IRS Outlines Arguments and Preparations for Trial on Derivative Swaps

     Year(s)   Deficiencies     Additions      Damages

STIPULATION OF FACTS:

[2] Respondent proposed a stipulation of facts on May 23, 2000. Petitioner did not respond to that proposed stipulation. On October 12, 2000, petitioner proposed an alternative stipulation of facts. Because petitioner's proposed stipulation was just received, respondent has not yet reviewed it in detail.

ISSUES:

[3] The issues are set forth in Respondent's Pre-Trial Brief filed simultaneously with Respondent's Trial Memorandum.

WITNESSES RESPONDENT EXPECTS TO CALL:

1. Ballantine, John

[4] During the years 1992 and 1993, Mr. Ballantine was the Chief Credit Officer at FNBC responsible for setting policy for taking on credit risk. Mr. Ballantine will be called to testify about credit exposure policies and decisions relative to swaps, tracking of swap default rates and collateral, and the dividing of real estate and non-real estate credit exposure.

2. Bouchard, Craig T.

[5] During 1992, Mr. Bouchard was head of Global Derivatives for FNBC, stationed in Chicago. Mr. Bouchard will be called to testify about FNBC's swap pricing, its swap business, swap profits and FNBC's profit and loss statements for swaps.

3. Boule, Suzanne J. (IRS Employee)

[6] Ms. Boule is currently a technical advisor with the Internal Revenue Service, and was the Chief, Financial Industry Studies, Internal Revenue Service Office of Financial Products and Transactions at the time of the IRS project to develop derivatives valuation software in conjunction with the Los Alamos National Laboratory ("Los Alamos Project"). If respondent does not prevail in his motion to exclude all evidence with respect to the Los Alamos Project, Ms. Boule may testify about the Los Alamos Project, depending on what, if any, evidence petitioner proffers.

4. Carroll, Edward R. (IRS Employee)

[7] Mr. Carroll is now with the IRS office of International, Compliance, National Office, and was formerly involved with the Los Alamos Project. If respondent does not prevail in his motion to exclude all evidence with respect to the Los Alamos Project, Mr. Carroll may testify about the Los Alamos Project, depending on what, if any, evidence petitioner proffers.

5. Carney, Owen (expert)

[8] The expert report of Mr. Carney, Bank Capital Markets Consulting, on bank regulatory concerns and bank reporting with respect to swaps has already been submitted to the Court. Mr. Carney will be called to introduce his report into evidence as his direct testimony and for redirect or rebuttal testimony, as necessary.

6. Commodities Future Trading Commissioner ("CFTC")

[9] An employee of the CFTC will be called to lay a foundation to introduce into evidence the CFTC study of derivatives entitled OTC Derivatives Markets and Their Regulation along with working papers 1, 2, 3A through 3E, 4 through 6 and 7A through 7C (Oct. 1993), if petitioner does not stipulate to the document.

7. Dalrymple, John (IRS Employee)

[10] Mr. Dalrymple is the current Commissioner of the Wage and Income Division, and the former IRS Chief Operations Officer, a position he held at the time of the Los Alamos Project. If respondent does not prevail in his motion to exclude all evidence with respect to the Los Alamos Project, Mr. Dalrymple may testify as to his role in the decision not to authorize additional funding for the Los Alamos Project, depending on what, if any, evidence petitioner proffers.

8. Bank One Corporation

[11] Bank One employees, or other associated individuals, will be called, only if necessary, to lay a foundation for entering into evidence the documents which the petitioner produced during formal discovery and narrative assertions in petitioner's formal discovery responses if the parties do not stipulate to these matters prior to trial.

9. Garrigus, Mark

[12] During 1991 though 1993, Mr. Garrigus was a Senior Vice President, Credit Policy Group and the Secretary of CRESCO (Credit Strategy Committee). Mr. Garrigus will be called to testify about the CRESCO loss reserve factors (risk factors), the methodology used in determining the factors, the problems with real estate-related exposures of FNBC during the years at issue, and the dividing of the real estate and non-real estate credit exposure. Mr. Garrigus will also testify on the conservative nature of petitioner's reserve policies.

10. Gendron, David (canadian Imperial Bank Of Commerce)

[13] Mr. Gendron, or another representative of Canadian Imperial Bank of Commerce, will be called to testify about the terms and price negotiation of a buy-out of a swap entered into with petitioner.

11. GMAC Capital Markets

[14] A representative from General Motors Acceptance Corporation will be called to testify about the terms and price negotiation of a buy-out of a swap entered into with petitioner.

12. Harvey, David M.

[15] During 1991, Mr. Harvey ran the Risk Insurance Department ("RID") that included petitioner's swap business. Mr. Harvey will testify about the allocation of petitioner's non-swap expenses to the RID Department, swap expenses, trader bonus pools, and swap profit and loss statements (also called management reports).

13. Hogan, Edward.

[16] Mr. Hogan is a managing director at Citibank, a counter party to some of the swaps which are at issue in these cases. Mr. Hogan worked in the Citibank derivatives department during the years at issue and will testify about the pricing of swaps by Citibank in the years at issue, the pricing of buyouts in the years at issue and how Citibank assess the profitability of swaps in the years at issue.

14. Hohurst, James H.

[17] During 1992, Mr. Hohurst was involved in overseeing swap trading at FNBC, reporting to Mr. Harvey, and will testify about swap pricing, FNBC's swap business, the required return on capital for swaps at FNBC, swap expenses, profit and loss statements (also called management reports) and allocation of trader bonuses.

15. Hsieh, David A.

[18] During 1987 through 1992, Dr. Hsieh was a consultant hired by petitioner to develop a statistical model that was used in petitioner's Variable Exposure Products (VEP) system. Dr. Hsieh will testify about his understanding that the model was to be used as a risk management tool, and not a credit risk valuation tool.

16. Jawar, Mary Baumann

[19] During 1989 through 1998, Ms. Jawar was a marketer in RID. Ms. Jawar will testify about her role in pricing swaps as a marketer, how FNBC priced swaps, her knowledge about how bonuses were determined, FNBC's credit problems due to real estate loan losses, swap buyouts and entering into swaps with her client base, which consisted of financial institutions.

17. Knudsvig, Paul

[20] Mr. Knudsvig, or a representative of Alliance Capital Management, will be called to testify about the terms and price negotiation of a buy-out of a swap entered into with petitioner.

18. Kozicki, Perry (IRS Employee)

[21] Respondent may call Mr. Kozicki, if necessary, to testify to summary schedules exchanged with petitioner if petitioner does not stipulate to respondents schedules.

19. Kopera, Bernard (IRS Employee)

[22] Mr. Kopera was the financial products specialist assigned to the audit of the 1992 and 1993 years of petitioner. If petitioner introduces pre-notice information, Mr. Kopera may be called to respond to the proffered evidence, if necessary.

20. Kwintiuk, Andrew (IRS Employee)

[23] Mr. Kwintiuk was the financial products specialist assigned to the audit of the 1990 and 1991 years of petitioner. Respondent may call Mr. Kwintiuk, if necessary, to testify to summary schedules exchanged with petitioner, if petitioner does not stipulate to respondent's schedules. If petitioner introduces pre-notice information, Mr. Kwintiuk may be called to respond to the proffered evidence, if necessary.

21. Lamond, Darcie J.

[24] During 1991 and early 1992, Ms. Lamond was involved in overseeing swap marketing and trading for Tokyo, London and Chicago, stationed in Chicago. Ms. Lamond will testify about FNBC's swap business in Chicago, swap spreads, swaps pricing, risk management, Devon Derivatives System ("Devon"), FNBC's standing in the inter-bank dealer market and her understanding of charges to the swap business for use of bank credit line resources.

22. Landusky, Mike

[25] Mr. Landusky is a former employee of the petitioner, now with IBM. Mr. Landusky, a computer technician, will testify about the Devon system, related software, and the environment which supported Devon.

23. Lieb, Barry (IRS Employee)

[26] Respondent may call Mr. Lieb, if necessary, to testify to summary schedules exchanged with petitioner, if petitioner does not stipulate to respondent's schedules.

24. Lowinger, Lynn Terese

[27] During the years at issue, Ms. Lowinger was the manager responsible for FNBC's Chicago traders, reporting to Mr. Hohurst and Ms. Lamond. Ms. Lowinger will testify about interbank dealers, Devon, her interaction with Arthur Anderson audit staff, and the system for monitoring short terms swaps.

25. McCafferty, Belinda (IRS Employee)

[28] Ms. McCafferty is the Area Manager, Governmental Liaison and Disclosure, Area 3, Atlanta Georgia. At the time of the Los Alamos Project, she was the Director, Office of Financial Products and Transactions, Office of Assistant Commissioner, (International). If respondent does not prevail in his motion to exclude all evidence with respect to the Los Alamos Project, Ms. McCafferty may be called to testify about the Los Alamos Project, depending on what, if any, evidence petitioner proffers.

26. McMahon, Larry

[29] Mr. McMahon is a former employee of the petitioner, now with IBM. Mr. McMahon, a computer technician, will testify about Devon, related software, and the environment which supported Devon.

27. Moravey, Joe

[30] During the years at issue Mr. Moravey was at Arthur Anderson and was one of the audit partners for the audit of FNBC. Mr. Moravey will be called to testify about the audit.

28. Nagel, Emerson Howell

[31] During the years at issue, Ms. Nagel worked on quantifying the credit exposure for FNBC products. Ms. Nagel will be called to testify about FNBC's Variable Product Exposure ("VEP") system and FNBC's Credit Exposure Model ("CEM") amount for swaps.

29. Nolan, Deborah (IRS Employee)

[32] Ms. Nolan, formerly Deputy Assistant Commissioner, International, is currently the Division Deputy Commissioner, Large and Mid-Sized Business. If respondent does not prevail in his motion to exclude all evidence with respect to the Los Alamos Project, Ms. Nolan may testify about the Los Alamos Project, depending on what, if any, evidence petitioner proffers. Ms. Nolan may be called to testify about the Los Alamos Project *

30. O'Brien, Patricia (Expert)

[33] The expert report of Dr. O'Brien on accounting conventions and the accounting practices of the petitioner has already been submitted to the Court. Dr. O'Brien will be called to introduce her report in evidence as her direct testimony, and for re:-direct and rebuttal testimony, as necessary.

31. Office of the Comptroller of the Currency ("OCC")

[34] An employee or custodian of the records for the OCC, will be called to lay a foundation, if necessary, for the introduction into evidence of the Bank Examiners Reports for FNBC for the 1990 through 1993 years.

32. Osterling, Steve

[35] Mr. Osterling is am employee of the petitioner, a project manager who oversaw the implementation of the Devon system. He will be called to testify about the Devon system and its capabilities for reproducing data.

33. Packer, Mary Bliss

[36] During the years 1990 through mid-1993, Ms. Packer worked at FNBC on profit center reports and, starting about mid-1993, she worked as a functional comptroller for derivative products with responsibility for calculating the carve-outs at issue. Ms. Packer will testify about calculating credit risk carve-outs, determining administrative expense carve-outs, swap default rates and her involvement in tracking swap defaults, profit center reports, and double-count procedures for determining bonuses.

34. Parsons, John (Expert)

[37] The expert report of Dr. Parsons on the valuation issues in these cases has already been submitted to the Court. Dr. Parsons will be called to introduce his report in evidence as his direct testimony, and for redirect and rebuttal testimony.

35. Pettersen, Bjorn

[38] During the years at issue, Mr. Pettersen traded Canadian dollar denominated interest rate swaps for FNBC out of Chicago. Mr. Pettersen will be called to testify about swap pricing and swap trading.

36. Price Waterhouse

[39] An employee or custodian of the records for Price Waterhouse will be called to lay a foundation, if necessary, for the introduction into evidence of the January 8, 1989 letter from Price Waterhouse to the Federal Financial Institutions Examination Counsel contending there is no precedent, either accounting or regulatory, for market value accounting accompanied by deferrals and amortization.

37. Raghavan, Vijay

[40] During late 1993 Mr. Raghavan worked at FNBC at modifying the VEP system and had other responsibilities over aspects of RID. Mr. Raghavan will be called to testify about VEP, trading statistics, trading and trade pricing.

38. Rauchenberger, Lou

[41] During the years at issue Mr. Raauchenberger worked at J. P. Morgan, a counterparty to some of the swaps at issue, in J.P. Morgan's derivatives business. He was responsible for valuations and information support. Mr. Rauchenberger will be called to testify about J.P. Morgan's practices with respect to credit risk and servicing cost adjustments for swaps.

39. Rhoten, Doug

[42] Mr. Rhoten brokered swap deals at Garvin Guy Butler, a major swap broker, in the years at issue and will provide factual testimony, based on his experience in the years at issue, on the primary swap market, the types of clients his firm dealt with, and how swaps traded and swap assignments.

40. Taranemko, Yuri

[43] Dr. Taranemko is a scientist formerly with the Los Alamos National Laboratory, who was involved in the Los Alamos Project. If respondent does not prevail in his motion to exclude all evidence with respect to the Los Alamos Project, Dr. Taranemko may be called to testify about his involvement with the development and testing of the Los Alamos software, depending on what, if any, evidence petitioner proffers.

41. Thomas, Phyllis S.

[44] During the years at issue, Ms. Thomas was a marketing manager for RID in Chicago. Ms. Thomas will be called to testify about the role of marketers in entering into and pricing swaps, as well as FNBC's pricing matrix.

42. Vitale, Dave

[45] During the years at issue, Mr. Vitale was Head of Corporate and Institutional Banking, Vice Chairman of the Board of Directors and a CRESCO member at FNBC. Mr. Vitale will be called to testify about petitioners Global Derivatives Product Financial Organization Summary, the hurdle rate of return on capital for derivatives and the derivatives business of FNBC.

43. Winn, James

[46] During the years at issue, Mr. Winn was a credit officer at FNBC. Mr. Winn will be called to testify about credit exposure, FNBC's credit exposure limits such as VEP limits ("VEPL") which allocated credit exposure amounts to the derivatives unit, CEM amounts, the credit approval process, FNBC credit classes, Risk Acceptance Criteria ("RAC") worksheets, FNBC's own risk classification, and FNBC's CEIS0185 Reports.

[47] Respondent reserves the right to call all witnesses listed in petitioner's trial memorandum, as necessary, if they are not called by petitioner, and impeachment witnesses, if any are necessary.

CURRENT ESTIMATE OF TRIAL TIME: Two weeks

SUMMARY OF FACTS: ¹

I. PETITIONER'S SWAP BUSINESS

[48] During the years at issue, petitioner's swap trades were transacted through its Risk Insurance Department ("RID"). RID traders were responsible for all interest rate derivatives, which included swaps, caps, floors, collars, options and combination deals, e.g., swaps with embodied options. RID traders were also responsible for currency and commodity swaps. During the years at issue, RID included Chicago, London, and Tokyo offices. Only swaps booked through the Chicago office are at issue in these cases.

[49] During the years at issue, two traders were responsible for the interest rate swap trades, one transacting trades in U.S. dollar denominated swaps and the other Canadian dollar denominated swaps. At some point during the 1990 through 1993 period, the U.S. dollar denominated swap trading was split between two traders, one for short-term swaps and one for long-term swaps. Departments other than RID traded exchange traded contracts, bonds and other instruments.

[50] RID had a front office, which essentially consisted of the traders. RID had a back office which handled the documentation such as ensuring that swap master agreements were executed, confirmations were received, and periodic payments were properly transacted. Other departments separate from RID handled accounting and tax reporting.

[51] In addition to traders, RID included marketers, who promoted petitioner's swap business to its customers. Marketers were assigned groups of customers, such as financial institutions. However, each customer was "owned" by a relationship manager, sometimes also referred to by petitioner as a banker, who was not part of RID. The relationship manager had overall responsibility for the customers' transactions, e.g., all bond issuances, letters of credit, loans, derivative transactions and other accounts. A credit officer, who was not part of RID, was also assigned to each customer.

A. TRADER'S PRICING OF SWAPS

[52] Swap prices are quoted by petitioner's trader to other dealers and to petitioner's marketers, who in turn quote prices to end-user customers. In their quotes, the traders do not build-in an increased spread for credit or administrative expense carve-outs, they rely on the prevailing market. Traders risk-manage their own portfolios, but are subject to trading limits.

[53] The traders enter into swaps with other dealers based on the trader's own quotes without a marketer being involved, with the possible exception of a dealer that petitioner has notified its traders as having substantial credit problems. The traders enter into swaps for petitioner's customers only after consultation with marketers since the traders do not interact directly with the end- user customers.

B. MARKETERS' ROLE IN PRICING SWAPS

[54] Marketers promote swaps to petitioner's customers, they do not execute trades. Marketers communicate the terms of a proposed swap for an end-user customer to a trader and obtain a price quote. The marketer can build in an additional spread, but cannot decrease the price quoted by the trader without the trader's approval. The trader must sign the trade ticket and is responsible for risk managing the swap once executed. The marketer has no responsibility for market risk management.

[55] For swaps with customers who knew the going market, the marketers generally could not increase the spread without losing the deal. While marketers were generally aware that there were administrative costs for swaps, increased spreads were not added if the customer would not do the deal with an increased spread. If a marketer recommends a decreased spread to a trader, it is because the customer said it could get a better deal elsewhere or to enhance the bank's overall relationship with the customer.

C. CREDIT OFFICERS

[56] Before a trade can be entered into, there must be a line of credit exposure approved for the counterparty, be it a dealer or end-user counterparty. Credit officers, and the approval process, are separate from RID. The credit line for derivative products is called the VEP limit. If the credit exposure of a swap exceeds the available credit, or no credit has been approved, the trader must obtain credit approval from a credit officer.

D. RISK MANAGEMENT

[57] Each trader is responsible for risk managing his or her own portfolio. Exposure on swaps are managed with other swaps and with other instrument such as Treasuries. Daily risk profiles are prepared and printed by petitioner's Devon Derivative System. The risk profiles and the other tools petitioner's traders use to risk manage their portfolio do not have daily reports on administrative expense carve-outs or credit risk carve-outs. These carve-outs are only calculated on a quarterly basis.

E. COMPENSATION

[58] At some time prior to year-end, the size of bonus pool for RID is determined by upper management. The pool is allocated between geographical areas and among traders and marketers. For the allocation, petitioner relies on management reports, also referred to as profit and loss statements, that can divide the profits between geographic areas and among the various deals, e.g., swap traders, option traders, etc.

[59] The marketer's bonus is based on (a) a double-count system for the spread the trade quoted over the Devon mid-market and (b) any add-on spread the marketer added to the deal. The double-count system lets both the trader and the marketer be credited with the spread over the Devon mid-market amount: full credit for the trader and one- half credit for the marketer. The marketer gets sole and full credit for any 'add-on' spread they obtain.

F. DEVON DERIVATIVES SYSTEM

[60] Petitioner used its then-current version of its Devon Derivatives System ("Devon"), with proprietary modifications, to value its swaps using a discounted cash flow methodology. This methodology forecasts the future cash flows for the swap from market data and determines the net present value of the cash flows. The mid- market valuation method which petitioner claims to have utilized for the years at issue is a recognized discounted cash flow methodology for deriving the value of a swap from market bid and ask data using the average of the market bid and ask for swaps, i.e., the mid- market, to project cash flows. Devon did not take into account the effect of collateral, netting or other such credit enhancement.

[61] Respondent's experts have been able to verify that petitioner's December 20, 1993 Devon values were mid-market values. Petitioner has not established that the same is true for for 1990 through 1992. Petitioner has not retained all the data necessary to establish whether its Devon values overvalue, undervalue, or correctly value its swaps in that: (a) except for year-end 1993, petitioner does not know the actual swap curves used by it to arrive at the projected cash flows of the swaps for the years at issue; (b) petitioner does not know what version of its valuation software was utilized at each year-end; and (c) petitioner does not know what modifications to the then-current version of its software were used for its year-end valuations. See Respondent's Pre-Trial Brief, Section III, Substantiation, for more detail on petitioner's failure to keep the necessary records.

II. CREDIT RISK AND ADMINISTRATIVE EXPENSE CARVE-OUTS

[62] In determining its swap income for tax purposes, petitioner uses a three-step process to arrive at the amounts it uses for swap values. The first step involves the use of the Devon Derivatives System to determine the net present values of the expected cash flows of the swaps, the Devon values. The second step is the calculation of the credit risk carve-out, which unlike the calculation of the Devon values, is not recalculated annually for each swap. The credit risk carve-out is a reduction to the Devon values in the year of inception and is then amortized back into income over the average weighted life of the swap. This process will be described in more detail below. Finally, the third step is the calculation of the administrative expense carve-out, which is an adjustment to the Devon values. That process is also described below.

A. CREDIT RISK CARVE-OUTS

[63] For the years at issue in these cases, petitioner calculated the credit risk carve-outs in two different ways. For the period from the last quarter of 1992 through 1993, petitioner used one method. For the period from 1990 through the third quarter of 1992, the petitioner used another method which it has yet to fully explain, other than to say that it was done by the Devon system.

[64] For the last quarter of 1992 through the 1993 period, at the inception of each swap, petitioner determines the initial credit risk carve-out. Carve-outs are only claimed at the inception of a swap. The carve-out is calculated based upon Credit Exposure Model ("CEM") amounts derived from the Variable Exposure Products ("VEP") system and the CRESCO loss reserve factors (also known as risk factors) derived from a counter party's risk classification. See the discussion below for each of these factors.

[65] The credit risk carve-out calculation is actually made on a quarterly basis, for each swap that is entered into that quarter. ² Credit risk carve-outs are claimed on other RID instruments, such as swaptions and options. To understand petitioner's credit risk carve-out account, it is important to know some background information on all types of instruments. First, petitioner claims credit risk carve-outs on all instruments traded by RID traders. This includes all types of swaps (U.S. dollar denominated interest rate swaps, Canadian dollar denominated interest rate swaps, currency swaps, commodity swaps and a type of swap petitioner refers to as a "COMB", shorthand for combination deal since such a swap usually has an embedded feature such as an option), options, interest rate guarantees (petitioner's reference for caps, floors, and collars), and swaptions.

[66] Second, while only swaps are at issue in these cases, the accounts to which credit risk carve-outs for swaps are posted also include credit risk carve-outs for other instruments which are not at issue. For example, in 1993, credit risk carve-outs for U.S. dollar denominated interest rate swaps, swaptions, and some Canadian dollar denominated swaps were posted to petitioners balance sheet account number 194-130-0007. Credit risk carve-outs for currency swaps, commodity swaps and some Canadian dollar denominated interest rate swaps were posted to other balance sheet accounts. Thus, there is no one account that can be examined to see the amounts at issue.

[67] Third, when the credit risk carve-out amounts are posted to income accounts they pour into unrealized trading income accounts which are unique to each instrument. While swaptions, U.S. dollar denominated interest rate swaps and some Canadian dollar denominated interest swaps are all posted to balance sheet account 1194-130-0007, they are posted to separate unrealized trading income accounts, e.g. Canadian dollar denominated interest rate swaps, U.S. dollar denominated interest rate swaps, and swaptions each have their own unrealized trading income account.

[68] To determine the carve-out amounts taken for each type of swap instrument at issue (i.e. U.S. dollar and Canadian dollar denominated interest rate swaps, currency swaps, COMB and commodity swaps), one must know each account to which credit risk carve-outs are posted. To date, respondent does not know all of the accounts. Petitioner has, however, produced a list of all swaps entered into in 1993 which, since carve-outs are only claimed on swaps at their inception, includes all swaps for which credit risk carve-outs were claimed in 1993. The list includes 488 swaps, 418 of which had carve- outs. The list has carve-outs which total to $982,417 for the five types of swaps at issue. Petitioner has provided detailed information about the postings for only one category of swaps, U.S. dollar denominated interest rate swaps. The credit risk carve-outs are posted each quarter to some balance sheet account. Also posted to the balance sheet account are amounts to record the amortization back into income of the credit carve-out amounts, to the extent they are so amortized. ³ The carve-outs and amortization are also posted to the respective unrealized income account to which the revaluation income for that type of instrument is posted.

[69] Petitioner never recalculates the credit risk carve-out amount for each swap again. Instead, when the Devon values for each swap are recalculated annually to determine swap values, petitioner simply subtracts the unamortized credit carve-out adjustment from the recalculated Devon values, less the recalculated administrative expense carve-out amount.

[70] Petitioner calculates the initial credit risk carve-out amount using several factors. First, petitioner calculates the CEM amount. Second, petitioner looks to the risk classification of the counter party to determine the appropriate risk factor (CRESCO loss reserve factor) to be used for each swap counter party. The risk factor provides the percentage that is multiplied with CEM to give the credit risk carve-out amount for each swap. Respondent expects to prove at trial that these factors used in arriving at the carve-out amount (CEM, risk class and risk factor) are not appropriate measures of current credit exposure. Although the petitioner's system may be appropriate for the purposes that it was established, i.e., risk management and measuring capital reserve adequacies, it is not an appropriate system for calculating fair market value.

1. CEM

[71] The CEM amount is a measure of the maximum credit exposure, or potential loss on the swap were a counter party to default and the petitioner was unable to recover anything. In other words, the CEM amount reflects petitioner's measure of the maximum amount of income it might possibly receive on each particular swap, which petitioner treats as the most it could lose on that particular swap. Unlike loans with fixed interest rates, the amount of income a swap may generate is not known at the inception of the swap. Nor is the amount of potential loss known.

[72] Petitioner originally used CEM for swaps in order to evaluate market risk and reserve adequacies. Petitioner's concept of CEM is not limited to swaps. The bank determines a CEM amount for all transactions with a customer, whether the transactions are be they loans, letters of credit, forward rate agreements or swaps.

[73] The VEP system is used to calculate the CEM amount for swaps. The systems used to calculate the CEM amount for other types of products is not known to respondent. Regardless of how the CEM amount is computed, under petitioner's policies, no transaction can be entered into with a customer if the transaction exceeds the determined allowable unused CEM amount that the bank has left for said customer for that type of product, unless a special exception is approved by a credit officer with sufficient authority to grant the approval.

[74] The CEM amount is calculated by inputting the transaction information, such as, notional principal amount, tenor, type of instrument and interest rate terms from the confirmations into the VEP system. Also inputted into the system for aggregation are the mark-to-market amounts (Devon values) for each particular swap. The VEP system then produces a CEM amount. The CEM amount was recalculated at various times throughout the year. During the period at issue here, the CEM amount was calculated quarterly, and eventually on a monthly basis. If the CEM amount exceeded the mark- to-market amount (Devon value) for a swap, during this time period, the CEM was recalculated on a monthly basis.

[75] Petitioner uses the CEM amount from its VEP system to calculate the credit risk carve-outs claimed for all swaps, short- term (eighteen months or less) and long-term. Yet, petitioners own internal audit report pointed out that the VEP system could not accurately calculate the CEM for short-term swaps. It is not clear to respondent what alternative systems were available in these years, but there was at least one alternative system, Trading Credit System ("TCS") that was used at some time late 1993 for short-term swaps. Thirty percent of the swaps on which petitioner claimed credit-risk carve-outs in 1993 were short-term swaps. Thus, the CEM amount used for credit risk carve-outs for 30 percent of the swaps is not even an accurate maximum potential credit exposure.

[76] The CEM is a statical simulation model implemented by a computer model. CEM begins with a statistical model of interest rates based upon historical interest rate data. It assumes that the interest rate in any future period is a random variable described by a probability distribution. The variance of the distribution measures the degree of uncertainty about future interest rates. See Expert Report of Dr. Parsons, Appendix C, for detailed discussion of the simulation model. Basically, it uses a Monte Carlo simulation with 10,000 potential variations of quarterly interest rates over the remaining term of the transaction. The statistical model used to produce the CEM amount does not return a single value, rather it provides a complete probability distribution of possible losses. Petitioner must then specify a cut-off level to get the CEM amount. For the period from the last quarter of 1992 through 1993, petitioner used an 80 percent cut-off level. ⁴ The choice of a cut-off level is arbitrary and a different cut-off level produces a different CEM amount, which in turn produces a different credit risk carve-out.

[77] The available aggregate CEM amount per customer is divided between various departments of petitioner, with the derivatives department having its own portion of the available CEM amount to work with on a per customer basis. Petitioner's relationship manager assigned to the customer, with the credit officer, is responsible for dividing the available aggregate CEM amounts between departments and among product types. The portion of the aggregate CEM amount allocated to interest rate derivative products such as swaps is called the VEP Limit ("VEPL"). A trader can enter into a swap without consultation with a credit officer if the CEM amount for the swap is within the available VEPL for that customer. The VEPL does not always distinguish between short-term swaps and the maximum tenor for swaps with that customer; the impact of this lack of differentiation is discussed in more detail in the section on risk factors. It is not known if the VEPL distinguish between swaps and other interest rate derivatives such as caps, floors and collars.

[78] While the CEM amount available under the VEPL may be a good measure of whether petitioner, itself, should enter into another swap with a particular customer based on existing exposure to that customer, it is not a measure of whether interbank swap dealers in the marketplace with less exposure to that customer, or higher tolerance for exposure to that customer, would be willing to enter into a swap with petitioner's customer at prevailing rates.

[79] While petitioner contends that its methodology excluded the potential of claiming credit risk carve-outs on any swaps where petitioner was the expected net payor of the projected cash streams (thus not subject to recognizable counter party credit risk when netting is considered) this claim cannot be substantiated. The numbers alone indicate that petitioner's claim is doubtful. For the 488 swaps petitioner entered into in 1993, only 70 had CEM amounts of zero. Thus, petitioners CEM calculation is such that petitioner is the expected net positive recipient of the cash flows in 82 percent of the swaps it entered into in 1993. No data was produced by petitioner for prior years from which a similar analysis could be calculated.

[80] The CEM amount does not reflect the impact of netting agreements. For the years at issue, most of petitioner's swaps are covered by ISDA Master agreements ⁵ with netting provisions, i.e., if the counter party does not pay FNBC on its leg of the swap, then FNBC does not have to pay the counter party on the other leg. Recognizing that netting has a large impact, petitioner changed the CEM cutoff level from 95 percent to 80 percent for the years at issue, but never tested to see whether this adjustment equated to the reduction in credit exposure generated by the netting provisions. In fact, petitioner never tested either the 95 percent or 80 percent cutoff level for purposes of determining what impact netting had on these numbers. Yet, the confidence level impacts the resulting CEM amount.

[81] Nor does the CEM amount consider the impact on the potential exposure amount of cross defaults agreements, offsetting swaps or any other individual variations. The CEM amount does not vary based on the counter party, the risk class of the counter party or collateral. It is a swap-by-swap calculation driven solely by the projected movement of interest rates, based on the interest rates in the confirmation for that swap, the tenor of the swap and the notional principal amount.

2. RISK CLASSES

[82] The second factor necessary for the calculation of the credit risk is the risk class assignment. Petitioner assigns a credit risk classification to each swap counterparty. Petitioner does not assign one risk class per customer; rather, the risk class ratings are assigned to types of products (called "facilities" by petitioner) which a customer is expected to transact. For example, a customer may have one risk class rating for interest rate derivatives such as swaps, and another rating for other products petitioner labels "trading products" which include instruments such as short-term forward rate agreements and other instruments. The risk class assigned for all swaps under a particular VEPL is generally assumed to be the same for all tenors (maturities) within that VEPL, unless special risk class approval is assigned to a single transaction. Thus, a customer may have a risk class of 2 assigned for a VEPL that covers swaps ranging from three months to ten years with the result that short-term swaps and long-term swaps have no differentiation in terms of credit risk.

[83] It is petitioner's credit officers who assign the risk class to a counter party. According to a former senior credit officer of petitioner, Mr. James Winn, many factors go into this determination, including reviews of financial statements, public debt ratings, other available public information, credit exposures existing with petitioner and the tenor of the swap for which the rating is to be determined. Mr. Winn also stated that a counter party's risk classifications (of which there could be several, depending upon how many facilities a counter party had) were reviewed at least annually, and more frequently if subsequent events warranted it. Generally, the relationship manager and the credit officer were responsible for monitoring the counter party at all times.

[84] During the period of the last quarter of 1992 through 1993, petitioner used a credit risk classification system consisting of nine classes. Below is a chart comparing the relationship between petitioner's risk classes (numerical designations) and the S&P ratings are as follows:

          Risk Class          S&P Equivalent

[86] The petitioners assigned risk classes are generally lower than the credit rating assigned to an entity by public rating institutions such as S&P. When respondent refers to a lower risk rating in this memorandum and in its Pre-Trial Brief, it generally means risk classes 4 through 10. A higher risk rating generally refers to risk classes 1 though 3 (investment grade). A lower risk rating by petitioner results in a higher risk factor being multiplied against the CEM amount and, thus, a larger credit carve-out amount.

[87] The description of risk classifications do not take into consideration collateral or any other type of credit enhancements. If petitioner's employees fail to input the correct risk classification for petitioner's customers into the VEP system, the system defaults to a risk class 3 which ultimately results in an overstatement of the credit risk carve-out amount, even under petitioner's approach, if the party was actually a risk class 1 or 2. Although the input of the risk class into the VEP system does not directly impact the CEM amount, when petitioner calculated the carve-out amount (CEM multiplied by the risk factor), if the risk class was not inputted, petitioner used a risk factor appropriate for a risk class 3, according to a former employee of petitioner, Ms Mary Bliss Packard, who calculated the credit risk carve-out for part of the time period at issue.

[88] Once the risk class was determined for a counter party for the swap transaction the next step was to determine the applicable corresponding risk factor.

3. RISK FACTORS

[89] The risk factors are multiplied by the per swap CEM amount to arrive at the credit risk carve-out amount for each swap. Which risk factor is used for the multiplication is solely dependent on the risk class assigned. For the years 1990 through about June 28, 1993, ⁶ one set of risk factors was used in petitioner's swap business, and thereafter, a new set of factors was used as follows:

 Risk Class    Old Risk Factors    New Risk Factors

[91] Petitioner does not track its swap defaults, according to its responses to formal discovery. However, in the early 1990's, petitioner did attempt to reconstruct its history of defaults from the inception of its swap business in the 1980s. Based upon a review of that attempt, it appears that only 5 or 6 swaps resulted in actual losses, with one more having a loss of penalty interest and another having the existence of the swap disputed. The exact time span which these figures cover is not known. In addition, these figures cannot be recast in percentage terms because it is not known how many swaps petitioner entered into during the period covered by its history.

[92] Petitioner's risk factors were originally developed for purposes of determining reserve adequacy requirements and not for the purpose of measuring the discount in the marketplace for credit risk. According to at least one former employee of petitioner from the Credit Policy Group, Mr. Mark Garrigus, petitioner developed conservative reserve policies. The risk factors may overstate the identified loss possible for non-real estate assets (risk factors for this category were used for swaps). According to Mr. Garrigus, the risk factors are based upon historical cyclical data, with a determination based upon a "going forward" business judgment.

4. REPORTING PERIOD

[93] While petitioner is a calendar year taxpayer, it did not claim credit risk carve-outs based on the fact that the start date of the swaps was within the calendar year. For example, on its 1993 return, petitioner claimed credit risk carve-outs for swaps entered into from December 2, 1992 through November 29, 1993. Thus, petitioner shifted its reporting period. No information on this aspect of petitioner's methodology is available for the years 1990 through 1992. Claiming the carve-outs in the wrong year is not allowable, even if the carve-outs themselves, were allowable.

5. DISCOUNT AMOUNT

[94] Petitioner reconstructed ⁸ Devon values for individual interest rate swaps as of December 20, 1993. ⁹ The credit risk carve-out amount for positive Devon values (swaps where petitioner is the net receiver of projected periodic payments) ranges between .01 percent and 138.03 percent of the reconstructed values provided. The credit risk carve-out amount for negative Devon values (swaps where petitioner is the net payor of projected periodic payments) ranges between .01 percent and 78.13 percent of the reconstructed values provided. Thus, for some swaps where petitioner is the expected net receiver, the reduction to income that petitioner is claiming for credit risk (without considering the further reduction for administrative expenses) totally offsets the value of the swap. In addition, the credit risk carve-out is sufficiently in excess of the Devon value to partially offset the values of other swaps. For example, in the swap with Korea Commercial Bank, the carve-out was 138 percent of the reconstructed value, with the reconstructed value being $284 and the credit risk carve-out $392.

[95] For swaps where petitioner expects to be the net payor, or party owing money, not the one exposed to loss of expected net positive exposure, petitioner still claims credit risk carve-outs. An example is the swap with Caterpillar Finance. The credit risk carve- out was 78.13 percent of the reconstructed value, with the reconstructed value being -$896, and the credit carve-out $700. Yet, petitioner has provided no data to respondent to support an assertion that the marketplace viewed petitioner as such a risky credit that petitioner would have to pay $1,596 ($896 + $700) to buyout or transfer the swap. Nor has petitioner shown that the marketplace would, despite the netting agreement, want to discount the swap to such a great extent in the event that Caterpillar ever became the net payor in the future.

6. CREDIT ENHANCEMENTS/COLLATERAL

[96] In reviewing the credit files of swap counterparties, respondent saw references to various forms of collateral, such as loan agreements, pledges, and guarantees. While petitioner assures respondent that the credit risk carve-out amount was adjusted to take into account these credit enhancements, petitioner has yet to explain the procedure for taking credit enhancements into effect, much less verify it was followed. Petitioner failed to keep records of its swap collateral, yet again, petitioner requests it be trusted without corroboration. Petitioner errs. Given respondent has located swaps which are collateralized with no apparent change to the risk class, it is up to petitioner to prove that the swaps for which it claimed credit risk carve-outs were not collateralized.

B. ADMINISTRATIVE EXPENSE CARVE-OUT

[97] For the years at issue, petitioner calculated administrative expenses carve-out on a portfolio basis; the carve-out is never amortized back into income. The administrative carve-out amount is petitioner's estimate of the aggregate of: (a) the direct and indirect future costs to manage its swaps to maturity; and (b) the allocated future costs of the entire bank that petitioner claims would be necessary to support its swap line of business, assuming no new swaps are transacted.

[98] The administrative expense carve-outs, in effect, reduce the year-end Devon values of outstanding swaps at year-end by the total projected future expense of $5,253,337, $3,318,920, $3,843,770 and $4,832,469 in each year 1990 through 1993, respectively. As petitioner is not in the first year of relying on Devon values, it is the difference in Devon values from year-to-year that impacts swap revaluation income, i.e., it is the difference in the mark-to-market income at each year-end which is posted to the tax returns. Thus, petitioner computes its administrative expenses separately from its Devon values, booking just the quarterly increases and decreases in administrative expenses. For tax purposes petitioner reports on its tax returns what it claims are the annual increases and decreases to these expenses. The amounts claimed were -$982,000, $1,934,417, -$524,850, and -$988,699 in 1990 through 1993, respectively (negative numbers decreasing the values, positive numbers increasing the values).

[99] Petitioner has failed to keep records that would allow respondent to audit the amounts claimed. The budget and management reports, that petitioner states it relied upon for its projection of future administrative expense, have not been produced. Petitioner states that none of the reports were retained. Petitioner has not explained how costs of the other departments of the bank were allocated to the derivatives department. The survey of other bank departments to determine the percentage of these costs allocated to new derivatives as opposed to already-booked derivatives was oral and no contemporaneous documentation of anything but the results is available. Petitioner offered only summary sheets of the amounts and oral statements based on the memories of employees who worked on the administrative expense carve-outs. Respondent did obtain a profit and loss statement for 1997, with historic data for the entire derivatives department for 1992 and 1993; however, the 1992 and 1993 administrative costs for interest rate derivatives cannot be reconciled to the numbers petitioner provided. Petitioner claims it does not have any similar historical data for 1990 and 1991.

[100] The amount of the projected administrative expenses varies form year to year. For 1991, there was, in effect, a net increase to the tax return values by the administrative expense carve-out. There is no indication that the costs of swap administration for petitioner decreased in 1991. There is no evidence that the costs of administering swaps in the marketplace varied in correlation with petitioners numbers. There is no evidence that the cost to administer swaps, if recognized in the market pricing of swaps, will vary based on the maturity of each dealers portfolios, i.e., whether it would take two or ten years to manage a current portfolio to maturity.

[101] For the indirect and direct costs of the swap business and the bank as a whole, it appears from the summary sheets and interviews that respondent has conducted with former employees, that petitioner includes fixed costs such as salary, rent for space, and computer costs, all of which may or may not have been fully paid for, or committed to, in the past. Petitioner considered 100 percent of its computer costs necessary for existing swaps, with none of said costs allocated to new swap business even though new swaps need the system. Petitioner assigned the cost of space for traders believed to be necessary to manage its swaps to maturity for its projections since, as stated by one former employee who was interviewed, they must sit somewhere. Two equivalent traders were assigned for purposes of the calculation. Petitioner may have had only two full-time traders for swaps (one for Canadian dollar and one for U.S. dollar swaps) in the years at issue for both old and new business, until another was added when U.S. dollar swap trading was split between short-term and long-term swaps.

[102] The method of allocating the costs of the bank to the derivatives business and reallocating to the swaps part of the derivatives business is not known. However, respondent has interviewed two former employees of petitioner that stated the allocation of the costs of the bank followed revenues.

III. BUYOUTS

[103] Occasionally petitioner terminates a swap prior to its natural maturity. The termination may be initiated by petitioner or by the counterparty. Terminations can be required if there is an actual deficit such as non-payment of a periodic payment when due. Terminations are also allowable if a counter party has a technical default, e.g., a credit trigger in the International Swap Dealers Association ("ISDA") agreement, due to a credit downgrade. Petitioner does not always exercise its right to require an early termination on a technical deficit. The ISDA agreement sets forth how the termination payment is to be determined.

[104] Most of the swaps at issue in this case are under either the 1987 or 1992 ISDA master agreement. For termination, the 1987 ISDA Master Agreement provides for bilateral and elective close-out netting of swap obligations. The 1992 ISDA master provides for similar netting provisions. Again, the ISDA agreement sets forth how the termination payment is to be determined.

IV. REGULATORY REPORTING

[105] Petitioner is a member of the regulated banking industry. As such, it is audited by and reports to various banking agencies, including the OCC. The petitioner maintains that it used the net values obtained from its adjusted mid-market methodology in its reports to these regulatory agencies. Petitioner's Issues Memorandum, at 14. During the years at issue, the petitioner was audited by various banking agencies, not to determine underreporting of income, but rather to ensure the safety and soundness of the petitioner as a depository institution, and compliance with their regulations and published regulatory guidance.

[106] No evidence of the written policies or procedures of the petitioner regarding its swap income reporting, required to be maintained by these regulatory agencies, was produced in response to respondent's discovery request, other than credit manuals, a traders' manual and limited training materials with respect to trading procedures. It is not clear what, if any, comprehensive written procedures and policies existed in 1990 through 1993.

[107] The petitioner has not shown that it reported either the credit risk carve-outs or administrative expense carve-out on any separate line of the call reports filed with the Office of the Comptroller of Currency ("OCC") or the Bankruptcy-9C reports filed with the Federal Reserve Bank ("FRB"). Other inconsistencies appear to exist with respect to mark-to-market valuation that petitioner undertook for regulatory capital purposes and tax. In its 1993 Year- End Call Report, petitioner stated that the mark-to-market value of its interest rate swaps was $3.2 billion. Yet, petitioner claims the fair market value of its interest rate swaps was $76,973,957 for 1993, and that the mark-to-market income on its 1993 tax return is the difference between this $76.9 million and the value of its interest rate swap portfolio at year-end 1992.

V. UNIFORMITY

[108] Not only does there appear to be a lack of conformity between the petitioner's book and regulatory accounting, there is a lack of conformity among dealers. The Group of Thirty follow up- survey shows that, during the years at issue, dealers in the swap industry took widely divergent approaches to the issue of credit risk, with many dealers not claiming any adjustments. Forty-eight percent (48%) of the dealers who responded valued derivatives using mid-market values without any adjustments whatsoever, and 37% used mid-market values with adjustments (response as of July, 1993). Of the group who used mid-market valuation with adjustments, 51% of those respondents took adjustments for credit risk, and 40% took adjustments for administrative costs. See Petitioner's Expert Report at pages 30-31, Group of Thirty, Derivatives: Practices and Principals, [sic] Follow-up Surveys of Industry Practice (1994), p. 12.

EVIDENTIARY PROBLEMS

Summary Schedules

[110] Respondent has submitted to petitioner a number of schedules which organize and present a variety of data obtained from petitioners records. Despite having some of those schedule since last spring, petitioner has yet to state its position as to the accuracy of any of the schedules. To the extent the parties fail to reach a stipulation regarding respondent's summary schedules, respondent intends to offer a witness who will testify regarding the schedule's preparation and the information reflected on the schedules.

Los Alamos Project

[111] Petitioner has indicated its intention to present evidence regarding the Los Alamos Project. Respondent will object to any evidence regarding the Project based on its lack of relevancy. Petitioner apparently believes that the government's project to develop software for valuing a wide range of exotic financial instruments and derivatives is probative of whether petitioner's own valuation method calculated a fair market value for the swaps at issue. Further, petitioner labors under the misconception that the project was terminated because the project team completely failed in attaining its goal. Respondent has several rebuttal witnesses who will testify to the contrary, if necessary.

Petitioner's Allegation of Arbitrariness

[112] Statements made by petitioner have lead respondent to believe that petitioner will attempt to present evidence that respondent was arbitrary in its determination. Some time ago respondent inquired of petitioner whether this perception was correct but no response has been received. Petitioner appears to be contending that respondent's agents failed to raise specific questions about the actual calculation of the credit risk and administrative adjustments. (Petitioner's Memorandum Pursuant to August 14, 2000 Court Order, at 22). While respondent disputes this characterization, respondent also submits that, even if true, the alleged facts regarding the revenue agent's actions would be irrelevant to the Court's inquiry.

Admissibility of Document -- Bates 1078

[113] Petitioner provided respondent with a document entitled Global Derivative Products Financial Organizational Summary -- 1997. The document is clearly a business record of petitioner and should be admissible as such. It provides a multi-year comparison of the revenues, expenses and profits of the Global Derivatives Unit. Despite repeated requests, petitioner has failed to identify the creator and purpose of this document. Absent a stipulation with respect to this document, respondent anticipates the need to serve an entity subpoena on the petitioner to produce an employee to identify and explain this document.

Petitioner's Responses to Interrogatories

[114] Respondent served interrogatories upon petitioner and received responses. Some responses were signed by an officer of petitioner, some by petitioner's counsel, and some were initially unsigned. For those responses that arrived unsigned, respondent made a subsequent request for signed responses. Respondent relied upon petitioner's responses and seeks petitioner's agreement to stipulate to those responses. Absent an agreed stipulation, respondent will seek admission into evidence of the petitioner's responses as an admission of a party-opponent under FRE 801(d)(2). If necessary, respondent will call an officer of the petitioner to testify with respect to the responses to interrogatories.

CFTC Study and Working Papers

[115] Respondent will request petitioner to stipulate to the October 25, 1993 Commodity Futures Trading Commission Study of Swaps and Off-Exchange Derivatives Trading and the accompanying Working Papers. The study was required by the Conference Report to accompany P.L. 102-546, the Futures Trading Practices Act of 1992. Absent a stipulation, respondent will seek its admission into evidence as a public record or report pursuant to Fed. R. Evid 803(8) or under the hearsay exception provided in Fed. R. Evid. 803(24) as a statement having circumstantial guarantees of trustworthiness equivalent to those found in other exceptions to the hearsay rule.

STUART L. BROWN

² The quarterly calculation actually has a one month lag. For example, swaps originated in December 1993 were not calculated until the following quarter.

³ This is a separate substantiation issue. See Respondent's Pre-Trial Brief, Section III on Substantiation.

⁴ Prior to this time, petitioner used a 95 percent cut-off level.

⁵ Standard swap forms developed by the association then known as the International Swap Dealers Association ("ISDA").

⁶ Respondent does not know if the new factors were used for the second quarter carve-outs.

⁷ In August of 1992, FNBC's Credit Strategy Committee ("CRESCO") approved the change in rates for the risk factors in the above table of new rates. These new rates called for zero credit risk carve-out for risk class 1 swaps.

⁸ Petitioner failed to keep records of the actual values used on the tax return and cannot prove the reconstructed values equal the tax return values.

⁹ The date petitioner uses as its year-end, though it should be December 31st.