Duke Energy Corp. v. United States, 49 F. Supp. 2d 837 (W.D.N.C. 1999)

Herman Spence, Robinson, Bradshaw & Hinson, P.A., Charlotte, NC, Robert W. Fuller, III, Robinson, Bradshaw & Hinson, Charlotte, NC, for Duke Power Company, plaintiff.

Alan J.J. Swirski, U.S. Dept. of Justice, Tax Division, Washington, DC, Lawrence P. Blaskopf, U.S. Department of Justice, Tax Division, Washington, DC, William K. Rounsborg, Trial Attorney, Tax Division, U.S. Department of Justice, Washington, DC, for U.S., defendant.

MULLEN, Chief Judge.

This matter is before the Court after a bench trial which took place from September 23, to September 25, 1998. Duke Energy Corp. ("Duke"^[1]) brought this action for a refund of federal income taxes paid after an audit of its 1985 tax return. Pursuant to that audit, certain Duke deductions were disallowed. Duke exhausted its administrative appeals of the audit, paid the tax in question in 1994, and filed this action in 1997.

In 1985, Duke invested in an investment program formulated by 21st Securities Corp. ("21st"), a broker/dealer located in New York. This program was designated *838 the "Preferred Dividend Capture Strategy." The 21st program was designed to maximize the receipt of preferred stock dividends and short interest rebates, and the payment of dividend equivalent payments ("DEPs") in order to take advantage of the disparate tax treatment between dividends received and DEPs.

As a part of the 21st program Duke purchased preferred stocks and sold other preferred stocks "short." A short sale is a sale of borrowed shares. In order to borrow securities the borrower is required to provide cash collateral. The amount of such collateral is negotiable but generally is pegged at 102% of the market value of the underlying securities. The lender of the securities, usually the owner, invests the cash and returns a portion of the interest earned thereon to the borrower in the form of a "rebate." These rebates are negotiated, and vary from day to day security to security, and broker to broker. Due to the extremely dynamic and individualized nature of the rebate market, rebates that are paid to borrowers are typically calculated as a portion of the actual rebates and paid in arrears. Certain stocks are difficult to locate because there is a great demand to sell them short by speculators and arbitrageurs. The greater the demand for a certain stock the lower the expected rebate. Short interest rebates are fully taxable.

Duke received dividends on the long positions. As a corporation and subject to the limitations which make up the issues in this case. Duke was entitled to deduct 85% of these dividends in computing its taxable income. 26 U.S.C. § 243 ("I.R.C. § 243"). This deduction is referred to as the "Dividends Received Deduction," or "DRD." In order to qualify for the DRD a corporate taxpayer must own the securities for at least 46 days I.R.C. § 246(c) (1). That 46 day holding period is tolled if the investor is short "substantially identical" securities. I.R.C. § 246(c) (4) (A). The DRD is also disallowed if the investor is short substantially similar property. I.R.C. § 246(c) (1) (B). As the stocks Duke held short paid dividends, Duke made DEPs to the beneficial owner of such securities. DEPs are fully deductible as long as the short position was held for at least 46 days. I.R.C. §§ 162(h), 264(h). Duke reduced its taxable income by the amount of these payments on its 1985 consolidated tax return.

The IRS determined that Duke was not eligible for the DRD under I.R.C. § 246(c) (4) (A). The IRS also required Duke to capitalize, or defer recognition of the DEPs. The IRS originally based these requirements on its determination that Duke had invested in the functional equivalent of bonds and that it was not entitled to the preferential treatment afforded corporate dividends. Later the IRS changed its position, arguing that Duke had invested in an economic sham program such that it never had a reasonable expectation of return other than the projected tax benefits of the program.

This Court has jurisdiction over this action pursuant to 28 U.S.C. § 1346(a) (1).

The parties have stipulated that the amount in controversy is $859,001.00 plus statutory interest from March 15, 1986.

In 1985, Duke decided to invest some of its excess cash in short term investments Duke did not want to expose itself to a significant risk of loss of principle, but desired to achieve a return greater than then available in Treasury Bills or commercial paper. Mr. Rich Osborne, Duke's current Chief Financial Officer, and Mr. Rhem Wooten, who was in charge of the Duke investment in the 21st program in 1985, testified that Duke interviewed and examined a number of different investment managers and programs Duke eventually decided to invest in the Preferred Dividend Capture Strategy program designed and managed by 21st. Duke also invested *839 in at least two other dividend capture programs in 1985.

Mr. Osborne described the program. The Preferred Dividend Capture Strategy program consisted of purchasing certain preferred stocks and selling "short" other preferreds. 21st intended that investments in the program be hedged, i.e. that market price changes in the securities held long would be offset to great extent by the market price movements in the securities sold short. Preferred stocks, which are fixed income securities, are typically priced according to the market rate of interest for similar securities. Therefore, the price movements of the longs and shorts were expected to have a high negative correlation. This had the effect of reducing the overall risk of loss Duke was exposed to. The nature of the program and the IRS minimum holding period requirements dictated that the trades be entered and exited approximately eight times per year. Each entrance and exit, over the course of a 46 day holding period, was known as a "roll." The evidence showed that all of the stocks utilized in the program were preferred stocks. None of the preferred stocks held long by Duke were in the same industry as the stocks which were sold short.

The program included short interest rebate. Mr. Osborne testified that Duke expected to receive between 2% and 3% per annum on its invested capital in the form of rebate. Short rebates were negotiated among 21st. Bear Stearns and the lending shareholder. Mr. Osborne also stated that he monitored the program and that Duke's expectations were satisfied.

Duke was exposed to a risk of loss on its investments in the 21st program. Under certain market conditions or in the event of a significant adverse change in the credit worthiness of a particular issuer, Duke would have incurred significant losses or made significant profits (depending on whether Duke had purchased or sold short stock of the issuer). The evidence was that no such market conditions or such events affecting particular issuers occurred during the approximate five-year period of Duke's investment in the program, therefore Duke did not suffer significant capital losses. Mr. Osborne testified that Duke attempted to mitigate these risks to some extent by only approving "A" rated or better preferred stocks for purchase. However, because Duke was at all times long and short the preferred stocks of different issuers in different industries, the risk of such loss was present throughout 1985.

Chris Pedersen, the individual in charge of the 21st trading room, testified that he obtained lists of preferred stocks which were suitable for the program from Bear Stearns, a broker dealer allied with 21st ("Bear"). He would then contact each 21st client to offer those securities for purchase or sale in the upcoming roll. After receiving the orders, Mr. Pederson investigated each stock and recorded the trades at what he thought was the appropriate market price. He testified that this due diligence process regularly included calling dealers, obtaining quotes, and comparing dividend yields to relevant market interest rates. To execute each trade, Mr. Pedersen filled in the price on a trade ticket, timestamped the ticket, and reported the trade to the appropriate market and regulatory agencies.

Duke's expert witness, Mr. Ed Myers, testified that all of the trades executed by 21st for Duke's account were reasonably close to the market price of those securities Mr. Myers has over thirty years experience in the securities industry, virtually all of which was in preferred stocks. During the course of his career Mr. Myers kept a weekly log of representative market yields for preferred stocks of various credit ratings, treasury instruments, municipal bonds. Mr. Myers considers this pricing matrix to be proprietary and it was filed with the Court under seal at his request. Because Mr. Myers was forthcoming and *840 knowledgeable, and because his opinions were derived from his experience, training, records, and the other evidence in this case, the Court finds that he was a very credible witness and has accorded his testimony significant weight in formulating its findings.

Mr. Myers testified that preferred stocks tend to trade rather infrequently and often trade in the Over The Counter market on a cash or next-day settlement basis. Trades which do not settle "Regular Way," five business days after trade date in 1985, are not reported to the consolidated pricing service. Trades Which are executed after four o'clock in the afternoon are also not required to be reported. Mr. Myers estimated that between 20 and 30 percent of the institutional trading in preferred stocks are not actually reported for pricing purposes. Therefore, the published New York Stock Exchange ("NYSE") trading prices for a particular security at a particular time may not be representative of the actual market for those securities.

21st set the prices for each trade based on Mr. Pedersen's determination of the market at that time. Although a specialist or market maker would have been available to provide a bid or offer for the securities at any given point in time, that price would have been based on several factors including the specialist's projection of his ability to liquidate the securities purchased or sold over time. Mr. Myers testified that he was only able to valuate the propriety of the trades in question by comparison to his matrix of yields. He believes that to simply refer to historical trading information published in the financial press would not be a reliable method for determining the propriety of those prices.

The government's expert, Mr. Donald J. Puglisi, evaluated Duke's investments in the 21st program 1985 and subsequent years. Mr. Puglisi concluded that the program must have been manipulated because the net gains and losses were too small to have been actually realized in the marketplace. Mr. Puglisi also compared the results of Duke's investment in the 21st program with the results that Duke would have obtained had Duke's trades been at the prices reported in the Standard & Poor's Daily Stock Record Mr. Puglisi was able to identify one trade in Potlatch Corporation preferred stock which he felt was not executed at the then existing market price. The government queried Mr. Myers concerning this trade on cross examination. His testimony was that the Wall Street Journal of December 30, 1985, reported that the market for Potlatch preferred was Bid 100, Offered 110 21st closed out Duke's position in that security on that day at a price of 117 1/4. Mr. Puglisi concluded that this trade was not executed at the market. On redirect examination the following exchange took place between Mr. Fuller, representing Duke, and Mr. Myers:

Q: What I want you to do is using this sheet, explain to us the process of that you went through in terms of determining that you thought the price at which Potlatch was traded was a fair price. And refer to your exhibits as necessary.

A: Okay. This is a trading review which goes from 10-1 through the end of the year, would be December 31. It appears that the total number of shares that traded during that period, best I can comply compile here, about 2, 220 shares. The average spread that was quoted on the NYSE, and it does not show any size as to bid or offer, was ten points. For no the at 10-1 through 11-1, it appears that the market was something like 98 to 110. And then the market became, on the 18th of November, 100 to 115. And then subsequent to that, it became 110 to 125 with basically no trading or transactions taking place 12-2 was through the 4th was 110 to 125. Then it became *841 approximately 105-110, and then subsequently 100 to 110. First of all, these are not viable markets. There's no size indicated. During the entire period of time, a total of 220 shares traded.

This is not a viable market for a professional that was asked to price and/or trade this. There basically is no viable market on the stock exchange as far as I would be concerned at the time. to value and establish a value for Potlatch, the way you back into it, you look at a value of some related securities. And in particular, at this particular time in the program, R.J. Reynolds 12 96 preferred, which is an A rated A-minus rated industrial preferred with an 11 percent sinking fund with most features pretty much the same as Potlatch except Potlatch was rated BBB-, one notch below the rating of R.J. Reynolds. We I took a look at where that was valued and where that traded. That was listed on the stock exchange again, but there was no major trading that took place on the exchange. It was apparently upstairs Over The Counter which would be normal. At the time, R.J. Reynolds traded about at 10 percent which would put an A-rated premium, industrial preferred at 10 percent. Potlatch traded at 11 percent. The initial purchase was at 11.56. When they unwound the roll. I think it was at 10.56, but 11 percent to round it out. A one percent spread for a slightly lower rated preferred and a slightly higher premium, in my judgment, is a reasonable valuation of Potlatch. There is really no other information in particular where you can put your finger and say this is where it should be priced except on a comparative value.

This testimony both supports Duke's contention that the trade was executed at the market and illustrates the pitfalls in relying on published prices for preferred stock prices in general. As Mr. Myers explained, these pitfalls are especially prevalent when the post hoc analysis is being performed on trades of large size as those in the 21st program undisputedly were. Although significantly different than the published market, the Potlatch trades were executed within a reasonable market range for those securities.

There are other differences between published and actual market prices. When a preferred stock begins trading on the first day after the trade date for which trades will settle on the record date of a declared dividend, i.e. the ex-date, its price is expected to fall by the amount of the upcoming dividend. Mr. Puglisi admitted that his analysis did not take these price changes into account. Published prices were based on regular way settlements. A particular stock would trade ex-dividend on the exchange while it was trading cumdividend in the institutional OTC market. The price in these two markets would be different at the same moment in time. Mr. Puglisi did not quantify the amount or degree of error these pricing data deficiencies introduced into his analysis, and he could not state either the magnitude or direction of any resulting errors.

Mr. Puglisi's determination that Duke's net gains and losses in the program were too small to be real amounts to a hollow conclusion unsupported by the evidence. Mr. Puglisi excluded from his calculations the "short rebate" received by Duke while participating in the 21st program, which added 2-3% to Duke's pre-tax return on its investment in the 21st program. This exclusion effectively contradicts Mr. Osborne's testimony that the rebate was expected by Duke. The Court concludes that Mr. Puglisi's exclusion was not supported by the evidence.

The movement in the price of a preferred stock attributable to changes in *842 market interest rates is called Interest Rate Risk. To the extent that the securities held long and short react the same to given interest rate fluctuations Duke was able to limit its exposure to Interest Rate Risk. The movement in the price of a preferred stock attributable to company specific factors is called Credit Risk. Credit Risk can not be hedged by offsetting long or short positions because the price fluctuations are company specific Duke illustrated Credit Risk by reference to the drastic decline in the market price of General Public Utilities preferred stock occurring at the time of the Three Mile Island nuclear reactor leak in March, 1979. Investors owning General Public Utilities preferred stock would not have been able to reduce the losses resulting from Three Mile Island by being short other preferred stocks. Duke was never long and short the same preferred stock. In fact, Duke was never long and short preferred stocks in the same industry. Therefore, Duke was subject to both company and industry specific Credit Risk.

An investor who invests in an economic sham is not entitled to treatment under the Code. Gregory v. Helvering, 293 U.S. 465, 55 S. Ct. 266, 79 L. Ed. 596 (1935). In Gregory, the taxpayer planned to transfer shares of stock from her wholly-owned corporation to herself. The taxpayer accomplished this goal by organizing a new corporation, transferring the stock to the new corporation, and then liquidating the new corporation. While these steps literally satisfied the requirements of the Code for a tax-advantaged reorganization, the taxpayer accomplished all the steps in a few days and could articulate no business purpose for the "reorganization." Accordingly, the Supreme Court held that the stock distribution should be taxed as a dividend, stating: "The legal right of a taxpayer to decrease the amount of what otherwise would have been his taxes, or altogether avoid them, by means which the law permits, cannot be doubted [Citations omitted]. But the question for determination is whether what was done, apart from the tax motive, was the thing which the statute intended." Id. at 469, 55 S. Ct. 266.

In the Fourth Circuit the issue of economic substance is controlled primarily by Rice's Toyota World, Inc. v. Comm'r, 752 F.2d 89 (1985). In that case a car dealer entered into a sale and leaseback transaction for computer equipment structured to generate accelerated depreciation and interest expense deductions in the early years of the lease. The Tax Court disallowed the deductions, finding that the taxpayer had no business purpose other than achieving tax benefits in entering the transaction and that the transaction had no economic substance because there was no reasonable possibility of a pre-tax profit. The Tax Court concluded, in essence, that the car dealer, who had no interest in entering the computer leasing business, had simply purchased a tax shelter in an attempt to reduce tax liability on other income. The Court of Appeals affirmed in part, disallowing the deductions on all aspects of the program except for interest payments on recourse debt financing. In so doing, the appellate court set forth a two-part test for courts to follow in analyzing economic substance cases: "To treat a transaction as a sham, the court must find that the taxpayer was motivated by no business purpose other than obtaining tax benefits in entering the transaction, and that the transaction has no economic substance because no reasonable possibility of a profit exists." 782 F.2d at 91.

The government and Duke agreed that the Rice's Toyota analysis provides the proper framework for deciding the issues here. It is important to note that the government argued that Duke not be allowed to consider the short interest rebates in performing this analysis. The *843 government based its position on 1) The lack of computational precision in the agreement between 21st, Bear and Duke to share rebates, and 2) Bear's and 21st's respective rights to retain rebates before passing on any residual amounts to Duke. The Court disagrees with this view. The evidence showed that Duke was motivated by the expected returns on the program as a whole. To piecemeal out the various income components and then argue that the remaining parts would not comprise a valid business purpose for making the investment is to ignore the holistic approach to the program described by Messrs. Wooten and Osborne.

If Duke had at any point received less than what it perceived to be a fair apportionment of the rebates in question it could have simply left the program. Mr. Pedersen and others testified that each roll was a discreet investment in its own right, requiring approval of each participant for each trade. Duke not only participated in the "Preferred Dividend Capture Strategy" in 1985, it continued to do so for a number of years afterward. The evidence that Duke expected and received a pretax return, therefore that they did not participate in a sham transaction or series of transactions, was persuasive, believable and essentially uncontradicted. Mr. Osborne and Mr. Wooten both described Duke's expectations clearly and credibly, such that the Court finds that Duke did have an expectation of a pretax return throughout its participation in the 21st program.

Duke considered the rebate as part and parcel of the total return generated. The evidence showed that the short interest rebate was at all times between 2 and 3% of the market value of the short positions. This return, while below the yield available on Treasury instruments at the time, constitutes a not insignificant amount, given the low overall risk of the portfolio. The Court finds, therefore, that the short interest rebate expected to be and actually earned by Duke, together with all of the other aspects of the program, constituted a valid business purpose for entering into the 21st program. The short interest rebate was an important portion of the business purpose, but the overall business purpose was to achieve a favorable rate of return given the low risk of the hedged positions. Given the hedging strategies employed in the program, the short interest rebate was reasonably forecast by Duke to be equal to the pre-tax profit in the program. As the evidence indicated that Duke reasonably expected to receive a pre-tax profit, and that they were reasonably motivated by that expectation to enter into the program, neither Rice's Toyota element is present.

The government also made reference to Seykota v. Comm'r, 1991 WL 86320, in support of its position that Duke's position in the 21st program amounted to a sham in fact or economic sham. In Seykota, the tax court disallowed the deductions taken by investors in the Merit Option Markets program and the Merit Stock Forward Contract program. These programs, like the 21st program, consisted of a closed circle of investors who purchased and sold securities at the direction of the promoter, who had advertised the program as a tax shelter. Viewing those programs as a whole, the court found them to be economic shams. The Court finds Seykotas inapposite to the case sub judice because the critical facts are dissimilar. The Merit programs set prices by computer and generated losses for each investor which equaled the losses requested. There is significant evidence here that the prices at which Duke traded were at the market. Duke's officers responsible for investing in the 21st program testified that they desired both a positive pre-tax return and an after-tax return on Duke's investment. There was evidence in Seykota that the computer generated prices were designed solely to achieve the desired tax effects without regard to actual trading conditions at the time. That evidence was bolstered by the explicit requests of the participants *844 to receive specific deductions in each year. The government attempts to make much of a Duke internal memorandum referring to the 21st program as a "closed circle," as if that is an admission that the program lacked economic substance. The greater weight of the evidence is that the program was not a sham. The only evidence of price manipulation is the unsupported conclusion of Mr. Puglisi and the government's contention that, according to Seykota, the mere ability to use subjective judgment in setting prices is enough to render the program illicit. Upon questioning by the Court during closing arguments Mr. Swirski admitted. "Our evidence is the implausibility of the results ..." Those results, standing alone, are not sufficient to overcome Mr. Myers' testimony that the trades were executed at the market.

Duke does not have to prove that there was no discretion involved in setting any of the trading prices in the program. Such a requirement would effectively outlaw a large number of legitimate investments simply because there was no exchange or reporting entity to refer to when determining the price at which to enter and exit the investments. Assuming, arguendo, that Duke bear's the burden of proof on this issue, the Court finds that Duke has disproved the Rice's Toyota elements by a preponderance of the evidence, therefore the "Preferred Dividend Capture Strategy," as invested in by Duke in 1985, was not an Economic Sham or Sham in Fact.

As the rolls were exactly 46 days long, a finding that there was substantial similarity at any given point in the program would toll the running of the clock for a sufficient period to deny Duke's DRD. I.R.C. § 246(c) (1) (B). The leading case construing the "substantially identical" language in the pre-1984 version of the I.R.C. is Hanlin v. Comm'r, 108 F.2d 429 (3d Cir. 1939). In that case the Court ruled that two municipal bonds issued by the same issuer, varying only slightly in maturity were substantially identical. (Emphasis added). The Court also held, however, that bonds of the Louisville Federal Land Bank were not substantially identical to those of the St. Louis and Wichita Federal Land Bank even though both bonds were backed by similar security and even though two separate banks were both liable for debts of the other. Id. at 430-31. The Court focused on the differing credit risks of the individual banks, holding essentially that instruments would not be substantially identical even if the instruments were of the same nature and were issued by similar issuers. Id. Treasury Regulation § 1.1233-1(d) (1) states that "ordinarily, stocks or securities of one corporation are not considered substantially identical to stocks or securities of another corporation."

In 1984, Congress broadened the scope of impermissibly similar securities by changing the word "identical" to "similar." Its broadening has not, however, so enlarged the scope of the statute as to eliminate the consideration of subtle but important distinctions between securities. For example, the Supreme Court has held that two mortgage pools were "materially different" because the "legal entitlements" associated with the mortgages were "different in kind and extent." Cottage Savings v. Comm'r, 499 U.S. 554, 111 S. Ct. 1503, 113 L. Ed. 2d 589 (1991). Two preferred stocks representing claims on different issuers which are not convertible into each other or in some other way exchangeable clearly represent different sets of legal entitlements. They are claims on the assets of two different entities and are subject to different indentures. Duke was never long and short the same security or two securities of the same issuer. The different issuers were always in different industries, therefore the business risk profiles were distinct as well.

Mr. Puglisi expressed an opinion that Duke was short and long substantially similar property because there must have been manipulation involved in achieving *845 the very low net capital gains and losses they experienced. In other words, his opinion that the property was substantially similar rests on the same discredited foundation upon which his conclusion that the program was an economic sham is based. As the Court has already determined that the trades were executed at market prices, it must reject Mr. Puglisi's conclusion on this issue as well. Mr. Puglisi further undermined his conclusion by assuming in his analysis that all of the preferred stock purchased and sold by Duke in the 21st program were perpetual stocks without call features or sinking funds. In actuality, approximately 30% of the preferred stock purchased and sold by Duke in the 21st program had call features or sinking funds. These funds amount to distinguishing features further underscoring Mr. Myers' opinion that the short positions were not substantially similar to the property Duke held long. Therefore, Duke has established, assuming that it has the burden of proof on this issue as well, that they were not required to treat the long positions and short positions of different preferred stock issues, of differing types, from different issuers, in different industries, as being substantially similar property for purposes of the DRD.

The Court's findings that the trades were executed at market, that Duke's participation in the 21st program was not an economic sham or sham in fact, and that Duke's investments were not in substantially similar property were guided significantly by the expert testimony presented. The Court makes an express credibility determination and finds Mr. Myers believable and Mr. Puglisi not believable. Mr. Myers has over thirty years of experience in trading the kind of securities at issue in this case. This real world, market experience is vastly different than the academic and internal corporate experience Mr. Puglisi brought to the stand. The basic questions to be decided here involve trading and invoke Mr. Myers' experience more than Mr. Puglisi's.

Mr. Puglisi testified that, based on his experience working with issuers and investors of preferred stocks and his academic research and knowledge, a hedge of the sort executed by 21st could not achieve the results he observed without some sort of manipulation. Given Mr. Myers' opinion that the individual trades were actually entered at market prices and Mr. Puglisi's inability to identify any that weren't Mr. Puglisi's opinion is not credible. It was, rather, an example of post hoc ergo propter hoc reasoning and did not stand up to Mr. Myers' scrutiny.

Mr. Myers was very forthcoming on direct and cross examination. Mr. Puglisi, on the other hand, appeared to be more an advocate for the Government's position than a neutral expert. The data relied upon by Mr. Puglisi was shown by Mr. Myers to be unreliable and incomplete. The market price reports Mr. Puglisi relied upon did not include a significant number of trades and did not account for important market factors. When confronted with these shortcomings in his report and underlying data Mr. Puglisi effectively offered only his "experience" to prop up his conclusions. Yet this experience is vastly different than Mr. Myers' in the substance of their respective professional activities. Mr. Myers possesses the type of experience that directly relates to the questions at issue, and his testimony was relied upon in making these determinations.

The Court finds that Duke bears the burden of proof in this matter. Duke has established by a preponderance of the evidence that it is entitled to the deductions in question.

For the reasons set out above, Plaintiff is hereby ORDERED to recover the stipulated amount.

[1] Duke Energy Corp., formerly Duke Power Company, entered into the transactions at issue through Church Street Capital Corp., a wholly owned subsidiary. Church Street and Duke filed consolidated tax returns in 1985, therefore for simplification they will be collectively referred to hereinafter as Duke.