Skip to main content

Re Rev. Proc. 96-41

Arbitrage and Rebate Committee Comments on Rev. Proc. 96-41

Editor's Note: The following comment letter on Rev. Proc. 96-41 and Notice 96-49 was submitted to the IRS by the Association's Arbitrage and Rebate Committee on December 17, 1996.

Internal Revenue Service
CC:DOM:CORP:R (Notice 96-49)
Room 5226
POB 7604
Ben Franklin Station
Washington, D.C. 20044

Gentlemen and Ladies:

We are writing on behalf of the Arbitrage and Rebate Committee of the National Association of Bond Lawyers ("NABL") to comment on Rev. Proc. 96-41 (the "Rev. Proc.") and Notice 96-49 (the "Notice") and to request to speak at the public hearing to be held regarding the Rev. Proc. and the Notice.

The National Association of Bond Lawyers was incorporated in 1979 for the purpose of educating its members and others in the law relating to municipal bonds, providing a forum for exchange of ideas as to law and practice, improving the understanding and knowledge in this field of law, providing advice and comment at the federal, state and local levels with respect to existing or proposed legislation, regulations, rulings and other action affecting state and municipal obligations, and providing advice and comment in proceedings before courts and administrative bodies through briefs and memoranda as a friend of the court or agency. NABL has over 3,000 members who actively participate in some aspect of public or private municipal finance practice. They are bond counsel, underwriter's counsel, municipal attorneys, issuer's counsel, corporate counsel, trustee's counsel, defense counsel in tax and securities-related proceedings, and others involved in municipal finance. The members of NABL who contributed to the preparation of these comments are: Jeannette M. Bond, David A. Walton, Jeremy A. Spector, David A. Caprera, Arthur M. Miller, Margaret C. Henry, William H. Conner, Julianna Ebert, Amy K. Dunbar, and Mitchell H. Rapaport.

INTRODUCTION

We appreciate the opportunity to comment on the Rev. Proc. and commend the IRS for the prompt issuance of the Notice and elimination of the one-year deadline in the Rev. Proc. as originally issued. However, we continue to believe that the Rev. Proc. and Notice create a situation which results in the IRS coercing issuers into paying to the IRS the difference between the so-called "spot price" for their escrow securities and the price actually paid. By "offering" issuers a declared maximum penalty of the difference between "spot price" and the price paid, the IRS' amnesty forces issuers to evaluate their transactions based upon the conclusions in the Rev. Proc. regarding pricing and the definition of fair market value. Issuers who do not take advantage of the Rev. Proc. and pay the penalty run the risk that an IRS agent will audit their transactions, resulting in a greater penalty or even the loss of tax-exempt status. This is a real risk, particularly since we understand that IRS National Office personnel recently provided training to agents regarding overcharging for escrow securities. Administrative convenience does not justify this coercive approach.

As discussed below, issuers are in an impossible situation as a result of the Rev. Proc. With at most a few exceptions, issuers have no idea whether they overpaid for their open market escrows and no practical way to determine whether or not they did. Bond counsel are in no position to evaluate the matter, and investment bankers generally shy away from second-guessing the pricings of others on a retrospective basis, given the subjective nature of such pricing considerations.

Given the huge volume of advance refunding transactions during the period from 1991 through 1994, when open market securities were commonly used in escrows, the Rev. Proc. will affect a tremendous number of issuers. The claim that the Rev. Proc. would only "affect" about 100 issuers is far from accurate since every issuer who used open markets must evaluate its situation under the Rev. Proc. As a result of the Rev. Proc., issuers who did not know or have reason to know of potential overpayments are being (or will be) forced to spend time and money attempting to self-audit their transactions. All of this seems unnecessary and unfair. The Rev. Proc. should either be revoked or revised so that it only includes the IRS' offer to resolve concerns about overcharging for investments through the closing agreement procedure for issuers that voluntarily come forward. Further, either the Rev. Proc. or another appropriate document should instruct IRS agents not to assert penalties against issuers (or the holders of their bonds) who did not know or have reason to know of overcharging for investments, provided that those issuers cooperate with any IRS or SEC investigations of the pricing of open market securities. Alternatively, with the issuer's consent, cases should be referred to the SEC, an agency that may be more able to police abuses in pricing matters and which has the power to obtain disgorgement of overcharges from securities providers. Where, however, an issuer knew or had reason to know it was being overcharged for its investments, the IRS should be free to pursue a recovery from the issuer.

GENERAL COMMENTS

(1) Retroactivity. The Rev. Proc. is an attempt by the IRS to impose rules retroactively against issuers where no such rules previously existed. Under the Rev. Proc., unless an issuer purchasing open market securities used the bidding process described in the Rev. Proc. or managed to pay no more than "spot price" (an extremely unlikely possibility) the Rev. Proc. unmistakably asserts that the issuer overpaid for those securities and may have caused its bonds to be arbitrage bonds. We believe that this retroactive approach is unwarranted and runs counter to the spirit of cooperation that we believe should exist between state and local governments and the federal government. Moreover, by imposing these rules retroactively, the IRS has effectively turned much of the proposed regulations relating to the purchase of open market securities into regulations that are retroactively effective.

(2) Disregard for past comments and industry practice. The Rev. Proc. and the Notice have the effect of retroactively prohibiting industry practices that have existed and been relied on by issuers for many years. These industry practices developed to deal with a SLGS program that, until recently, was often not a viable investment alternative. Until relatively recently, it was common for issuers to purchase open market securities on a negotiated basis and rely on certifications made by the providers of those securities. The IRS has every right to look at that practice and conclude that more restrictive prospective rules are needed. It is unwarranted, however, to use hindsight to outlaw retroactively the reliance on practices the IRS and Treasury were long aware of and never objected to until now. Not only were the IRS and Treasury aware of these practices, they were implicitly approved by the IRS in 1992 and 1993. In the proposed arbitrage regulations issued in 1992, the IRS requested comments on whether specific rules were needed for the purchase of open market securities. The IRS received a number of comments to the effect that such rules were not needed. Some of these comments, including those of NABL and the American Bar Association, described industry practices in detail. After having received and considered such comments, the final regulations contained no new rules or safe harbors for purchases of open market securities.1 Not surprisingly, this was widely viewed as an IRS approval of these industry practices. In light of this, it is unwarranted and unfair for the IRS three years later to in effect retroactively impose bidding rules on the entire industry. If the IRS believes it made a mistake in not imposing bidding rules in 1993, it must accept some responsibility for the situation, not try to shift it to innocent issuers. What it should not now do is impose a set of rules that have the effect of forcing every issuer that did not foresee that such rules would apply to pay--in fact, overpay--significant amounts to the IRS or risk loss of tax-exemption.

(3) Strong-arm approach. The Rev. Proc. and the Notice are a clear attempt by the IRS to coerce issuers into paying the so-called "differential" amount. This approach unfairly attacks those who did not benefit from the overpricing. Although the Notice removed the one-year deadline on the availability of the IRS' settlement offer, issuers are still faced with a choice: pay the differential amount or run the risk that an IRS examining agent will audit its transaction. If a transaction is audited, the IRS agent will come armed with the legal assumptions in the Rev. Proc. and assert, based on the statements in the "background" section, that the issuer paid too much (e.g., because it paid for remote contingencies, etc.), relied on nonconclusive certifications, and, therefore violated the arbitrage rules. In this situation, the Rev. Proc. states that the result upon audit "could be different from the terms of" the Rev. Proc. The clear intent is that those who do not settle up on their own may have their bonds taxed or, if the agent is generous, have to pay "lost taxes," an amount likely to be far greater than the differential. The Rev. Proc. applies this strong-arm approach to every issuer in the country. We believe this is the wrong way for the IRS to deal with municipalities with which it believes it has, as repeatedly stated by IRS staff, a "special relationship."

A fairer approach would permit an issuer to provide the relevant facts to the IRS so that the IRS or the SEC could pursue the wrongdoers if it wished to do so. By this method, issuers can avoid the extensive administrative and litigation costs that will be involved in any attempt to recover damages from the investment providers.

(4) Yield restriction and reasonable expectations. Under Section 148(c) of the Code and '1.148-2 of the regulations, a bond is an arbitrage bond if the proceeds are reasonably expected to be used to acquire higher yielding investments (or to replace funds so used) or if the issuer intentionally uses any proceeds in a prohibited manner. Thus, compliance with the arbitrage yield restriction rules requires that the issuer have the proper reasonable expectations on the issue date and that it thereafter did not take any intentional act to violate the arbitrage rules.

In the context of the Rev. Proc., did an issuer that purchased open market securities at a price that is subsequently determined to have exceeded fair market value lack the proper reasonable expectation or commit an intentional act to earn arbitrage? The regulations contain the following definition of "reasonable expectation:"

An issuer's expectations or actions are reasonable only if a prudent person in the same circumstances as the issuer would have those same expectations or take those same actions, based on all the objective facts and circumstances. Factors relevant to a determination of reasonableness include the issuer's history of conduct concerning stated expectations made in connection with the issuance of obligations, the level of inquiry by the issuer into factual matters, and the existence of covenants, enforceable by bondholders, that require implementation of specific expectations. For a conduit financing issue, factors relevant to a determination of reasonableness include the reasonable expectations of the conduit borrower, but only if, under the circumstances, it is reasonable and prudent for the issuer to rely on those expectations.

'1.148-1. Thus, in order to satisfy the reasonable expectations requirement, an issuer's actions in purchasing open market securities would have to be consistent with those of a "prudent person in the same circumstances." Issuers satisfied this requirement in several ways, including the issuer's own review and diligence, review by a financial advisor, review by another broker-dealer, and certifications provided by the seller of the securities. We understand that only the reliance on the certification of the seller is being questioned by the Rev. Proc. We believe that reliance on the certification of the seller of the securities satisfies the reasonable expectations standard. First, reliance on such certifications was a common practice in the industry for many years and, therefore, similarly situated persons were "taking those same actions."

Second, issuers were generally advised by their bond counsel that reliance on such certifications was appropriate. Third, the legislative history to the amendments to section 6700 states that reliance on certifications of experts is appropriate in connection with tax-exempt financings:

bond counsel, issuer's counsel, and underwriters' counsel would be entitled to rely upon a feasibility study conducted by an engineering firm reputed to be an expert in the subject matter and area of study, unless such counsel independently knew or had reason to know information bringing into question the results of that study. Absent that, counsel would not be required to question the assumptions underlying, or results reached by, the study. Similarly, bond counsel would be able to rely, as to matters of fact or expectation relevant to his or her opinion, on information provided by other parties (including the issuer) absent actual knowledge or a reason to know of its inaccuracy or the use of statements not credible or reasonable on their face.

Omnibus Budget Reconciliation Act of 1989, House Report, pp. 1398. These statements strongly support the notion that issuers acted reasonably in relying on statements of investment providers where they did not know or have reason to know of information that would cause them to question those certifications and that issuers need not have questioned the assumptions or conclusions of those certifications.

Fourth, the IRS specifically reviewed the long-standing practice of relying on provider certifications in connection with the finalization of the arbitrage regulations and provided no indication whatsoever that this was not an acceptable practice. Finally, the certification itself was provided by a person with whom the issuer probably had a previous relationship and whose statements the issuer had no reason to believe might be false.2 If such a person overcharged for the securities or lied in the certification, it would have violated various securities, tax, and criminal laws, and violated NASD procedures. Given all of these factors, we believe a prudent person would reasonably have relied on such a certification. We also note that Black's Law Dictionary defines "prudent" as judicious, careful, circumspect, sensible; synonymous with cautious. Similarly, "prudence" is commonly associated with care and diligence and is contrasted with negligence. We believe that reliance by issuers on certifications satisfied this standard and that, in doing so, issuers complied with the requirement that they reasonably expected that they would make only fair market value purchases.

Similarly, we do not see how an issuer that did not know or have reason to know that it overpaid for securities has taken an "intentional act" to violate the arbitrage rules. With respect to the definition of intentional acts, the regulations contain the following statement:

The taking of any deliberate, intentional action by the issuer or person acting on its behalf after the issue date in order to earn arbitrage causes the bonds of the issue to be arbitrage bonds if that action, had it been expected on the issue date, would have caused the bonds to be arbitrage bonds. An intent to violate the requirements of section 148 is not necessary for an action to be intentional.

'1.148-2(c). To have intentionally overpaid for open market securities, an issuer must have known that it was doing so.

Based on this analysis, we believe that an issuer who did not know or have reason to know that it was overpaying for open market securities and who relied on a certification from the provider of the securities did not violate the arbitrage yield restriction rules as a result of the purchase of those securities.

(5) Arbitrage rebate consequences. Given that there has been no violation of the yield restriction rules by an innocent issuer, the only other potential arbitrage violation would be with respect to the arbitrage rebate rules. Under those rules, rebate must be computed based on the fair market value of the investments purchased and an issuer who overpaid for its investments may owe rebate based on the difference between fair market value and the price actually paid.

Can the bonds of an issuer who unknowingly overpaid for open market securities but who relied on a provider's certification be declared taxable as a result of the failure to pay rebate in a timely manner? Section 1.148-3(h) of the regulations provides as follows:

(1) In general. The failure to pay the correct rebate amount when required will cause the bonds of the issue to be arbitrage bonds, unless the Commissioner determines that the failure was not caused by willful neglect and the issuer promptly pays a penalty to the United States. If no bond of the issue is a private activity bond (other than a qualified 501(c)(3) bond), the penalty equals 50 percent of the rebate amount not paid when required to be paid, plus interest on that amount. Otherwise, the penalty equals 100 percent of the rebate amount not paid when required to be paid, plus interest on that amount.

(2) Interest on underpayments. Interest accrues at the underpayment rate under section 6621, beginning on the date the correct rebate amount is due and ending on the date 10 days before it is paid.

(3) Waivers of the penalty. The penalty is automatically waived if the rebate amount that the issuer failed to pay plus interest is paid within 180 days after discovery of the failure, unless, the Commissioner determines that the failure was due to willful neglect, or the issue is under examination by the Commissioner at any time during the period beginning on the date the failure first occurred and ending on the date 90 days after the receipt of the rebate amount. For purposes of section 1.148-3(h)(3), willful neglect does not include a failure that is attributable solely to the permissible retroactive selection of a short first bond year if the rebate amount that the issuer failed to pay is paid within 60 days of the selection of that bond year. Generally, extensions of this 180-day period and waivers of the penalty in other cases will be granted by the Commissioner only in unusual circumstances.

We believe that issuers that acted "prudently" could not be guilty of willful neglect and, therefore, cannot have taxable arbitrage bonds under the rebate rules. Further, the worst case scenario for these issues is that they owe the unpaid rebate plus interest plus a 50 percent penalty. Under the regulations, this penalty is automatically waived if the rebate amount plus interest is paid within 180 days "after the discovery of the failure." We do not believe that issuers should be considered to have discovered a failure merely because the IRS asserts questionable pricing theories in the "background" in a Rev. Proc. This would force every issuer to self-audit each of its transactions. Further, by the time the IRS decides what to do with the Rev. Proc., the 180 days will likely have run.

Given all this, it is in fact questionable whether the Rev. Proc. (even ignoring all its technical problems) provides anything that is not already available to issuers under the regulations, since the rebate provisions permit issuers to correct their errors.

(6) Ignoring available penalties. By issuing the Rev. Proc. and the Notice and holding issuers accountable under what is in effect a strict liability standard for failure to purchase at fair market value, the IRS seems to have deliberately decided not to impose the penalties under section 6700 that Congress provided to the IRS to pursue the real wrongdoers in municipal finance abuses. Although we disagree entirely with the overall approach of the Rev. Proc., the IRS could just as easily have directed the Rev. Proc. at providers of open market securities who signed fair market value certifications that were false or incorrect. Those providers could have been informed that, unless they paid the differential, they would run the risk of having to pay a $1,000 per bond penalty. Congress intentionally revised section 6700 in 1989 so that the IRS could impose significant penalties on the wrongdoers in municipal finance transactions. To ignore or reject this available penalty and focus on issuers is the wrong approach. The IRS seems to have chosen to go after the easiest party to attack: issuers of outstanding bonds with limited resources and a limited ability to fight back.

In 1989, Congress conducted a general re-examination of the Internal Revenue Code's penalty provisions. A group of administration and Congressional staff, including staff of the Ways and Means Committee, then-Congressman Beryl Anthony, and the IRS, together with representatives of NABL, successfully advocated the modification and expansion of section 6700 so that enforcement efforts could be properly targeted at the "deserving" participants in a tax-exempt bond transaction (rather than the bondholders or issuer). We understand that there may be a concern at the IRS that section 6700 may only provide a penalty of $1,000 per bond issue. We believe that such a reading of section 6700 in light of the legislative history is nonsensical. To read the amendments in this manner leads to the conclusion that Congress acted in a wholly impractical manner. It makes no sense, and certainly does not deter abuse, to provide a $1,000 penalty in transactions that average in the millions of dollars. Congress certainly is not so naive as to enact a new penalty that would have absolutely no deterrent effect and that would cost more to enforce than it would produce. Statutes are not to be interpreted in such an illogical manner.3 Furthermore, the IRS specifically sought to and did clarify confusion in the Circuits regarding the application of section 6700 as it had been applied to tax shelters to make clear that the penalty was not to be applied on a per transaction basis but on a per share, per plan, or per denomination basis.

More specifically, section 6700 focuses on a "plan or arrangement," which term was modified in 1989 to include tax-exempt bonds. See Omnibus Budget Reconciliation Act of 1989, House Report, pp. 1397-1399. The organization of a plan or arrangement or the sale of each interest in a plan or arrangement is a separate activity for purposes of applying the section 6700 penalty; the penalty for each sale is $1,000. See Omnibus Budget Reconciliation Act of 1989, Conf. Report, pp. 657-658. In the context of municipal bonds, the issuance or sale of each bond is a "plan or arrangement." Neither the statute nor the legislative history used the term "issue of bonds," clearly evidencing that it is not the "issue" that is the plan or arrangement; it is each bond. Further, even if plan or arrangement refers to the issue, the term "interest" must refer to each bond. The clear purpose of the changes to section 6700 was to treat bonds like tax shelters. To interpret "plan or arrangement" or "interest" to mean the entire bond issue would clearly subvert this intent.

(7) The IRS should identify and clearly state its goal. We suspect that no matter how pervasive the IRS believes investment pricing abuses may have been, it does not believe that issuers knowingly participated in (or even should have known they were participating in) many of these abuses.4 As the Government Finance Officers Association stated in its comments on the proposed regulations, "yield burning" was predominantly a sales abuse visited upon unknowing and innocent state and local officials. If the IRS has a different view of this matter, it should say so publicly since that will make its policies and positions more understandable.

If we are correct that the IRS believes that investment pricing abuse was in most cases something perpetrated on both the IRS and governmental issuers, we believe that the IRS should take a fresh look at the Rev. Proc. The first step in such a fresh look is for the IRS to identify what it is trying to achieve. The Rev. Proc. seems to have as its goal the recovery from state and local governments of something approximating the difference between what issuers paid for their open market securities and what they would have paid to purchase the same securities for next day delivery (with no contingencies). We question the purpose of this approach. It cannot be to punish the wrongdoer, since most issuers neither knew of nor should have known that they were paying too much (if that is, in fact, the case). Further, such issuers did not benefit from the "wrongdoing," since they are out of pocket the amount of any overpayment. Issuers cannot easily recover these overpayments from their securities providers; it is likely that they will have to sue and prove in court that the providers sold the securities at more than fair market value. If all this is correct, most issuers settling up under the Rev. Proc., despite having neither known of their "wrongdoing" nor benefitted from it, will have to pay municipal funds to the IRS. The only goal this approach fosters is to hold state and local governments absolutely accountable for any improprieties involving their tax-exempt bonds regardless of any knowledge or benefit. We know of no other areas in the tax law that the IRS enforces in this manner or that may effectively force a taxpayer to turn itself in and pay a penalty for a wrong committed by another.

We believe that there is no real benefit to the above-described approach, other than as a revenue raising function. In the tax-exempt bond area a little enforcement goes a long way. Further, in our view, the targeting of issuers who knew (or should have known) of investment pricing abuse and of the securities providers who benefitted will be effective to limit the scope of this and other types of abuse in the future. In short, if the intent of the drafters of the Rev. Proc. was to write a legalistically perfect cure to overcharging for investments, without regard to the fact that it may force them to settle by overpaying to the IRS any possible overpayment for its investments, the Rev. Proc. has succeeded. On the other hand, if the intent was to eliminate abuses, the IRS needs to re-think its approach. When the IRS finalized the arbitrage regulations, it indicated that it would concentrate its efforts on effective enforcement, not on industry-wide rulemakings. Three years later, the IRS should recognize that investment pricing abuses by issuers are more the exception then the rule. The IRS should enforce and regulate accordingly, not by issuing a Rev. Proc. that essentially declares on a retroactive basis that all purchases of open market securities other than on a bidding basis involved so-called "yield burning." It is appropriate for the IRS to condition any non-enforcement against issuers who neither knew nor should have known of investment overcharging on their cooperation in the IRS' enforcement efforts.

(8) Need to distinguish wrongdoing. There may well be situations where issuers overpaid for their open market securities and knew (or should have known) that they were doing so. We believe that such issuers do not represent more than a relatively small percentage of the market. The far greater percentage is made up of issuers who purchased securities at a price that reflected a mark-up for certain contingencies and in reliance on certifications from the investment providers to determine fair market value. These issuers did not know of any overpricing, did not benefit from such overpricing, yet will bear a significant financial cost if punished for allegedly overpaying for securities. We believe that there is no need for the IRS to punish innocent issuers merely because the IRS can easily do so, particularly when there are other parties to whom the IRS can appropriately direct its enforcement activities.

(9) Failure to suspend the Rev. Proc. While the Notice's elimination of the one-year deadline for entering into settlement agreements was helpful, we strongly believe that the Rev. Proc. should be suspended and that IRS agents should be instructed not to raise fair market value issues or rely on the Rev. Proc. until the policy and legal questions raised by the Rev. Proc. are finalized. Despite the elimination of the one-year deadline, issuers remain at serious risk that agents will examine their transactions and use the Rev. Proc. to extract penalties significantly in excess of those provided under the Rev. Proc. In addition, any inference intended by the Notice that the IRS would hold enforcement actions in abeyance was eliminated by the IRS' recent training of agents regarding overpayments for escrow securities. As long as this threat exists, issuers are put in a dangerous position. Further, as described above, the continued applicability of the Rev. Proc. threatens to serve as notice to issuers of a rebate violation for purposes of determining whether a rebate penalty is owed. NABL continues to favor the withdrawal of the Rev. Proc. Alternatively, the Rev. Proc. could be revised and reissued without the substantive provisions contained in the background section.

(10) The Rev. Proc. eliminates the value of safe harbors. One of the many disturbing aspects of the Rev. Proc. is what it means for issuers who do not follow IRS safe harbors. Even if issuers had perfect hindsight and knew that the bidding safe harbors for guaranteed investment contracts applied equally by analogy to open market securities, still those provisions are only safe harbors. For example, an issuer who solicited three or more bids but only received two from disinterested parties might determine that, despite a technical failure to comply with the safe harbor, its procedure had been sufficient to establish a market price purchase. The Rev. Proc., however, effectively turns the safe harbors into explicit rules on a retroactive basis. Any issuer who deviates from the safe harbors identified by the Rev. Proc. is considered to have overpaid and must cure this noncompliance by paying the differential to the IRS or risk loss of tax-exemption. If this is the IRS' view of how safe harbors work, we suggest that safe harbors are a trap and unnecessary (and "rebuttable presumptions" even more so).

(11) Pricing practices and roles of other regulators. We cannot comment on those portions of the Rev. Proc. in which the IRS questions the propriety of mark-ups for contingencies and delayed closings since, as bond counsel, we do not profess to possess the expertise to judge the correctness of these points. However, we believe that the IRS will find relatively consistent mark-ups whether there was significant arbitrage or not. What is necessary before imposing rules such as those contained in the Rev. Proc. is that the IRS and the securities providers gather information on the pricing of open market securities under these various scenarios and evaluate the real forces involved in pricing decisions. In this context, it seems totally unfair that the SEC and the NASD could investigate the pricing of open market securities and determine that there is no liability (that is, no overcharging for investments) yet issuers would have been compelled by the Rev. Proc. to make payments to the IRS.

Regardless of whether the IRS views a customary market risk as remote and, therefore, not worthy of compensation, if such compensation has been standard in competitively bid transactions and transactions in which issuers incurred significant negative arbitrage, that is a strong indication that the IRS is wrong in its premises. Some have suggested that, despite these facts, issuers still paid too much (that is, even where issuers were in a significant negative arbitrage situation, they overpaid). It is unwarranted for the IRS to retroactively impose its view as to how much an issuer (or any person) should pay for a contingent, delayed delivery of securities. The regulations and the Rev. Proc. define fair market value as what a willing buyer would pay a willing seller in a bona fide arm's length transaction. The Rev. Proc. questions valuations where one party to the purchase lacked a financial incentive to obtain the best price. Where there is significant negative arbitrage, the issuer has every incentive to obtain the best price and is a willing buyer in an arm's length transaction. That the IRS thinks the issuer paid too much in such a transaction should be irrelevant since the transaction is, by definition, at "fair market value." If issuers paid the same prices (and paid for the same risks) in situations where there is no negative arbitrage, those purchase prices equally satisfy the definition of fair market value.

It is also clear to us that the pricing of open market securities is a complex, fact-specific process. The strongest evidence of this is that the National Association of Securities Dealers ("NASD"), which requires that sales occur at a "fair price," has a complex series of factors to take into account in determining whether pricing is fair. Given this complexity, it is unlikely that the IRS can appropriately conclude that issuers who paid more than the so-called spot price paid too much.

(12) The purported lack of a financial incentive. Both in the Rev. Proc. and in its subsequent defense of the Rev. Proc., the IRS has made much of the purported lack of a financial incentive to invest escrows at market rates. We believe that this point has been greatly exaggerated and that state and local officials responsible for these investments and securities providers certifying to the pricing of the securities have significant incentives in this regard. First, as the Rev. Proc. amply demonstrates, issuers who overpay for open market securities risk having their bonds declared taxable or paying significant penalties. Second, the individuals who sign the issuer's arbitrage certificates and the securities provider's fair market value certificates risk the section 6700 and other, more serious, penalties for tax and securities fraud. Finally, state and local officials who pay too much for open market securities risk violating local laws that prohibit paying more than fair market value for anything as they carry out their fiduciary responsibilities. Contrary to the IRS' apparent belief, these are significant disincentives to intentionally or recklessly "burn" yield.

(13) Inability to challenge IRS determinations. Under current law, when the IRS challenges the tax-exempt status of an outstanding issue of bonds, as a practical matter, the issuer has few options available if it disputes the IRS' conclusions. While the issuer can argue with the agent's findings and request technical advice from the National Office, once these options have been exhausted, the issuer has little practical choice but to accede to the IRS' demands. Further, throughout any such controversy, the issuer must do everything possible to make sure that the IRS continues to be amenable to a closing agreement. Unlike other "taxpayers," except in the rarest of circumstances, an issuer of outstanding tax-exempt bonds has little ability to litigate the tax issues that the IRS has raised. The reason for this is that, even assuming that the issuer, rather than a holder, can litigate on its own, the issuer will have to disclose to the market that its bonds are being audited, that the issuer has chosen to take the IRS to court over the issue, and that an adverse court decision could result in a loss of tax-exempt status. This situation would likely result in an immediate reduction in the price of the issuer's bonds, lawsuits against the issuer, and potential difficulties in issuing bonds in the future. In the case of variable rate bonds, this would result in an immediate effect on interest rates.

In the context of the Rev. Proc., an issuer that is audited by the IRS is already out of realistic options; the Rev. Proc. sets forth the National Office position and, therefore, the issuer's only choices are to pay up or litigate. Because of these circumstances, we believe that the IRS needs to be very careful in implementing its tax-exempt bond audit program. For the most part, in public discussions of the audit program, the IRS has generally indicated that it is proceeding cautiously in this area. The Rev. Proc. and the Notice, however, are wholly inconsistent with a careful, considered approach. For this reason, we believe the Rev. Proc. and the approach of the Notice should be reconsidered.

(14) Use of Rev. Proc. to make law. We also believe that it is inappropriate for the IRS to use the background section of a Rev. Proc. to attempt to make law, particularly in a retroactive manner. Outside of the Rev. Proc. and the Notice, the IRS has been justifiably cautious in issuing retroactive guidance in the tax-exempt bond area. This makes it all the more inappropriate for the IRS to use a section labeled as "background" to state its views regarding the law. In fact, were it not for the background section, it is likely that the municipal bond industry would have viewed the Rev. Proc. as a positive development -- an attempt to provide relief to issuers who knew or suspected that they overpaid for their open market securities. Moreover, much of the background section of the Rev. Proc. contains inappropriate statements of factual rather than legal conclusions, such as those relating to the proper pricing of contingent, forward delivery securities.

For example, Sections 2.15 and 2.16 of the Rev. Proc. claim that the simultaneous purchase of open market securities and SLGS constitutes the receipt of a call option, the value of which must be included in any settlement. Leaving aside the accuracy of the typical factual situation as stated in the Rev. Proc., this is a clear-cut example of substantive law and rules being announced and made effective retroactively through a revenue procedure. There are no case law, regulations, or rulings cited in support of the IRS' position. The conclusions reached are wholly inappropriate in the context of a revenue procedure. To the extent that the IRS believes in the correctness of its approach, it should promulgate it in the form of a revenue ruling, rather than through a closing procedure, particularly one with the coercive elements of the Rev. Proc.

(15) Failure to obtain industry input. We believe that many of the problems with the Rev. Proc. could have been avoided if the IRS had attempted to obtain industry input before releasing a document in "final form" and not subject to comment (at least until the industry uproar occurred) and which is in effect to this day. There is only so much that can be learned about the proper pricing of open market securities without getting input from the securities industry. Given that so much of the Rev. Proc. is based on the proper pricing of open market securities, industry input should have been a vital part of the process.

(16) Future Rev. Procs. Although we have many concerns with the Rev. Proc. as it applies to open market escrows, we are also concerned that the IRS not view the Rev. Proc. as an available format for future enforcement actions. In other words, many of the comments included herein apply equally to any use of the format used in the Rev. Proc. to "suggest" noncompliance and offer a purportedly reasonable settlement offer. Prior to the Rev. Proc., the IRS has only used this approach in rare, unusual, and extremely limited circumstances. This should continue to be the case.

In the rare situations where, in the past, the IRS has used the type of approach taken in the Rev. Proc., it has always stated a clear standard that did not leave issuers wondering where they stood. For example, both in Rev. Ruls. 80-328 and 82-101, the IRS issued revenue rulings clearly describing the facts, the law as the IRS saw it, and its legal conclusion. In those situations, issuers clearly knew if they had a problem and knew the potential consequences of failing to make use of the IRS' closing agreement offer.

The Rev. Proc. on the other hand, provides little certainty. There is no statement in the Rev. Proc. that tells issuers that they do or do not have a problem. For example, the Rev. Proc. states that the risk of non-settlement and provider certifications are "factors" to be considered in determining whether an issuer paid in excess of fair market value. Issuers must ask their counsel whether they have a problem and counsel must respond that there is no problem if a market price was paid. An issuer who, based on this advice, does not choose to settle under the procedure in the Rev. Proc. will be uncertain as to what the IRS' position will be if the issuer's bonds are challenged. It is possible to interpret the Rev. Proc. as an attempt to force issuers into settling out of IRS concern that it could not make its views hold up in court and that, in any event, would tax the IRS' resources if it pursued a full legal challenge. We believe that if the Rev. Proc. is revised, in fairness it should provide certainty to issuers rather than the in terrorem rules it currently provides.

(17) Equal treatment of issuers. The Rev. Proc. indicates that the penalty under the Rev. Proc. (the differential) is not necessarily available to issuers who do not voluntarily enter into a settlement agreement. If this penalty is the right amount, we fail to see why it should not apply equally to all issuers, including those who do not self-audit and voluntarily settle with the IRS. We believe that all "taxpayers" should be treated equally. We question this approach, both in the context of the Rev. Proc. and future enforcement actions that might take a similar approach. The IRS should not be establishing post-closure self-audit rules and penalties especially through revenue procedures.

(18) Ability to determine the differential. The Rev. Proc. is also lacking in the practicality of its "remedy." Except for the most sophisticated issuers, we suspect that issuers will not have access to the information that would enable them to determine the differential, assuming that they understood what was required. The intra-day changes in the market for both spot and forward pricing are such that we believe issuers will have great difficulty in making a practical application of the Rev. Proc. As we understand it, experts can review the same escrows on the same day and conclude that different prices are nonetheless each "fair market value." So some adjustments must be made to clarify and simplify the determination of the "remedy."

(19) Spot price as fair market value. Despite the clear statements to the contrary by some at Treasury and the IRS, we continue to be concerned that in applying the Rev. Proc., IRS auditors will equate spot price with fair market value (or to state it another way, that issuers, given the Rev. Proc., will not be able to "prove" fair market value and, therefore, by default, spot price is the closest thing). We have similar concerns about the presumption that a delay of more than 30 days between pricing and settlement is "bad." We know of at least two major metropolitan areas that, as a result of local law, must have pricing and settlement separated by more than 30 days.

SPECIFIC COMMENTS

(1) Two certificate procedure. As stated throughout these comments, we advocate the withdrawal of the Rev. Proc. and significant changes in any new versions of the Rev. Proc. Until recently, some issuers used a two certificate process in purchasing open market securities. Under this procedure, both the provider of the securities and another expert (e.g., a financial advisor or a primary dealer) certified that the purchase was made at a price not in excess of fair market value. Even if the IRS makes no other changes in the Rev. Proc., this procedure should be effective in establishing fair market value and the Rev. Proc. should be revised to clarify that, like a bidding process, for purposes of applying the fair market value rules retroactively, issuers who relied on this two certificate procedure are presumed to have satisfied the tax laws. This clarification is needed by issuers both in trying to decide whether to avail themselves of the offer contained in the Rev. Proc. and to ensure that zealous auditors will not use the statements in the Rev. Proc. regarding certifications to challenge purchases using the two certificate procedure.

(2) Negative arbitrage. Similar to the two certificate procedure, many issuers and their counsel relied on the fact that the yield on the securities that they purchased was well below the bond yield to determine that a market price was paid for the securities (since even the IRS ought to concede that an issuer would not willingly give away its own earnings to a securities provider). The interim final arbitrage regulations issued in 1992 adopted this approach for investment contracts. The Rev. Proc. should be revised to clarify that a material amount of negative arbitrage (that is, at least 12.5 basis points) establishes a fair market value purchase.

(3) Form 10001. Form 10001, which is required to be filed to be eligible for a closing agreement, requires issuers to submit a significant amount of documentation. Much of this documentation seems designed to enable the IRS to determine whether the issuer's bonds might have violated any other rules rather than to provide the information necessary to pay the differential. Given that the form must be signed under penalty of perjury, copies of the Form 8038 or 8038-G, confirmations, and a statement explaining the determination of the spot price and differential should be sufficient. Copies of the offering document, escrow agreement, and arbitrage certificate are unnecessary and will only make issuers more suspicious of the IRS' motivations in issuing the Rev. Proc. and the benefit of a closing agreement offer.

REQUEST FOR COMMENTS

In the Notice, the IRS requested specific comments on several questions. We wish to respond to certain of these questions:

 (i) Appropriate time period for closing agreement requests. If the Rev. Proc. retains a legal background section that suggests that a vast majority of open market securities were overpriced, there is no significant reason to provide a deadline on availability of the settlement offer, since issuers who do not take advantage of the settlement offer will always be at risk. If, on the other hand, the Rev. Proc. is revised to more effectively identify the limited class of transactions in which issuers knew or should have known that improper conduct was occurring, then a deadline is appropriate and a one-year period could be satisfactory.

  (ii) Determining spot price. In the period since the release of the Rev. Proc., we have spent a significant amount of time discussing "spot price" with issuers, financial advisors, investment bankers, and others. While we are not expert regarding the pricing of securities, we would like to pass on our understanding of the issues relating to "spot price." We understand that it is extremely difficult and costly to attempt to recreate "spot price" on a historical basis as of the time that securities were purchased. Investment providers tell us that here is also great uncertainty as to the meaning of spot price: is it the interdealer price, the interdealer price plus markup, the prevailing market price (institutional or retail?), the bid price or the ask price? Does it matter if the dealer is a primary dealer or not? What if there are no contemporaneous prices? The only easily obtainable, published information that we are aware of is the closing prices in The Wall Street Journal and similar publications. We understand that, for a fee, there are also other sources of end of the day prices, as well as sources for real time prices. We are told that these sources of information have several significant defects: First, the end of the day prices ignore intra-day price volatility which, we understand can be significant. Second, these prices do not take into account any of the peculiarities of contingent, delayed-delivery purchases or open-ended offers. We are also told that the price quotes are often based on estimates, and therefore are generally not reflective of the securities purchased by issuers for escrows. Given these problems, the published prices need to be adjusted to take these factors into account. For example, to adjust for volatility and forward delivery, the higher of the opening or closing prices for a day could be adjusted upward by an arbitrary number of basis points simply to achieve rough justice. We note that such a mechanism would still ensure that many purchases that were significantly overpriced would result in a payment to the IRS that issuers would have to make out of their own pockets.

Even assuming that an appropriate reference price can be determined, we understand that it is still unclear how the "spot price" is to be determined, creating further, unacceptable ambiguities. For example, does "spot price" allow for the "extra" accrued interest to the actual day of settlement or does it merely reflect the price, including accrued interest, for regular settlement? A similar problem exists for securities that are traded and quoted on a yield basis: does "spot price" mean the price as of next day settlement or is it the price on the actual settlement day determined based on the yield quoted for next day settlement? Common sense indicates that these differences can be significant and result in ambiguities that a "simple" closing agreement procedure should not contain.

(iii) Other methods to determine closing agreement amounts. We do not have any specific suggestion for other closing agreement amounts. We do wish to emphasize that no fixed amount or formula can be established for all transactions. The IRS should provide flexibility to determine the settlement amount for determination on a case-by-case basis. While this will be more difficult for the IRS and issuers, it is the only fair way to proceed. Further, a more open-ended Rev. Proc. might get more issuers to volunteer to settle with the IRS. It does not make sense to have a program that requires the same amount to be paid regardless of the facts. Any such voluntary program would still be more effective than reliance on the examination program.

We would be pleased to discuss these matters further and provide any additional assistance. Any questions should be directed to Mitchell H. Rapaport at (202) 457-5305.

Sincerely,
Mitchell H. Rapaport