolidates from a former group, the corporation's actual distribution history is subtracted from the former group's distribution history and is available to the departing member for purposes of computing any three year distribution average following the deconsolidation.⁶⁹ That is, the departing member takes only its own distribution history and is not allocated a portion of the group's overall distribution history.

New members. If a corporation that is an Applicable Corporation with regard to a CERT occurring in a separate return year joins a consolidated group ("pre-existing CERT member"), the group is treated as a single Applicable Corporation with regard to that CERT in the consolidated return year of the acquisition and any succeeding year.⁷⁰

Departing members. If a corporation deconsolidates from a group that is treated as an Applicable Corporation with regard to a CERT, the departing member and the former group are both treated as Applicable Corporations following the deconsolidation.⁷¹ Note that the departing member may be treated as an Applicable Corporation with regard to a CERT of a former group only if the group engages in the CERT, or a pre-existing CERT member joins the group, on or before the date of the deconsolidation.⁷²

New members. For purposes of determining the group's CERIL cap, the group's three-year interest average includes the interest history of all of its members (and their predecessors); thus, new members contribute to the group's three-year interest average.⁷³ With respect to a new member that had a CERT before it joined the group, any CERT costs incurred by such member during separate return years prior to the acquired corporation's inclusion in the group are treated as having been incurred by the acquiring group.⁷⁴

Departing members. A portion of a group's CERT costs and interest history is allocated and apportioned to a departing member if the group engages in the relevant CERT, or a pre-existing CERT member joins the group, on or before the date of the deconsolidation.⁷⁵ The portion of the group's total CERT costs and interest history that is allocated to a departing member equals (i) the group's total CERT costs, or interest paid or accrued,⁷⁶ in each consolidated year multiplied by (ii) a fraction, the numerator of which equals the value of the departing member immediately after its deconsolidation, and the denominator of which equals the value of the entire group immediately prior to the deconsolidation.⁷⁷ The allocated and apportioned interest is subtracted from the group's interest history and is unavailable to the group (or any other group member) for purposes of computing a three-year interest average of the group (or any other group member) after the year of deconsolidation.⁷⁸

As described above, in applying the CERT rules, intercompany transactions are generally disregarded under the Proposed Regulations.⁷⁹ For example, interest expense attributable to an intercompany obligation is not taken into account in computing the CERIL or three-year interest average of a group.⁸⁰ However, a transaction between group members is not disregarded if a party to the transaction becomes a non-member pursuant to the same plan or arrangement.⁸¹ In such a case, any transaction between group members, including a potential ED, is tested on a separate entity basis under the CERT rules.⁸²

If a member departs from a group, the departing member may make an irrevocable election to relinquish the carryback of all NOLs (and attributable portions of CNOLs) to taxable years of the former group and any preceding years.⁸³ This Carryback Waiver Election is binding on the departing member and any group of which it may become a member.⁸⁴ If the former member becomes a member of another group immediately after its deconsolidation from the former group, the Carryback Waiver Election is available only to the common parent of the acquiring group.⁸⁵

If a Carryback Waiver Election is made, the default rules related to departing members of a consolidated group discussed above do not apply. The departing member is not treated as an Applicable Corporation with regard to any CERT of the former group, none of the former group's CERT costs are apportioned to the departing member, and the departing member is treated as having paid or accrued zero interest during the period of its inclusion in the former group and preceding years.⁸⁶ The group retains the interest history that would otherwise be allocated and apportioned to the departing member under Prop. Reg. § 1.1502-72(d)(3)(ii).⁸⁷ However, even if a Carryback Waiver Election is made, the departing member retains its distribution history.

IV. Discussion of Recommendations

The Proposed Regulations clarify the application of the CERT rules in both the separate and consolidated return contexts. In particular, the Proposed Regulations resolve a number of ambiguities and fill in numerous gaps under the Current Rules in the separate return context and provide a thoughtful framework for integrating these separate return rules with consolidated return principles. While we do not agree with all aspects of the Proposed Regulations, there is much to commend them, and we applaud the government for providing much needed guidance in this area. Our focus below is on matters where we feel that further consideration of other alternatives is warranted.

As a general matter, final Regulations could adopt one of two competing approaches to the statute. The first would support broad final Regulations that reconcile many of the statutory inconsistencies under the Current Rules. This approach would reflect the view that Congress intended Treasury to exercise its broad regulatory authority to construct a reasonable set of rules to "smooth out" the various inconsistencies of a sweeping statute to target only those acquisitions and distributions identified in the legislative history that reduce corporate equity. The second approach would nevertheless pursue a more modest set of objectives. As such, it would leave the basic statutory framework largely intact, due in substantial part to the difficulty of discerning Congressional intent on a number of fundamental issues.

Because the Tax Section did not reach a consensus view on which of these approaches should inform any future guidance, this Report often presents different alternatives for consideration rather than consensus recommendations.

Simply put, a stock-for-stock acquisition does not reduce corporate equity and thus does not implicate the concern most clearly animating the CERT rules (i.e., funding an acquisition through carrying back losses generated from interest deductions attributable to debt which itself effectively funds the acquisition). In recognition of this point, the legislative history to the CERT provisions contemplates that Treasury and the Service could exclude tax-free stock-for-stock acquisitions from the definition of a MSA by Regulation.⁸⁸ As a policy matter, the exclusion of stock-for-stock acquisitions should not depend upon whether the acquisition is taxable or tax-free. In either case, the transaction does not reduce corporate equity.

The Preamble nevertheless justifies the treatment of stock-for-stock acquisitions as MSAs and stock consideration as a CERT cost on parity grounds: because a stock-for-stock acquisition is not fundamentally different from a cash acquisition funded by an issuance of stock, both acquisitions should be treated as MSAs. We agree that achieving parity is a reasonable objective, provided it can be achieved in an administrable fashion. However, we do not believe that achieving parity mandates that the CERT rules apply to all stock-for-stock acquisitions. We therefore recommend that final Regulations exclude stock-for-stock acquisitions from the CERT limitations.

As discussed below, relief from the CERT limitations could be achieved by adopting one of the following alternatives. First, the government could retain the rule that MSAs include stock-for-stock acquisitions (whether taxable or tax-free), but exclude any consideration paid in the form of stock from the definition of CERT costs. Boot would therefore continue to be considered a CERT cost. Under this approach, stock-for-stock acquisitions would generally not be subject to a CERT limitation. Second, a narrower alternative would be to exclude taxable and tax-free stock-for-stock acquisitions from the definition of an MSA. This approach would address our basic concern about subjecting such acquisitions to the CERT limitations. It could be simpler than the first approach, but ultimately may involve its own complexities. For example, the government may conclude that it is necessary or appropriate to create special rules for transactions involving a mix of stock and other consideration. In such a case, the cash or property component would be equity-reducing. If the consideration paid were composed of slightly more stock than cash, would the acquisition be entirely excluded from MSA status, or would the boot still constitute a CERT cost?

The Tax Section did not reach a consensus view on whether final Regulations should adopt certain of the correlative rules, described below, which could be applied in addition to the first or second approach for dealing with stock-for-stock acquisitions. To achieve symmetry among economically similar transactions, the correlative rules would (i) decrease CERT costs by the proceeds of any equity offering undertaken by the acquiror in connection with a cash acquisition and (ii) increase CERT costs by the costs of any share redemption undertaken by the acquiror in connection with the stock-for-stock acquisition. Under these correlative rules, a cash acquisition funded through an issuance of stock for cash would be treated in the same manner as a pure stock-for-stock acquisition (i.e., the CERT limitations effectively would not apply in either case). In addition, a stock-for-stock acquisition coupled with a share redemption would be treated in the same manner as a cash acquisition (i.e., the CERT limitations would apply in either case).⁸⁹

If final Regulations were to adopt both components (i.e., one of the two approaches, together with the correlative rules), they would (i) exclude from the CERT limitations a class of transactions that do not reduce corporate equity, and (ii) create parity between economically similar, but formally distinct, transactional forms. On the other hand, the approach would deviate from the statutory language by effectively excluding certain cash acquisitions from the CERT limitations. Moreover, it would require a degree of tracing to establish which share issuances and redemptions were undertaken by the acquiror in connection with a cash or stock acquisition. To make these correlative rules more administrable, they could apply only to issuances and redemptions that occur within six months of the CERT.

As a policy matter, one might argue that the Proposed Regulations should not draw a sharp distinction between a taxable stock acquisition and a taxable asset acquisition of all or substantially all of target's assets followed by a liquidation. In either case, aggregate corporate equity is reduced (except to the extent that the consideration is paid in the form of acquiring corporation stock). Nevertheless, the treatment of such taxable asset acquisitions as not being CERTs appears to be mandated by the Code, particularly in light of the exception for transactions in which a section 338 election is made.

With regard to tax-free asset acquisitions, there are two competing views:

Under one view, it is better to treat such acquisitions as MSAs and provide an exclusion from the scope of CERT costs for the stock consideration (similar to the ED rules) than to indirectly subject them to the CERT rules, either in whole or in part, through the back door. Economically, neither a tax-free stock acquisition nor an equivalent tax-free asset acquisition effects a reduction in corporate equity. This is particularly true if final Regulations adopt the single entity regime for stock acquisitions by a consolidated group, effectively folding a target corporation into the consolidated group upon its acquisition for CERT purposes. If a consolidated group is treated as a single entity, it does not make sense for the CERT consequences to depend on whether a target corporation joins the group in a tax-free stock acquisition or is merged into the group in a tax-free asset acquisition.

Under another view, final Regulations should accord more weight to the statutory language, in particular the distinction between stock acquisitions and tax-free asset acquisitions, excluding only the latter from the definition of a CERT. If final Regulations adopt this view, we recommend that they clarify that a tax-free asset acquisition is not an ED simply because the consideration provided is actually or constructively distributed by the target corporation to its shareholders. We do not believe that the successor rules were intended to apply in that scenario, as the transaction is an asset acquisition to which the statute does not apply.

Except in the case of asset reorganizations (which could be equated with an MSA as discussed above), we believe that the CERT rules should not be triggered by intercorporate transfers of assets in exchange for stock of the acquiring corporation because such transfers, whether to a newly-formed corporation or an existing one, do not reduce corporate equity. Although intercompany transfers within a consolidated group generally are excluded from the definition of an MSA because of the treatment of the group as a single entity, the Proposed Regulations would apply where a corporation transfers assets to a non-group member (e.g., to a controlled foreign corporation). These transfers do not reduce the transferor corporation's corporate equity and, in the case of transfers to which section 351 applies, merely represent continued ownership of corporate assets in modified corporate form. Even if the transferor receives boot in the transfer, the boot only reduces corporate equity if the transferor distributes the boot to its shareholders. We believe the ED rules adequately police these situations.

We do note that the method adopted by the Proposed Regulations for apportioning CERT attributes to departing members of a consolidated group seems ill-suited to addressing these concerns. Since the Proposed Regulations apportion CERT attributes based on relative fair market value of the departing and remaining members, a larger amount of CERT costs would be attributed to the "underleveraged" entity despite the fact that the "overleveraged" entity would likely be the entity generating most of the associated interest expense in the future. Accordingly, if final Regulations continue to treat a tax-free spin-off as a CERT transaction, we recommend that Treasury and the Service consider adopting an alternative methodology that would allocate a larger share of the CERT costs to the entity that is likely to incur most of the related interest expense (discussed further below).

Under an alternative view, the CERT rules should not apply to a corporate distribution. Final Regulations should provide that the distribution of stock of a controlled corporation under section 355 does not constitute a corporate distribution for CERT purposes. Accordingly, under this rule, section 355 distributions of stock would be outside of the ambit of the CERT rules, except to the extent of any boot distributed in connection with the spin-off. Correlative adjustments would be required to allocate CERT-related attributes (e.g., the CERT costs, interest histories, etc.) between the distributing and controlled corporations.⁹⁵ Identical rules could be applied to other divisive stock distributions, such as under section 311.

This alternative view is animated by three concerns. First, a spin-off divides a corporation into two corporations. As such, it divides corporate equity into smaller pieces, but does not reduce the aggregate amount of corporate equity (except to the extent of any boot). To the extent that new leverage is introduced, there is a corresponding increase in corporate assets. To the extent that existing leverage is apportioned between the distributing and controlled corporations, overall corporate equity remains constant. On an overall basis, the system is made whole in that interest-generated refunds from the "overleveraged" corporation generally would be balanced by increased taxable income recognized by the "underleveraged" corporation. Moreover, since the two or more businesses involved may have different economic profiles and appropriate capital structures, identifying abusive situations in which leverage is inappropriately allocated to one side or the other inherently will prove challenging and complex. Second, there is no economically equivalent transaction to a spin-off which would be respected for U.S. federal income tax purposes and which effect a corporate equity reduction. For example, a distribution of cash by the distributing corporation and use of such cash by the distributing corporation's shareholders to purchase the stock of the distributing corporation's controlled subsidiary generally would be recast as a spin-off.⁹⁶ Third, applying the CERT rules to the distribution of controlled corporation stock in a spin-off creates an impediment to corporate divisions, thereby frustrating section 355 policy.

We recommend that final Regulations exclude stock distributions under either section 305(a) or section 305(b) from the definition of corporate distributions for CERT purposes. Whether taxable or tax-free, distributions of stock do not reduce corporate equity. They simply divide corporate equity into more units.

In terms of the technical aspects of this rule, we would offer two observations. First, in adopting this rule, the government appears to be focused on a scenario where a portion of the acquisition of corporation (i.e., an MSA) is in the form of, or recast as, a redemption of target stock.¹⁰⁰ We agree that a redemption of target stock related to the acquisition of target is sufficiently connected to the MSA that the redemption should be considered as part of the MSA. However, if the acquiring corporation redeemed a portion of its own stock in connection with an MSA, we do not believe that such redemption bears a sufficient nexus to the MSA to be considered part of the MSA for CERT purposes. Accordingly, we recommend that the government clarify that only redemptions of target stock be considered part of the MSA. Any redemption of acquiror stock should be independently tested as an ED.

Second, under the Proposed Regulations, the inclusion of an integrated distribution as part of the related MSA does lead to a taxpayer-favorable result where the distribution and the MSA straddle a year end. In such case, the distribution is not treated, by itself, as a CERT even if, when tested separately, it would constitute an ED. Accordingly, the year of the distribution is not an LLY and the distribution is deemed to be a CERT cost in the later year of the MSA.¹⁰¹ On the whole, this seems an acceptable trade-off (i.e., the rule as a whole balances taxpayer-favorable and taxpayer-unfavorable components) and abuse could be policed by an anti-abuse rule if necessary (e.g., separately testing the distribution as an ED where a principal purpose for straddling a year-end is to take advantage of general rule treating the distribution as part of the later MSA). As a policy matter, we believe that there are strong arguments in favor of treating a distribution by the target corporation as part of an MSA where the distribution and MSA are part of the same plan. As an initial matter, the rule in the Proposed Regulations avoids the complexity of applying the ED and MSA rules separately to different components of an integrated transaction, which otherwise could result in the application of multiple limitation calculations. In addition, in the context of a leveraged acquisition, absent the rule, the location of the acquisition debt would determine how the series of transactions is tested. If the debt remains at the acquiring corporation, the transaction potentially would constitute an MSA. If the debt was assumed by the target corporation, the transaction potentially would constitute an ED.¹⁰² These two transactional formats create similar economic outcomes and should be analyzed in the same manner for CERT purposes to avoid creating artificial distinctions. Indeed, in other contexts, the Code has accorded similar treatment to these two transactional formats.¹⁰³

While we believe treating a distribution by the target corporation as part of an MSA where the distribution and MSA are part of the same plan is sensible in light of the underlying purposes of the statute, we note that the statute makes certain distinctions between EDs and MSAs which are difficult to justify when viewed from the prism of the overriding objective of limiting the tax advantages of debt-financed reductions of corporate equity. For instance, under the statute, the potential application of the CERT limitations to EDs may be mitigated by statutory thresholds based on the distributing corporation's distribution history or corporate value, and by the offset rules for stock issuances, whereas there are no similar forms of relief provided in the statute with respect to MSAs. Under the Proposed Regulations, through collapsing certain EDs with and into MSAs, these benefits are lost, and such an outcome may be viewed as unfair given that general tax principles will, in many cases, respect the existence of a distribution and related share acquisition as two separate transactions.¹⁰⁴

Thus, the rule in the Proposed Regulations integrating EDs into related MSA raises the correlative issue as to whether the mitigating provisions applicable to EDs ought to be extended to MSAs. For example, as a policy matter, consideration should be given to integrating all CERT activity of a taxpayer occurring within a year or, alternatively, accomplished pursuant to a single plan. That is, all EDs and MSAs could be aggregated if they occur during the same year or, alternatively, pursuant to the same plan, and the CERT mitigations would be applied to this aggregate CERT amount. In this manner, artificial and often inappropriate distinctions between MSAs and EDs would be eliminated. We acknowledge that such a rule is clearly beyond the literal language of the statute. However, the statute does provide Treasury and the Service with broad regulatory authority to exempt transactions that do not have the effect of reducing corporate equity. Therefore, it would not be unreasonable to provide equivalent rules for MSAs and EDs when they have the same overall impact on the amount of corporate equity.¹⁰⁵

Consideration should be given to whether final Regulations should adopt a different definition of debt-like preferred stock. For example, non-participating preferred stock that is mandatorily redeemable on a specified date but that carries a small voting interest does not qualify as section 1504(a)(4) stock in spite of the predominant debt-like characteristics of the instrument. Conversely, perpetual preferred stock may qualify as section 1504(a)(4) stock in spite of its equity-like features. If Congress only intended to exclude debt-like preferred stock, the current statutory definition is both under- and over-inclusive.

We therefore suggest that Treasury and the Service consider modifying the definition of preferred stock for ED purposes. To remedy one instance in which the reference to section 1504(a)(4) appears to be under-inclusive, final Regulations could provide that section 1504(a)(4) is to be applied without regard to clause (A), which excludes voting stock. A broader approach would substitute the definition of preferred stock under section 351(g), which we believe provides a better definition of debt-like preferred stock. At the same time, however, we acknowledge that Treasury and the Service may prefer not to exercise their broad regulatory authority in an area where Congress has already provided a clear statutory definition.

We are concerned that this rule will require tracing, which seems fundamentally inconsistent with the fungibility principles underlying the general rules in the Proposed Regulations governing the determination of Allocable Interest Deductions. Because neither Reg. § 1.263(a)-4 nor Reg. § 1.263(a)-5 provides a methodology for determining when debt facilitates an acquisition, such a determination is not otherwise required and likely cannot be easily made in a number of instances.

Of course, tracing of debt would be avoided if borrowing costs were disregarded for purposes of identifying CERT costs and determining the amount of CERIL. While this approach has the benefit of simplicity and recognizes that amortizable borrowing costs are not interest under the Code, it permits costs closely associated with debt-financed equity reductions to generate deductions that could be carried back and used to offset pre-CERT income.

One way to capture these costs without a tracing regime would be to exclude the costs of facilitative borrowings from CERT costs and instead treat deductions for borrowing costs on all debt, whether or not it facilitates a CERT, as interest expense for purposes of the CERT rules. Economically, borrowing costs increase the cost of incurring debt and typically are recovered over the term of the debt. In this sense, these costs are akin to interest and should be treated as such in order to achieve the Congressional objective identified in the Preamble.

Treasury and the Service have a sound basis, as a matter of statutory interpretation, for reaffirming the Avoided Cost Method. The Code itself adopts the Avoided Cost Method in absence of Regulations to the contrary, and the legislative history does not include a Congressional directive contemplating the adoption of Regulations employing another method as a general matter or even in specifically identified circumstances (outside of the context of Unforeseeable Events). Second, the parenthetical in section 172(h)(2)(B) directing the use of the Avoided Cost Method without applying the tracing rule of section 263A(f)(2)(A)(i) (which, by analogy, would have required allocation to the CERT of debt financing directly attributable thereto) suggests that Congress viewed adversely the application of tracing methodologies in this context.

As a matter of administrability, it also is preferable not to require or permit tracing of debt to specific expenditures. It is true that certain indebtedness is closely associated with certain expenditures and that the proceeds of certain borrowings can be used only to fund specified expenditures by law, regulation, or contract. Thus, project finance borrowings may be closely associated with the construction of a particular project (rather than a CERT) and, conversely, a corporation may incur acquisition funding through borrowings that are closely associated with a CERT (rather than other corporate expenditures). Nevertheless, we believe that most corporate treasury functions do not associate particular sources of funding with particular uses as a general matter, viewing the financing issue more as a question of how to fund an aggregate level of expenditures from all available sources of cash flow. Moreover, we think that a tracing regime would be difficult to administer for taxpayers and the government alike, as it would be difficult to identify and apply tracing factors across taxpayers in a consistent manner.¹⁰⁶ Accordingly, we do not recommend that Treasury and the Service adopt a general tracing methodology to replace the Avoided Cost Method.

For similar reasons, we do not recommend a change to the rule that interest expense for debt allocated under the Avoided Cost Method to CERT costs be calculated using a weighted average interest rate. From a corporate treasury perspective, the use of the average interest rate is imprecise because cash flow freed up from refraining from engaging in a CERT likely would not be used to pay down debt pro rata. Typically, a corporate treasurer would take into account various factors (e.g., contractual limitations on prepayments, prepayment penalties and make-whole provisions, etc.) in determining which debt instruments would be repaid in such case. Nevertheless, attempting to apply a tracing analysis would be fraught with administrative difficulties.

Notwithstanding our concerns about the use of tracing regimes, we also recognize that the Avoided Cost Method is somewhat harsh in its application because the allocation of corporate debt first to CERT costs in not economically realistic in many cases. As one example, the Avoided Cost Method potentially results in a limitation on NOL carrybacks where, in a single year, a corporation incurs debt to grow organically and also engages in a CERT. Under the Avoided Cost Method, the debt is allocated first to the CERT costs. As described above, under a tracing regime, the debt might be more properly allocable to non-CERT costs. Even without undertaking the administrative complexities involved in a tracing regime, it may be more economically realistic to allocate the debt to corporate expenditures for organic growth and the CERT pro rata.

In this vein, we believe that Treasury and the Service should consider whether it is possible to develop a regime for allocating debt to CERT costs that is more economically realistic than the Avoided Cost Method and that employs irrebuttable presumptions so as not to rely on tracing. As one example, sources of cash generated in the year of a CERT and subsequent LLYs could be allocated pro rata to uses of cash including CERT costs. For this purpose, sources of cash would include (a) cash on hand at the beginning of the year, (b) positive net cash flow from operations, (c) gross cash from investment activities (such as dispositions of assets), and (d) gross cash from financing activities (e.g., proceeds of issuances of debt or equity). Uses of cash would include (i) negative net cash flow from operations, (ii) gross cash expenditures for non-CERT investing activities (e.g., capital expenditures), (iii) gross cash expenditures for non-CERT financing activities (e.g., interest payments), (iv) CERT costs, and (v) cash on hand.

Such a presumptive regime would avoid any need for tracing. However, it would create its own interpretive difficulties. For example, it could involve questions as to which cash flows fall into which buckets.

In the alternative, the government could consider applying a limited tracing regime for certain types of debt that cannot realistically be considered to have funded the CERT. As one example, Regulations could exclude from the Avoided Cost Method debt of a target corporation that joins the group in an MSA (i.e., not allocate any of the target debt to the MSA). The basis for this exception would be that funds used to accomplish an MSA of a target corporation could not have been used to repay the target's own debt in absence of the MSA.

Certain events, such as a sudden and catastrophic loss or interruption of business operations due to natural or man-made disasters, seem to comfortably fit within the concept of an Unforeseeable Event. Other events, such as the 2008 financial crisis, are less clearly included in, or excluded from, the concept, but should be viewed as within the spirit of the exception because they result in unexpected operating losses.

It seems unlikely that final Regulations could provide a precise definition of an Unforeseeable Event. Nevertheless, Treasury and the Service should consider providing guidance for making the determination, or even a safe harbor. To assist you in considering this, we set out here several principles that could be included in final Regulations to provide guidance or establish a safe harbor.

The determination of whether an event is unforeseeable should be made based on an objective, rather than a subjective, standard. The question should not turn on whether the particular taxpayer did or did not expect that the event would occur, but rather on whether it would have been objectively reasonable under the circumstances for a taxpayer to have not foreseen that the event would occur. Only events that have a low likelihood of occurrence (e.g., no more than a one-in-ten chance of occurrence) as of the time of the CERT should be eligible for the exception. Moreover, an Unforeseeable Event should arise only from circumstances beyond the taxpayer's control. An Unforeseeable Event should be viewed as adverse, within the meaning of section 172(h)(2)(E), if it results in an unexpected need for expenditures or a loss in revenue; in either case, the Unforeseeable Event may cause the taxpayer to have an NOL (or to enlarge an existing NOL). In addition, an event should only be extraordinary if it requires aggregate expenditures, or results in an aggregate loss of revenue, in excess of a threshold percentage of the average indebtedness of the taxpayer outstanding in the year of the CERT (e.g., 10 percent). We also suggest that Treasury and the Service consider explicitly providing that significant adverse market events and extraordinary adverse changes in market conditions are Unforeseeable Events.

Nevertheless, the Proposed Regulations make no similar adjustment where the LLY is of normal length but one or more of the years in the interest lookback period is a short year. Moreover, if the corporation (including any predecessor thereof) was not in existence for the entirety of the interest lookback period, the Proposed Regulations deem the corporation to have been in existence for the remainder of the period and deem it to have accrued no additional interest during such remainder.

We believe that the failure to adjust the CERIL cap for a short year in the interest lookback period is problematic in that it changes an unwarranted, pro-taxpayer result to an unwarranted, pro-government result. Making an adjustment similar to the adjustment made for a short LLY would not be complex and would facilitate a more neutral and accurate comparison.

We also believe that the rule for an incomplete interest lookback period is distortive because it unrealistically deems the corporation to have a complete period with no additional interest. Although we understand the reluctance to hypothesize what interest the corporation would have accrued if it were in existence for the full length of the interest lookback period, it seems that the CERIL cap should be computed using the shorter period during which the corporation was in existence as long as the corporation existed for at least some minimum period (e.g., one full year).

The legislative history contemplates that Regulations will fashion a rule to prevent a CERT limitation from being imposed merely because of interest rate fluctuations. Factoring out interest rates fluctuations arising after the CERT, although helpful, does not fully address the conceptual issue. We believe that it should be possible to more fully factor out interest rate fluctuations without undue administrative complexity, and we believe this should be administered in a neutral manner rather than solely in favor of taxpayers as the legislative history suggests.

In our view, this can be accomplished by adjusting the CERIL cap upward or downward for interest rate fluctuations, which would be measured by reference to changes in the applicable federal rates. In effect, the CERIL cap, as determined without regard to any rule factoring out interest rates, would be multiplied by a fraction, the numerator of which would be the applicable federal rate for the LLY and the denominator of which would be the applicable federal rate for the interest lookback period. The applicable rate for the LLY would be the average of the mid-term applicable federal rates for such year. The applicable rate for the interest lookback period would be the weighted average of the mid-term applicable federal rates for such period, determined by reference to the taxpayer's average debt outstanding each year.

While imperfect, this methodology would adjust the CERIL cap for changes in market rates without adjusting for changes in issuer-specific rates, which (as noted in the Preamble) may increase because of the incurrence of leverage in connection with the CERT itself.

The single entity approach adopted by the Proposed Regulations is grounded in the strong single entity directive of section 172(h)(4)(C). It is also supported by policy and administrative considerations. As a policy matter, single entity analysis is consistent with treating a consolidated group as an economic unity, the policy consideration that has driven the adoption of single entity rules in a number of contexts.¹⁰⁷ As an administrative matter, application of single entity principles, while not simple, promotes consistent results and avoids the excessive complexity and inconsistencies that inevitably would accompany a separate company regime for members of a consolidated group.

On the whole, therefore, we support the approach taken in the Proposed Regulations to integrating the CERT rules and the consolidated return Regulations. Our comments and recommendations regarding certain aspects of the proposed rules are set forth below.

A potentially more controversial application of this principle is embodied in Prop. Reg. § 1.172(h)-5(b)(1). Applying the merger construct, the Proposed Regulations provide that if a corporation becomes a member of the group as a result of a CERT, no portion of a CNOL attributable to a CERIL may be carried back to a separate return year of the new member. This rule appears to parallel the prohibition in section 381(b)(3) on carrying back losses incurred by an acquiring corporation to a pre-reorganization year of the target corporation. Although the consolidated return rules do not generally prohibit loss carrybacks to prior separate return rules of new members,¹¹⁰ the prohibition on the carryback of a CERIL seems appropriate on balance because of the strong single entity directive of section 172(h)(4)(C).¹¹¹

This apportionment is made based on relative fair market value. While we acknowledge that such an allocation is administrable and that it parallels other allocations in the corporate division context,¹¹² we believe consideration should be given to other allocation methods in the CERT context.

We acknowledge that from a conceptual perspective, an allocation of attributes based on relative net fair market values has intuitive appeal because it reflects the fact that the group has divided into two pieces. Nevertheless, an allocation based on relative net fair market values will not necessarily match CERT histories and attributes to post-deconsolidation CERIL-generating capacities.¹¹³ A departing member without any material debt may be unlikely to generate a CERIL going forward, yet would be allocated a ratable portion of the group's CERT costs and three-year interest average. Conversely, a highly-leveraged departing member may be allocated a relatively small portion of the group's CERT costs and three year average, notwithstanding that the departing member has a significant amount of potential CERIL-generating debt.

Distribution history. The Proposed Regulations do not apply the corporate division construct across all CERT attributes. As one example, they do not apply this construct in determining the three-year distribution average for a departing member. In such case, the departing member takes with it only the history of its actual distributions to non-members; no apportionment of the group's distribution history is made.

Not requiring such an apportionment, even in a case where a Carryback Waiver Election is not made, is problematic. First, it is inconsistent with the corporate division principle. Second, it fails to take into account that the departing member actually may have funded prior distributions by the group through intercompany distributions. For example, if a wholly owned subsidiary makes a distribution to its common parent, and the parent then distributes such proceeds to its shareholders, the subsidiary's distribution to the parent will be disregarded in determining whether the subsidiary potentially engages in an ED subsequent to a future deconsolidation. However, if a second subsidiary engages in an MSA while the first subsidiary is a member of the same group, the first subsidiary is allocated a portion of the CERT costs related to the MSA when it departs the group regardless of the fact that the subsidiary did not directly participate in the MSA.

We recommend that final Regulations provide for an apportionment of the group's distribution history to a departing member. In this context, it may make sense to apportion the distribution history based on relative net fair market values on the theory that such an allocation reflects a division of remaining corporate equity and thus, perhaps, future distribution-making capacity.¹¹⁵

Intercompany transactions. The treatment accorded intercompany transactions under the Proposed Regulations also diverges from the corporate division model in an important respect. In general, the Proposed Regulations accord single entity treatment to intercompany transactions, disregarding intercompany stock acquisitions, distributions, and share issuances.¹¹⁶ However, the Proposed Regulations provide that (i) intercompany transactions between group members are not disregarded if they occur pursuant to a plan or arrangement in which a party to the transaction becomes a non-member, and (ii) in such case, transactions (such as distributions and stock acquisitions) are tested for CERT status on a separate company basis.

We believe that final Regulations should eliminate this override of single entity principles. Consistent with the single entity directive of section 172(h)(4)(C), the configuration of assets and liabilities within a group prior to deconsolidation, and correspondingly the restructuring of ownership of such assets and liabilities among members in contemplation of a deconsolidation of one or more members, should be irrelevant to the application of the CERT rules. Thus, for example, a distribution of unwanted assets from a subsidiary member to its parent member prior to a spin-off or sale of the subsidiary should not be an ED. To provide otherwise as a general matter would be inconsistent with treating the group as a single entity.¹¹⁷

Carryback waiver election. As discussed above, the default rule under the Proposed Regulations is that a departing member is apportioned a portion of the group's CERT attributes based on the relative net fair market values of the departing member and the remaining members. However, if a Carryback Waiver Election is made, the departing member relinquishes the carryback of all NOLs (and attributable portions of CNOLs) to taxable years of the former group and any preceding years and is apportioned none of the group's CERT attributes (including the interest history of the departing member). This election is procedurally similar to other elections to waive the carryback of CNOLs by departing members of a consolidated group.¹¹⁸

It should be noted that the Carryback Waiver Election differs from the existing waiver elections in at least one significant respect, which could be viewed as bearing on the question of whether the Carryback Waiver Election should be a unilateral election of the departing member (or the common parent of an acquiring group). In the context of the CERT rules, the default rule would require the selling group to provide substantially more information to the departing member (or the common parent of the consolidated group that acquires the departing member) than is required by existing waiver elections. In addition to the amount of existing CNOL and relative fair market value of the departing member, the selling group would also have to provide information regarding the group's interest history, CERT history, and CERT costs. Because the Proposed Regulations provide that the Carryback Waiver Election is made unilaterally by the departing member (or the common parent of the acquiring group), the selling group does not have the option to withhold the above-mentioned information. Unless the default rule is changed, to protect itself from this level of intrusion, the selling group will have to negotiate provisions in the sale agreement requiring the departing member (or the buyer) to make the Carryback Waiver Election.

On balance, we recommend maintaining the default rule (i.e., apportionment of CERT attributes to the departing member in the absence of a Carryback Waiver Election made unilaterally by the departing member). As discussed above, because this rule is consistent with the default rule for other carryback elections, it avoids having different default rules in similar contexts. Although somewhat burdensome, we believe that administrative difficulties will exist regardless of the default rule chosen for waivers in the CERT context and that the market will adjust to the problems discussed above.

V. Conclusion

The Proposed Regulations provide much needed guidance in a complex area of the tax law and we hope that Treasury and the Service will finalize them as soon as possible. The recommendations and alternatives described above seek to move in the direction of achieving consistent treatment of economically equivalent transactions while ensuring that the underlying policy concerns of the CERT rules are addressed. Although there was no clear consensus on just how far the final Regulations should go, we do propose clarifications and modifications to the Proposed Regulations, sometimes as consensus recommendations and other times as alternatives, for the Service and Treasury to consider, with the broad objective of providing CERT rules which are more narrowly tailored to addressing transactions that actually reduce corporate equity, and which are administrable by the government and taxpayers alike.

"Applicable Corporation" means (i) the acquiring corporation in an MSA, (ii) the target corporation in an MSA, or (iii) the distributing corporation in an ED (section 172(b)(1)(E)(iii)).

"Avoided Cost Method" means an allocation method which generally allocates indebtedness to a CERT to the extent that a corporation's indebtedness would have been reduced if amounts expended for CERTs during the applicable measurement period instead were used to repay debt (sections 172(h)(2)(B) and 263A(f)(2)(A)(ii), but without regard to section 263A(f)(2)(A)(i)).

"Carryback Waiver Election" means an irrevocable election by a departing member to relinquish the carryback of all NOLs (and attributable portions of CNOLs) to taxable years of the former group and any preceding years (Prop. Reg. § 1.1502-72(e)(1)).

"CERIL" means the corporate equity reduction interest loss, which is the excess (if any) of (i) the NOL for the LLY over (ii) the NOL for such taxable year determined without regard to any Allocable Interest Deductions otherwise taken into account in computing such loss.

"CERIL Cap" means the limitation on the CERIL for any LLY, which is the excess (if any) of (i)the amount allowable as an interest deduction during the LLY, over (ii) the three year interest average (section 172(h)(2)(C)).

"CERT Costs" means the pool of costs treated as arising from a CERT.

"CERT" means a corporate equity reduction transaction.

"CNOL" means a consolidated NOL (Reg. § 1.1502-21(e)).

"Code" means the Internal Revenue Code of 1986, as amended.

"Current Rules" means the existing statutory rules and limited administrative guidance concerning CERTs.

"departing member" means a member of a consolidated group which leaves the group.

"distribution lookback period" means the three taxable years immediately preceding the taxable year of a corporate distribution (section 172(h)(3)(C)(ii)(I)).

"ED" means an excess distribution (section 172(h)(3)(C)).

"former group" means the consolidated group from which a departing member leaves.

"interest lookback period" means the three taxable years preceding the taxable year of the CERT(section 172(h)(2)(C)(ii)).

"LLY" means a loss limitation year (i.e., the taxable year of the CERT and the next two succeeding taxable years) (section 172(b)(1)(E)(ii)).

"lookback period year" means any year within the distribution lookback period or the interest lookback period, as the case may be.

"MSA" means a major stock acquisition (section 172(h)(3)(A)(i)).

"new member" means a corporation which joins a consolidated group.

"NOL" means a net operating loss (section 172(c)).

"potential ED year" means the year in which a distribution or series of distributions constituting an ED occurs.

"Preamble" means the preamble to the Proposed Regulations (REG-140668-07).

"pre-existing CERT member" means an Applicable Corporation with regard to a CERT occurring in a separate return year which joins a consolidated group.

"Proposed Regulations" means the proposed regulations (REG-140668-07) consisting of Prop. Reg. §§ 1.172(h)-0 through 1.172(h)-5, Prop. Reg. § 1.1502-72, and certain proposed amendments to existing Reg. § 1.1502-21.

"Regulations" or "Reg." means the Treasury regulations promulgated under the Code.

"section" means a section of the Code.

"§" means a section of the Regulations.

"Service" means the Internal Revenue Service.

"three-year distribution average" means the average of the corporate distributions over the distribution lookback period (section 172(h)(3)(C)(ii)(I)).

"three-year interest average" means the average of corporate interest deductions over the interest lookback periods (section 172(h)(2)(C)(ii)).

"Treasury" means the U.S. Treasury Department.

"Unforeseeable Event" means an unforeseeable extraordinary adverse event (section172(h)(2)(E)).

² All "section" references are to the Internal Revenue Code of 1986, as amended (the "Code") and all § references are to Treasury regulations ("Regulations" or "Reg.") thereunder. A glossary of defined terms used in this Report is included as Annex A.

³ Sections 172(h)(3)(A) and (b)(1)(E)(i).

⁴ Section 172(a); Reg. § 1.172-1(a).

⁵ Section 172(b)(1)(A). The Court in United States v. Foster Lumber Co., 429 U.S. 32, at 42-43 (1976), described the rationale behind the section 172 deduction:

⁷ H.R. Rep. No. 247, 101st Cong., 1st Sess. 1250 (1989).

⁸See, e.g., TAM 9743001 (June 4, 1997) ("Congress enacted the CERT rules as part of the Revenue Reconciliation Act of 1989 to discourage the increase of corporate indebtedness generated by certain transactions such as leveraged buyouts and recapitalizations, and other transactions in which corporate equity is replaced with debt.").

⁹ S's outstanding stock value is already reflected in P's stock value and thus is not counted again.

¹⁰ Pub. L. No. 101-239 (Dec. 19, 1989).

¹¹ Section 172(b)(1)(E)(i).

¹² H.R. Rep. No. 247, 101st Cong., 1st Sess., 1251 (1989).

¹³ Section 172(h)(3)(A).

¹⁴ Section 172(h)(3)(B)(i).

¹⁵ Section 172(h)(3)(D)(ii).

¹⁶ Section 172(h)(3)(B)(ii). For acquisitions occurring prior to October 10, 1990, an MSA did not include an acquisition in which a corporation acquires stock of another corporation which, immediately before the acquisition, was a member of an affiliated group (within the meaning of section 1504(a)) other than the common parent of such group. Former section 172(m)(3)(B)(ii)(II) was repealed by H.R. 5835, 101st Cong., 2d Sess. (1990) (stating that the problematic "aspects of leveraging are present in all debt-financed CERTs, including the acquisition of a subsidiary corporation of an affiliated group. In addition, where a subsidiary corporation is acquired from an affiliated group, the sale proceeds may not necessarily remain in corporate solution. This results because the selling group may distribute the proceeds to its individual shareholders, thus reducing corporate equity").

¹⁷ Section 172(h)(3)(C).

¹⁸ Section 172(h)(3)(E)(ii).

¹⁹ Section 172(h)(3)(E)(i).

²⁰ Sections 172(b)(1)(E)(iii)(I) and (iii)(II).

²¹ Section 172(b)(1)(E)(iii)(III).

²² Section 172(h)(1).

²³ Section 172(h)(2)(A).

²⁴ Section 172(h)(2)(B).

²⁵ Section 172(h)(2)(C).

²⁶ A taxpayer is treated as having no Allocable Interest Deductions (and thus no CERIL) for any taxable year if the amount of such deductions (without regard to this rule) is less than $1 million. Section 172(h)(2)(D).

²⁷ Section 172(h)(2)(E)(i).

²⁸ Section 172(b)(1)(E)(ii).

²⁹ Section 172(h)(4)(A).

³⁰ Section 172(h)(4)(C).

³¹See ILM 200305019 (Jan. 10, 2002).

³²See ILM 200305019 (Jan. 10, 2002). See also TAM 200432014 (Aug. 6, 2004).

³³ Section 172(h)(5).

³⁴ H.R. Rep. No. 247, 101st Cong., 1st Sess., 1252 (1989). See also TAM 9743001 (June 4, 1997) ("Congress enacted the CERT rules as part of the Revenue Reconciliation Act of 1989 to discourage the increase of corporate indebtedness generated by certain transactions such as leveraged buyouts and recapitalizations, and other transactions in which corporate equity is replaced with debt" [emphasis supplied]).

³⁵ H.R. Rep. No. 247, 101st Cong., 1st Sess., 1251 (1989).

³⁶ REG-140668-07 (Sept. 13, 2012).

³⁷ Prop. Reg. § 1.172(h)-1(c)(1), (2), and (3).

³⁸ Prop. Reg. § 1.172(h)-1(d)(2).

³⁹ Id.

⁴⁰ Prop. Reg. § 1.172(h)-1(f) (also providing that the value of the fraction may not exceed 100% and that no distributions are deemed made (in excess of amounts actually distributed) in a distribution lookback period year that is shorter than the year of the potential ED).

⁴¹ Prop. Reg. § 1.172(h)-4(d).

⁴² Glenn R. Carrington, Tax Accounting in Mergers and Acquisitions ¶ 1002.1.1 (2012).

⁴³ Prop. Reg. § 1.172(h)-1(d)(1).

⁴⁴ Id.

⁴⁵See Prop. Reg. § 1.172(h)-1(b)(1); see also Prop. Reg. § 1.172(h)-5(b)(2), Example (ii).

⁴⁶ Prop. Reg. § 1.172(h)-1(b)(2).

⁴⁷ Prop. Reg. § 1.172(h)-2(a)(2).

⁴⁸ As defined in Reg. § 1.263A-9(c)(5)(ii) and (iii).

⁴⁹See Prop. Reg. § 1.172(h)-2. Note that allocable interest deductions are not netted against a taxpayer's interest income. Prop. Reg. § 1.172(h)-2(b)(5).

⁵⁰See Prop. Reg. § 1.172(h)-2(b)(2) (noting that the Avoided Cost Method "effectively allocates interest to a CERT to the extent that the taxpayer's interest costs could have been reduced if the taxpayer had not engaged in the CERT").

⁵¹ Prop. Reg. § 1.172(h)-2(b)(3)(i).

⁵²Id.

⁵³ Prop. Reg. § 1.172(h)-2(b)(3)(ii).

⁵⁴ Prop. Reg. § 1.172(h)-2(b)(3)(i) and (ii). Note the reference to section 162(k) is sensible given the potential integration of certain EDs with MSAs. The Preamble states that, because most CERTs occur under circumstances that already require application of section 263(a), invoking those rules should result in greater administrability.

⁵⁵ Under Prop. Reg. § 1.172(h)-2(b)(3), the determination of whether costs "facilitate" an MSA or ED is made without regard to Reg. §§ 1.263(a)-5(c)(1) and 1.263(a)-4(e)(1)(iv) (excluding borrowing costs). The Preamble also states that Congress objected to the carryback of NOLs resulting from leveraging that directly or indirectly enables CERTs and, therefore, it is appropriate to include borrowing costs in total CERT costs.

⁵⁶ Prop. Reg. § 1.172(h)-3(b)(1).

⁵⁷ Prop. Reg. § 1.172(h)-3(c)(1).

⁵⁸ Prop. Reg. § 1.172(h)-3(c)(2).

⁵⁹ Prop. Reg. § 1.172(h)-2(b)(6).

⁶⁰ Prop. Reg. § 1.1502-72(a)(2)(i).

⁶¹ Prop. Reg. § 1.1502-72(a)(2)(iv)(B) (also providing that, therefore, beginning on the day on which the pre-existing CERT member is included in the group, no CERIL is computed with regard to the member, independent of the CERIL computed for the group). See also Prop. Reg. § 1.1502-72(c)(3) (noting that CERT costs of an acquired member are no longer separately identified as CERT costs incurred by the acquired corporation); Prop. Reg. § 1.1502-72(d)(3)(i) (providing that, for purposes of the CERT rules, the interest from the separate return years is no longer separately traced as interest paid or accrued by the acquired corporation).

⁶² Prop. Reg. § 1.1502-72(a)(2)(i) and (iii).

⁶³ Prop. Reg. § 1.1502-72(f)(2)(i). See also REG-140668-07 ("[c]onsistent with single entity treatment under section 172(h)(4)(C), the proposed regulations provide that the distributions relevant for purposes of computing an ED of a consolidated group generally include only non-intercompany distributions").

⁶⁴ Prop. Reg. § 1.1502-72(f)(2)(ii).

⁶⁵ Prop. Reg. § 1.1502-72(f)(4)(i).

⁶⁶ Prop. Reg. § 1.1502-72(f)(3)(i). See also Prop. Reg. § 1.1502-72(f) and Prop. Reg. § 1.1502-72(a)(2)(iii).

⁶⁷ Prop. Reg. § 1.1502-72(f)(3)(i).

⁶⁸ Prop. Reg. § 1.1502-72(f)(3)(iii). See also Prop. Reg. § 1.1502-72(d)(3)(iii).

⁶⁹ Prop. Reg. § 1.1502-72(f)(2)(ii).

⁷⁰ Prop. Reg. § 1.1502-72(a)(2)(iv)(A).

⁷¹ Prop. Reg. § 1.1502-72(b)(1).

⁷²See Prop. Reg. § 1.1502-72(b).

⁷³ Prop. Reg. § 1.1502-72(d)(1).

⁷⁴ Prop. Reg. § 1.1502-72(c)(3).

⁷⁵ Prop. Reg. § 1.1502-72(c)(4)(i). Prop. Reg. § 1.1502-72(d)(3)(ii)(C).

⁷⁶ This amount includes any combination of interest history pursuant to Prop. Reg. § 1.1502-72(d)(3)(i).

⁷⁷ Prop. Reg. § 1.1502-72(c)(4)(ii).

⁷⁸ Prop. Reg. § 1.1502-72(d)(3)(ii)(C) and (D).

⁷⁹ Prop. Reg. § 1.1502-72(a)(2)(iii). Intercompany transactions are as defined in Reg. § 1.1502-13.

⁸⁰ Id.

⁸¹ Prop. Reg. § 1.1502-72(a)(2)(iii).

⁸² Id.

⁸³ Prop. Reg. § 1.1502-72(e)(1). Rules pertaining to the time and manner of making the election are provided in Prop. Reg. § 1.1502-72(e)(1).

⁸⁴ Id.

⁸⁵ Prop. Reg. § 1.1502-72(e)(1).

⁸⁶ Prop. Reg. § 1.1502-72(b)(2).

⁸⁷ Prop. Reg. § 1.1502-72(d)(3)(ii)(B).

⁸⁸ The limited reference to tax-free acquisitions may reflect the recognition that boot paid in an otherwise tax-free stock acquisition (e.g., a transaction qualifying under section 368(a)(1)(A) and (a)(2)(E)) does reduce corporate equity.

⁸⁹ These rules would have to be coordinated with the ED rules to eliminate duplication. For example, a share issuance that decreases CERT costs could not reduce the acquiror's EDs pursuant to section 172(h)(3)(E). In addition, a share redemption that increases CERT costs would not be treated as an ED.

⁹⁰See Rev. Rul. 69-6, 1969-1 C.B. 104.

⁹¹ Section 368(a)(1)(A), (a)(1)(C), (a)(1)(D), and (a)(1)(G).

⁹² Section 172(h)(3)(B)(ii).

⁹³ A section 361(c) distribution to a single corporate shareholder cannot reduce corporate equity. The same is generally true in the case of a liquidating distribution of property by a corporate subsidiary to its parent under section 332.

⁹⁴ Prop. Reg. § 1.172(h)-4(a).

⁹⁵ In the case of a divisive reorganization under section 368(a)(1)(D), these attributes could be apportioned under rules similar to the rules for a deconsolidation of a member of a consolidated group.

⁹⁶See Rev. Rul. 83-142, 1983-2 C.B. 68; Rev. Rul. 80-154, 1980-1 C.B. 68.

⁹⁷ It is unclear under current law whether many section 305(a) transactions constitute more than a "tax nothing." See Eisner v. Macomber, 252 U.S. 189 (1920). If a "tax nothing," it would seem appropriate that they not constitute a "transaction" such as an MSA or ED.

⁹⁸ Prop. Reg. § 1.172(h)-1(c)(3).

⁹⁹ Prop. Reg. § 1.172(h)-5(c).

¹⁰⁰See Prop. Reg. §§ 1.172(h)-1(d)(3), Ex. 2 (2012); 1.1502-72(b)(3), Ex. 1 (2012). See also Rev. Rul. 79-273, 1979-2 C.B. 125 (X desired to purchase the stock of unrelated P for cash, but X did not want P's subsidiary, S, so X formed Z with cash, Z merged into P, and the P shareholders received the cash and S stock treated as X directly acquiring a portion of the P stock for cash, and P's redemption under section 302(a) of its remaining stock with S stock); Rev. Rul. 78-250, 1978-1 C.B. 83 (the 65% shareholder of X eliminated the 35% minority by forming Y with all of his X stock, Y merged into X, and the majority Y shareholder's stock was converted into X stock while the minority X shareholders received cash treated as in section 302(a) redemption of their X stock); Rev. Rul. 73-427, 1973-2 C.B. 301 (X acquired 97.9% of Y's stock for cash, and X acquired the remaining 2.1% by forming S with X voting stock, S merged into Y, and Y distributed the X stock to its minority shareholders in exchange for their Y stock treated as though X transferred its stock directly to the minority shareholders under section 1001).

¹⁰¹ Prop. Reg. § 1.172(h)-2(b)(4)(iii).

¹⁰² Typically, this would be accomplished through the use of a transitory merger subsidiary which initially would incur the acquisition debt and would merge into the target corporation. To the extent of the debt financing, the transaction would be treated as a redemption by the target corporation.

¹⁰³See section 382(e)(2) (treating redemptions and corporate contractions as reducing the loss corporation's equity value for section 382 purposes).

¹⁰⁴See, e.g., Rev. Rul. 79-273, 1979-2 C.B. 125 (1979), and Rev. Rul. 78-250, 1978-1 C.B. 83.

¹⁰⁵ Section 172(h)(3)(E)(ii).

¹⁰⁶ Treasury and the Service have rejected tracing methodologies in a number of other contexts, citing administrative complexity. For example, the government ultimately determined that a presumptive regime should be adopted under Reg. § 1.1502-36, in part in light of the complexities in tracing non-economic gains and income.

¹⁰⁷See, e.g., § 1.1502-13 (generally adopting single entity principles in the treatment of intercompany transactions).

¹⁰⁸ Prop. Reg. § 1.172(h)-4(c)(2).

¹⁰⁹ Prop. Reg. § 1.1502-72(d).

¹¹⁰ In this regard, the apportionment rules of Reg. § 1.1502-21 reflect a mixture of single entity and separate company principles.

¹¹¹ From the standpoint of conceptual consistency, this view is consistent with the discussion above concerning the treatment of asset reorganizations as CERTs. Arguably, it makes sense to treat a stock acquisition and resulting fold-into the consolidated group as a transaction subject to Section 381(b)(3) for CERT purposes only if an actual asset reorganization of a target into the consolidated group is treated as CERT.

¹¹²Cf. Reg. § 1.312-10(a) (generally allocating earnings and profits in a divisive "D" reorganization based on the relative values of the assets of the distributing and controlled corporations).

¹¹³Cf. Reg. 1.1502-9 (allocating a group's consolidated overall foreign loss between departing members and remaining members based upon the members' capacity to produce foreign income which could absorb the overall foreign losses measured by the value of the assets held by each entity that are reasonably expected to produce foreign source income in the same foreign tax credit basket as the overall foreign losses).

¹¹⁴ The cap would be necessary to prevent the distortion that would otherwise result if the departing member essentially had a disproportionate share of the group's debt, but the departing member's debt significantly exceeded the aggregate CERT costs of the group.

¹¹⁵ We have an identical concern and recommendation with respect to the group's stock issuance history as it may affect future ED determinations.

¹¹⁶ Prop. Reg. § 1.1502-72(a)(2)(iii).

¹¹⁷ It may be appropriate to adopt a narrower anti-abuse rule where a target subsidiary makes a leveraged distribution in connection with its own acquisition. Such a rule may be viewed as appropriate to ensure that economically similar transactions are treated in the same manner. For example, if an acquiring corporation borrows funds and purchases an unleveraged target corporation, the CERT costs include the entire acquisition costs. If a target subsidiary borrows cash to fund an intercompany distribution in anticipation of the acquisition, an anti-abuse rule may be needed to ensure that the CERT costs include the leveraged distribution.

¹¹⁸See Reg. § 1.1502-21T(b)(3)(ii)(C)(2) and (3) (2003).