Trustmark Corporation and Subsidiaries v. Commissioner., United States Tax Court - Memorandum Decision, T.C. Memo. 1994-184, Docket No. 18372-89., Filed April 28, 1994
Trustmark Corporation and Subsidiaries v. Commissioner.
United States Tax Court - Memorandum Decision
T.C. Memo. 1994-184
Docket No. 18372-89.
Filed April 28, 1994.
Louis G. Fuller, John R. Hutcherson, George P. Hewes III, 248 E. Capital St., P.O. Drawer 119, Jackson, Miss., Paul B. Eason, Philip C. Cook, Timothy J. Peaden, Terence J. Greene, and Sidney O. Smith, Jr., for the petitioner. Albert L. Sandlin, Jr., Gary F. Walker, and Clinton M. Fried, for the respondent.
Memorandum Findings of Fact and Opinion
PARKER, Judge: Respondent determined the following deficiencies in petitioner's Federal income taxes:
In August 1983, Trustmark National Bank entered into an agreement to purchase the assets and liabilities of Canton Exchange Bank. This transaction constituted a taxable asset purchase for Federal income tax purposes.
The issues for decision relate to whether Trustmark National Bank is entitled in 1983 and 1984 to amortization of the core deposit intangible asset it purchased, thus increasing the amount of net operating loss carryback available in taxable years 1973 and 1974. Specifically, the issues are: (1) whether the core deposit intangible has an ascertainable value separate and distinct from goodwill/going concern value and has a limited useful life, the duration of which can be ascertained with reasonable accuracy; (2) if so, whether the values and amortization schedules used by petitioner are reasonable; and (3) whether petitioner may amortize the core deposit intangible on an accelerated basis.
Unless otherwise indicated, all section references are to the Internal Revenue Code in effect for the years at issue, and all Rule references are to the Tax Court Rules of Practice and Procedure.
Findings of Fact
Some of the facts have been stipulated and are so found. The stipulation of facts, the first supplemental stipulation of facts, the second supplemental stipulation of facts, the third supplemental stipulation of facts, the fourth supplemental stipulation of facts, and the exhibits attached thereto are incorporated herein by this reference.
Trustmark Corporation and its subsidiaries, with the exception of Trustmark National Bank, are Mississippi business corporations with their principal places of business in Jackson, Mississippi. Trustmark National Bank is a national banking association with its principal place of business in Jackson, Mississippi. The principal business of petitioner Trustmark Corporation and its subsidiaries is commercial banking. Petitioner Trustmark Corporation and its subsidiaries prepared consolidated Federal income tax returns on an accrual basis for the taxable years in issue.
Trustmark Corporation is a one-bank holding company, incorporated under the Mississippi Business Corporation Act in 1968 under the name of First Capital Corporation. On March 14, 1990, the name of First Capital Corporation was changed to Trustmark Corporation. The principal asset and major source of revenue of Trustmark Corporation is the stock of Trustmark National Bank. This stock accounts for approximately 95 percent of Trustmark Corporation's operating and net income. Trustmark Corporation owns approximately 99.5 percent of the outstanding stock of Trustmark National Bank. Trustmark Corporation also owns all of the stock of FS Corporation, whose sole activity is developing automobile financing, and of First Building Corporation, which manages and operates an office building in Jackson, Mississippi.
Trustmark Corporation is a bank holding company registered under the Bank Holding Company Act. As such, it is required to file an annual report and provide any other information that the Board of Governors of the Federal Reserve System may require. The Board of Governors of the Federal Reserve System may examine Trustmark Corporation and its subsidiaries. Additionally, every bank holding company must obtain prior approval of the Board of Governors of the Federal Reserve System before it may acquire substantially all of the assets of any bank, or ownership or control or any voting shares of any bank, if, after the acquisition, it would own or control, directly or indirectly, more than 5 percent of the voting shares of the bank.
Trustmark National Bank was originally chartered in 1889 as a State bank in Mississippi under the name of Jackson Bank. Jackson Bank merged with State National Bank, which had received its State charter in 1907 and its national charter in 1913. Jackson Bank then reorganized as a national bank and changed its name to Jackson State National Bank. In 1949, Jackson State National Bank merged with Capital National Bank, forming the First National Bank of Jackson. In September of 1985, the First National Bank of Jackson changed its name to Trustmark National Bank.
Trustmark National Bank is the second largest bank in Mississippi. As of December 31, 1982, Trustmark National Bank had total assets of approximately $1,437,861,000 and deposits of approximately $1,146,038,000. Trustmark National Bank provides a full range of consumer banking services, including checking accounts, negotiable order of withdrawal (NOW) accounts, savings programs, personal and business loans, money transfers, safe deposit facilities, as well as other interest-bearing time accounts. Trustmark National Bank also offers its customers MasterCharge and VISA credit cards.
Trustmark National Bank is a full-service commercial bank. Its corporate banking department provides a variety of financial services to businesses, including commercial and industrial loans, equipment financing, mortgage financing, term loans, small business loans, and letters of credit. Trustmark National Bank provides an outlet for data processing services, including automated payroll services, check reconciliations, and correspondent banking services.
Trustmark National Bank's trust department provides personal trust and agency services, custody of securities, investment management services, self-employed retirement plans, and corporate pension and profit-sharing trusts. Trustmark National Bank competes with other financial institutions, including national and state banking and savings and loan associations, personal loan companies, and credit unions in its service area for all types of deposits, loans, trust, and other financial services.
At the time of the acquisition in August 1983, Trustmark National Bank had 48 offices in the following cities and towns in Mississippi: Jackson, Clinton, Pearl, Brandon, Ridgeland, Columbia, Gloster, Liberty, Greenville, Leland, Greenwood, Hattiesburg, Petal, and McComb. The branch offices of Trustmark National Bank were located within a 100-mile radius of its home office in Jackson, Mississippi, the State capital of Mississippi and the county seat of Hinds County. As of December 31, 1982, Trustmark National Bank had approximately 1,147 employees and approximately 3,621 shareholders.
Hereinafter the term petitioner will refer to the First National Bank of Jackson as it was known at the time of the acquisition in 1983, or Trustmark National Bank as it is now known.
Northeast Jackson was one of the city's major growth areas in 1982 and 1983. Towns in south Madison County, which is just north of Hinds County, and particularly the towns of Ridgeland and Madison immediately to the north of Jackson's city limits, were growing rapidly in the early 1980s. They had become bedroom communities for many people who worked in Jackson, including many of petitioner's existing customers.
Petitioner had already opened a branch in Ridgeland and was aggressively pursuing new Madison County deposits. Petitioner viewed the acquisition of an existing bank with a high level of customer deposits as an expeditious means of achieving its goals. The Canton Exchange Bank in Madison County was such a bank. It was highly regarded and had good, stable deposits. The Canton Exchange Bank had good customer relationships throughout the whole county. For these reasons, petitioner was interested in that bank and wanted to acquire it.
On August 30, 1983, petitioner entered into an agreement with the Canton Exchange Bank (hereinafter the Acquired Bank) to purchase the assets and assume the liabilities of the Acquired Bank. Petitioner had actively considered the purchase of the Acquired Bank for at least 3 months before August 30th. At the time petitioner made its offer, the Acquired Bank was also considering an offer from Deposit Guaranty National Bank in Jackson, Mississippi.
Frank Day (Day) was the chairman of the board and chief executive officer of petitioner at the time of the acquisition. As such, he exercised the greatest authority in determining the offering price, subject only to the approval of petitioner's board of directors. Clyde Edwards (Edwards) was an executive vice president of petitioner at the time of the acquisition, and he was particularly knowledgable about the Madison County banking community. Edwards advised Day as to the offering price that was likely to be acceptable to the Acquired Bank.
John Tullos (Tullos) was an executive vice president and cashier of petitioner at the time of the acquisition. His primary responsibility in the acquisition of the Acquired Bank was the preparation of the application to the Office of the Comptroller of the Currency (OCC).
David R. Carter (Carter) is currently petitioner's chief financial officer and secretary/treasurer of Trustmark Corporation. In 1983 he was a vice president and comptroller of petitioner. Carter, along with others, determined the price to pay for the Acquired Bank. He used a computerized model to determine petitioner's earnings potential at various purchase prices. He was ultimately responsible for allocating or causing to be allocated to the core deposit intangible asset the amount claimed on petitioner's tax returns.
In June or July 1983, Tullos participated in determining the bid price that petitioner would offer to the Acquired Bank. Meetings were held attended by Tullos, Day, Edwards, Carter, and others. In Tullos' opinion, two things influenced the price to be offered: First, the desire to acquire this bank for an amount at which petitioner could make earnings and a profit from the deal and, secondly, the desire to outbid petitioner's closest competitor, if possible.
In considering the purchase of a bank, the amount of core deposits is a material consideration in the projected earnings analysis. 1 The account balances drive the earnings potential for the bank, especially at a bank like the Acquired Bank which had a relatively small number of accounts.
1 The core deposit intangible asset is also referred to as the deposit base intangible. Similarly the terms core deposit, deposit core, and deposit base are used interchangeably.
In making its decision to purchase the assets of the Acquired Bank, petitioner did not endeavor to assign a specific dollar value or specific weight to any of the following factors: Entry into a new market, location of acquired bank, favorable leases, trade name, workforce in place, long-term supply or service contracts, or goodwill.
When advising petitioner, Tullos looked at the assets of the Acquired Bank as well as relationships with the depositors "because if you acquire a bank and you don't have any deposits left in it, * * * you are in trouble." In Tullos' opinion, the key factors in the acquisition were the core deposits, the stable base of deposits that the Acquired Bank had, and the coverage that the bank had throughout Madison County. Petitioner was looking to acquire a solid deposit base. Thus, acquiring the deposits was more important to petitioner than the Acquired Bank's name. There was no other factor in petitioner's acquisition of the Acquired Bank that was as important to petitioner as the core deposits.
Among the factors influencing petitioner's decision to purchase the assets of the Acquired Bank was petitioner's interest in expanding its business in Madison County. Petitioner desired to become the dominant bank in Madison County, and acquiring the largest existing bank in that area was the most expeditious way of accomplishing that goal.
The Acquired Bank was a State bank organized in 1880 and incorporated in Mississippi in 1913. At the time of the acquisition, it had the largest share of bank deposits in Madison County, its primary market area. The Acquired Bank offered a full range of consumer and commercial banking services but did not offer trust services.
At the time of the acquisition, the Acquired Bank was a going concern and possessed a good local reputation and a large share of the bank deposits in Madison County. The Acquired Bank was managed by F.E. Allen, who had been an officer of the Acquired Bank since 1930 and a director since 1932.
At the time of the acquisition, the Acquired Bank had a main office and four branch offices. Its main office was located in Canton, Mississippi, the county seat of Madison County, which is just north of Hinds County. The main office of the Acquired Bank was located approximately 25 miles north of petitioner's home office. In addition to this office in Canton, the Acquired Bank had an office in a shopping center and a drive-in facility in Canton. It also had branch offices in Madison, about 11 miles south of its main office, and in Ridgeland, about 15 miles south of its main office.
At the time of the acquisition, petitioner had a Madison County branch in Ridgeland that was approximately 1.6 miles from the Ridgeland branch of the Acquired Bank. As of July 23, 1983, petitioner had approximately $12.1 million of deposits in accounts with Madison County addresses, an amount equal to 18.7 percent of the Acquired Bank's deposits in accounts with Madison County addresses.
Core deposits are both stable and relatively insensitive to interest rate changes. Petitioner's earnings potential following the acquisition depended upon the ratio of core deposits to so-called "hot money" and other deposits sensitive to interest rate changes. The higher the ratio, the higher the earnings potential. Petitioner was first contacted about making the 1983 acquisition of the Acquired Bank in 1982 or early 1983, when Flora Rimmer (Rimmer), executive vice president of the Acquired Bank, placed a telephone call to Edwards, petitioner's executive vice president, advising Edwards that the Acquired Bank had been approached recently by other banks expressing an interest in acquiring it. Rimmer advised Edwards that if petitioner had any interest in making an offer, it should do so in the near future. On May 24, 1983, petitioner submitted an offer to purchase the assets, franchise, and goodwill of the Acquired Bank for $8,500,000 in cash. The Acquired Bank rejected this offer.
On June 22, 1983, Deposit Guaranty National Bank (Deposit Guaranty) of Jackson, Mississippi, offered the Acquired Bank cash in an amount equal to 150 percent of the audited book value of the Acquired Bank. Six days later, petitioner offered cash in the amount equivalent to $55 per share for all of the assets and liabilities of the Acquired Bank. The total amount of this offer was approximately $9,800,000. The Acquired Bank rejected both offers.
On June 30, 1983, the Acquired Bank requested petitioner and Deposit Guaranty to resubmit their bids in terms of both the total cash offered and the amount per share. On July 7, 1983, petitioner offered to purchase all the assets, including properties, goodwill, and business and the liabilities of the Acquired Bank for $10,700,910 cash, or $60.05 per share. Deposit Guaranty submitted a bid of $10,250,000, or $57.52 per share. On July 7, 1983, the board of directors of the Acquired Bank accepted petitioner's offer. On July 12, 1983, petitioner's board of directors approved the purchase of the assets of the Acquired Bank and the assumption of its liabilities.
In July of 1983, petitioner engaged Touche Ross & Co. to perform an analysis of the useful life and value of the deposit base of the Acquired Bank. On August 19, 1983, Touche Ross & Co. furnished to petitioner a report on its analysis of the deposit base of the Acquired Bank as of June 30, 1983 (the Touche Ross Report). In reliance on the Touche Ross Report, petitioner allocated to the deposit base of the Acquired Bank an amount of purchase price equal to the value determined in that report. For financial, regulatory, and income tax accounting purposes, petitioner recorded on its books a core deposit intangible asset in an amount equal to the value determined in that report. Petitioner has been depreciating that amount using the amortization schedule set forth in that report. Petitioner submitted the Touche Ross Report to the OCC as part of its application for OCC approval to purchase the assets of the Acquired Bank. 2
2 Petitioner does not rely on the Touche Ross Report as an expert report in this case, and therefore that document is not part of the evidentiary record of the case.
On August 30, 1983, the Acquired Bank adopted a plan of complete liquidation pursuant to section 337. On August 31, 1983, petitioner submitted an application to the OCC for approval to purchase the assets of the Acquired Bank. The acquisition of the Acquired Bank by petitioner did not require approval by any State regulatory agency. On November 14, 1983, the OCC approved petitioner's application. On December 15, 1983, the OCC officially certified its approval of petitioner's purchase of assets and assumption of liabilities of the Acquired Bank. It further certified its approval for petitioner to establish branch offices at the banking locations of the Acquired Bank.
On December 14, 1983, pursuant to the Agreement to Purchase Assets and to Assume Liabilities, dated August 30, 1983, petitioner acquired all of the assets and assumed all of the liabilities of the Acquired Bank. As of December 14, 1983, 39.54 percent of the total assets shown on the statement of condition of the Acquired Bank consisted of loans; 25.98 percent consisted of investment securities; 25.5 percent consisted of Federal funds; 5.81 percent consisted of cash; 2.38 percent consisted of accrued interest receivable; 0.71 percent consisted of premises and equipment; 0.07 percent consisted of prepaid expenses; and 0.01 percent consisted of other real estate. As of the date of acquisition, the Acquired Bank charged a maximum rate of interest of 13.5 percent on commercial loans, a maximum add-on rate of interest of 12 percent on installment loans of $2,500 or less, and a maximum rate of interest of 13.5 percent on installment loans over $2,500. Terms of particular loans may have varied based on ordinary credit considerations.
On December 15, 1983, the former main office and branch offices of the Acquired Bank opened for business as branch offices of petitioner. After the acquisition and until September 1985, petitioner operated the former offices of the Acquired Bank under the name "Canton Exchange Bank, branch bank of the First National Bank of Jackson". In September 1985, petitioner changed the name of the First National Bank of Jackson and all of its branches to Trustmark National Bank.
None of the Acquired Bank's employees or management was discharged or resigned as a result of the acquisition. All telephone numbers of the Acquired Bank remained the same after acquisition by petitioner.
Immediately following petitioner's acquisition of the Acquired Bank, a form letter was prepared and mailed to all of the Acquired Bank's customers, including depositors, advising them of the acquisition. Based upon petitioner's written marketing plan developed at the time of the acquisition, petitioner believes "statement stuffers" announcing the acquisition were included in the monthly bank statements of the Acquired Bank's checking account customers. No copy of such statement stuffers can presently be located.
At the time of the acquisition, petitioner placed advertisements announcing the acquisition in The Madison County Herald, a newspaper published in Madison County; The Northside Sun, a weekly neighborhood newspaper serving north Jackson and south Madison County; and in The Clarion Ledger, a newspaper of statewide circulation published in Jackson. Petitioner prepared one radio spot that ran on the local radio station, WMGO, Canton, Mississippi, for 2 weeks after the acquisition. Petitioner also placed advertisements announcing the acquisition on two billboards located in Canton and Ridgeland. Petitioner, however, did not advertise on television. Notice of the filing of the OCC application was published in the legal advertising section of The Clarion Ledger on September 12, September 26, and October 11, 1983. The closing of the transaction was announced in news articles appearing in the December 14, 1983, edition of The Clarion Ledger and in the December 15, 1983, edition of The Madison County Herald. The Acquired Bank's acceptance of petitioner's purchase offer had also been reported in The Clarion Ledger on July 10, 1983.
Petitioner obtained the Acquired Bank in a transaction that constituted a taxable asset purchase for Federal income tax purposes. Petitioner thus obtained a basis in the assets of the Acquired Bank equal to the cost thereof. Upon its purchase of the assets of the Acquired Bank, petitioner allocated $4,744,366 of the purchase price to the core deposit intangible asset and $305,994.59 to goodwill and going concern value.
Petitioner allocated a portion of the purchase price of the Acquired Bank to the acquired real property in an amount equal to the appraised fair market value of the real property as established by an appraisal performed by Wortman & Mann, Inc. Petitioner allocated a portion of the purchase price of the Acquired Bank to the acquired loans in an amount equal to the appraised fair market value of the loans on the acquisition date. In recording the loan portfolio, petitioner made no adjustment to the loans of the Acquired Bank, except for an increase in loan loss reserves due to credit risk, because petitioner determined that the acquired loans bore market rates of interest.
Petitioner allocated a portion of the purchase price of the Acquired Bank to the acquired securities in an amount equal to the appraised fair market value of the securities on the acquisition date. Values for local and other securities not listed in the Wall Street Journal were determined by informal consultation with the Jackson office of Scharff and Jones.
The total purchase price paid by petitioner for the assets of the Acquired Bank was $10,700,910 plus the assumption of all of the Acquired Bank's recorded liabilities. No separate allocation was made in the acquisition agreement as to the amount of the purchase price that was attributable to the core deposit intangible asset or to any other asset. The purchase price of the assets of the Acquired Bank was the result of arm's-length negotiations.
Petitioner allocated the purchase price of the Acquired Bank in the following manner:
Petitioner allocated the purchase price among the investment portfolio, loans, fixed assets, and the core deposit intangible (or core deposit base), and allocated the remaining amount to goodwill and going concern value. 3 The parties have stipulated that the foregoing purchase price allocations made by petitioner are not in dispute, except for the allocations made by petitioner to core deposit base and goodwill/going concern value.
3 The parties' stipulation paragraphs 83 and 87 seem to contain inconsistent figures as to the amount allocated by petitioner to goodwill and going concern value ($305,994.59 vs. $485,057).
The OCC defines core deposits as "The deposit base of demand and savings accounts which, while usually not legally restricted, are generally based on stable customer relationships so that the bank can expect the use of these funds for an extensive period of time, often many years." The OCC defines the core deposit intangible as "the present value of the future net income stream associated with a bank's core deposits."
Members of the banking industry contrast core deposits with what is known as "hot money". "Hot money" is sensitive to even relatively small fluctuations in interest rates and is likely to be moved among various types of investments in order to take maximum advantage of higher interest rates. In contrast, core deposits are the most stable deposits of a bank or financial institution. The holders of these accounts do not shift money among accounts based on changes in the interest rates.
Petitioner contends that the core deposits in this case consist of the Acquired Bank's demand deposit accounts (DDA's) ≈ including regular checking, money market, and NOW accounts ≈ and savings accounts (SDA's) ≈ including Golden, Silver, and regular savings accounts. 4 Core deposits are a relatively stable, low-cost source of funds and are one of the largest income-producing resources for a bank. In the absence of core deposits, a bank would have to obtain funds at a higher cost, thereby reducing its income. Thus, lower-cost funds are more valuable to a bank than higher-cost funds.
4 Petitioner also contends that core deposits may also include insured certificates of deposit for amounts less than $100,000. Since petitioner assigned no value to such certificates of the Acquired Bank, however, they are not at issue in this case. Petitioner's experts also excluded the five Super NOW accounts from their consideration. As will be discussed in the Opinion below, the Court concludes that petitioner has failed to establish that the Acquired Bank's money market deposit accounts and NOW accounts were insensitive to interest rate changes. Hence, the money market deposit accounts and NOW accounts will be excluded from the core deposits in this case.
A depositor's account with a bank represents a liability of the bank. Accordingly, petitioner recorded deposits of the Acquired Bank as liabilities on its statement of condition. Petitioner also recorded the portion of the purchase price of the Acquired Bank that it allocated to the deposit base as an asset on its statement of condition. Subject to contractual and regulatory limitations, petitioner must pay depositors as they withdraw their funds. Prior to the time it submitted its offer to purchase the assets of the Acquired Bank, petitioner did not conduct a formal analysis or statistical study of the deposit accounts or any other asset of the Acquired Bank. Petitioner has not conducted any internal studies to determine the existence of, or the specific costs associated with, any favorable borrowing rates, precommitments (formal or informal) to supply loans, or cross-subsidies to core depositors provided in the form of below-cost check clearing, credit cards, telephone transfers, automatic teller machines, direct deposit services, financial advice, notary services, safety-deposit boxes, traveler's checks, cashier's checks, and business contacts.
The costs of maintaining core deposit accounts include a portion of the costs of maintaining the physical facilities of the bank, a portion of the marketing and advertising expenses of the bank, and the costs of providing services to maintain deposit accounts. Core deposit accounts and depositor relationships also afford the bank the opportunity to generate additional income from non-core deposit accounts and other banking services.
Beginning in the late 1980s and up to the present time, banks have bid for and acquired the deposit liabilities of failed thrift institutions through the Resolution Trust Corporation (RTC). In these transactions, the bank pays money to assume the deposit liabilities. In such circumstances, the bank may or may not purchase any other assets or properties of the thrift institution.
Deregulation of the banking industry was not a discrete event but had been gradually emerging since the mid-1970s. Deregulation may have contributed to the runoff of depositors, but it did not necessarily affect the outcome of statistical analyses performed for 1983. The major changes in interest-rate deregulation had taken place between 1979 and 1981. By 1982 the markets had settled. Deregulation was, however, one of a number of factors that created some uncertainty in predicting the behavior of accounts. In this case, it was better to look at data from only 1983 because conditions in the banking industry in the immediately preceding year were atypical and also were unlikely to recur.
At the time of the acquisition in this case, both petitioner's deposit accounts and the accounts of the Acquired Bank were insured by the Federal Deposit Insurance Corporation. Both petitioner and the Acquired Bank were required to pay premiums to obtain such Federal deposit insurance. Federal deposit insurance permitted both petitioner and the Acquired Bank to attract deposit funds at a lower cost than uninsured funds.
Each year there were three or four separate verifications of the account information that petitioner's data processing department stored on its computer tapes. These verifications of account information were done by computer experts in petitioner's internal audit department, by the OCC, and by petitioner's external auditors. In addition, petitioner had a quality assurance department that assessed any errors and determined how such errors were made.
The Acquired Bank's computer system and petitioner's computer system used different account status codes. On petitioner's computer system, a status code of "2" signifies an open account, "4" signifies a closed account, and "5" signifies a dormant account. On the Acquired Bank's computer system, a status code of "C" signified a closed account and "D" signified a dormant account. Under petitioner's current operating procedures, an account is assigned a status code of "4" (closed) in petitioner's computer system as soon as it has a zero balance, even if the status code must later be changed because of a change in the account balance. The Acquired Bank had a different computer system, and this procedure did not apply to accounts at the Acquired Bank at the time of the acquisition. Accounts with zero balances can sometimes be valuable to the bank if the account balance fluctuates depending upon, for instance, when the account holder's paycheck is deposited.
Petitioner created a set of diskettes containing files extracted from the original raw data tapes produced by its computer system and by the Acquired Bank's computer system (the extract files). In creating the extract files, petitioner converted the old account status codes used by the Acquired Bank's computer system to the new account status codes used by petitioner's computer system. Therefore, in the extract files, an account has a status code of "4" ( i.e., closed) if it had a status code of either "C" or "4" on the original data tapes, a status code of "5" ( i.e., dormant) if it had a status code of either "D" or "5" on the original data tapes, and a status code of "8" ( i.e., closed and waiting to be purged) if it had a status code of "8" on petitioner's original data tapes. Otherwise, the account has a status code of "2" ( i.e., open).
In creating the extract files from the original computer tapes, petitioner relied exclusively on the account status code to determine the status of an account. An account, even an account with a current zero balance, was not assigned a status code of "4" ( i.e., closed) unless the record of that account on the original computer tapes had a status code of "C" in the case of data from the Acquired Bank's computer system or "4" in the case of data from petitioner's computer system.
The deposit accounts acquired by petitioner from the Acquired Bank on December 14, 1983, included the following demand deposit and savings accounts:
The acquired demand deposit accounts (DDA's) included 277 accounts which had a current zero balance and which did not have a status code of either "C" or "D" as shown in the Acquired Bank's computer records at the time of the acquisition. The acquired savings accounts (SDA's) did not include any accounts which had a current zero balance and which did not have a status code of either "C" or "D" as shown in the Acquired Bank's computer records at the time of the acquisition.
Of the 5,461 acquired SDA's, the number of accounts that remained open at the following dates, together with the aggregate current balance of those surviving accounts, are shown in the following table:
Of the 9,064 acquired DDA's, the number of accounts that remained open at the following dates, together with the aggregate current balance of those surviving accounts, are shown in the following table:
A. Thomas E. Doerfler, Ph.D.
Dr. Thomas E. Doerfler (Dr. Doerfler) was engaged by petitioner to determine the useful life of the core deposit intangible asset. Dr. Doerfler is a statistician with Arthur D. Little, a management consulting firm in Cambridge, Massachusetts. He has performed lifing analyses in other core deposit cases before this Court, including Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. 463 (1988), affd. without published opinion 900 F.2d 266 (11th Cir. 1990), affd. per curiam [91-1 USTC ╤ 50,043], 919 F.2d 1492 (11th Cir. 1990), and Peoples Bancorporation v. Commissioner [Dec. 48,226(M)], T.C. Memo. 1992-285. 5
5 "Lifing analysis" is a term used to describe the determination of the remaining useful life of an asset.
1. Overview of the Lifing Analysis
In performing the lifing analysis, Dr. Doerfler considered three factors: (1) the number of open accounts at or near the date of acquisition; (2) the age of the accounts at that time; and (3) the rate at which the accounts were expected to terminate as a function of their current age. Dr. Doerfler began his analysis with the assumption that the existing accounts will continue to be closed out at rates consistent with the rates that have been observed in the past. He further assumed that the data used for the analysis was not inherently biased. Bias in data exists when the data that the statistician uses does not correspond to the goals of the analysis.
In order to determine the remaining useful life of the core deposit intangible asset, Dr. Doerfler observed account closures in 1983, and from that determined the probability that an account of a certain age will close in a given year. Dr. Doerfler did not, however, have data for DDA's that both opened and closed in 1983, which would have given him a more precise estimate of the number of accounts closed.
Next, Dr. Doerfler used the results obtained to "form an estimate of the cumulative probability of being closed or, equivalently, the percent remaining open to any age." The third step was to estimate a survival curve ≈ i.e., to characterize the observed account closures with an appropriate probability model.
In the final step, Dr. Doerfler estimated the remaining life of the existing accounts using a series of calculations and formulae. Dr. Doerfler referenced a treatise by Dr. Wayne Nelson as the source of these calculations. Nowhere in his report, however, did Dr. Doerfler explain his calculations. At trial Dr. Doerfler explained the theory behind his methodology. He did not, however, explain the specific formulae he used. 6
6 We think that an expert report should be self-contained and self-explanatory within its four corners and set forth in clear language all assumptions, formulae, computations, and any other bases for the conclusions reached therein. As will be discussed below, Dr. Doerfler's report is inadequate in this respect.
Dr. William Q. Meeker (Dr. Meeker), a statistics professor at Iowa State University, considered Dr. Doerfler's methodology to be reasonable. Dr. Meeker's "major complaint with the report was that it was incomplete in explaining some of the things that were done." As for Dr. Doerfler's use of this methodology, Dr. Meeker testified that it was not necessary that the units begin the experiment at the same time and be subjected to the same conditions over a relatively lengthy observation period. These methods are frequently used in medical studies.
Dr. Meeker concluded that the real statistical issue (and the critical assumption in Dr. Doerfler's report) was whether 1983 was a representative year.
2. Obtaining the Data
Dr. Doerfler's report stated that all of the deposit account data was provided by petitioner. Dr. Doerfler, however, did not receive the data directly from petitioner. L. Glenn Harper (Harper), another of petitioner's expert witnesses, acted as liaison between Dr. Doerfler and petitioner. Harper actually obtained the data (the extract files) from petitioner. Harper's role as an expert was to report on the actual attrition of the acquired accounts between the acquisition date and December 31, 1989; his role in data collection was essentially that of a conduit.
Harper received the data on magnetic tapes from Leo McLeod, petitioner's vice president in charge of data processing. Harper used the account status codes, as described earlier, to determine if each account was open, closed, or dormant. Harper also determined the accuracy of the data, including tracking accounts affected by the "false closure effect". He described the false closure effect as follows: "if a customer closed their account but kept their money in the bank, that [sic] we would be treating that as a closed account when, in fact, they kept their money in the bank and it should not have been treated as a closed account; so it is falsely accounted for as a closure." Harper did not change any of the data before sending it to Dr. Doerfler. The same data was made available to respondent's experts. 7
7 At trial the Court expressed concern about the fact that Dr. Doerfler obtained the data from Harper rather than directly from petitioner. However, the Court is now satisfied that there was no impropriety in Harper's handling of the data and that he was nothing more than a conduit for transmittal of the data. Respondent's continuing complaints on brief about Harper's role in the data collection are baseless.
3. Deficiencies in the Data
The magnetic tape data of the Acquired Bank went back to December 31, 1982. Some of the accounts did not have proper opening dates. An employee of petitioner, Lewis Greer, obtained the signature cards for Harper and, after looking at approximately 50 cards out of a total of 1,500, Harper and Greer determined that accounts in this category were "old" accounts ≈ i.e., that they were opened prior to October 1, 1969. Extrapolating from his analysis of the other data, Dr. Doerfler determined an average age for these accounts of 15 1/2 years. Dr. Doerfler found out later that the actual average age for these accounts was 16 1/2 years.
There were some other problems with the data. Of the NOW accounts, 220 out of 482 accounts had unknown opening dates. The amount of money attributable to those accounts included some $7,756 for accounts under $1,500 and over $2 million for accounts over $1,500.
Dr. Doerfler found some discrepancies between the end of the year DDA listings for 1982 and 1983. Dr. Doerfler testified that ≈
There were some discrepancies in which the [DDA] account numbers agreed but social security numbers differed. There was ≈ another example would be for savings accounts: not so much a data deficiency or mistake, but something of major concern was the ≈ on the 1983 date of acquisition savings account file there was a very large number of accounts that had closing dates of sometime during 1982.
I struggled with that, whether or not to include those accounts, because it would have expanded our window of observation beyond 1983; it would have extended sometime into 1982, preferably all of '82, but as I was able to justify for myself that all ≈ I did not have ≈ first of all, I did not have all 1982 closures; therefore the data would have been inherently biased. So I avoided that.
And I was also concerned that I did not even have all 1982 closures within, say, a contiguous two- or three- or four-months period.
So in order to avoid bias I simply used 1983 closures, because I was under the impression that I had a complete data set in that regard.
Further, there were approximately 109 out of some 9,000 DDA's with the same account number but different opening dates and approximately 100 DDA's with obscure account type codes. Dr. Doerfler made an educated guess of the opening dates for four of the DDA's, out of a total of 11,113 accounts. These educated guesses constituted routine data "cleansing" or data "massaging".
In the checking account portion of the DDA's, there were 1,304 accounts with unknown opening dates. Dr. Doerfler estimated an age for these accounts by extrapolating from the other data. Dr. Meeker agreed generally that Dr. Doerfler's procedures were reasonable methods of dealing with accounts with unknown opening dates.
While there was some imprecision resulting from the problems with the data, the data was reasonably adequate for statistical analysis, except for the NOW accounts as to which over 45 percent had unknown account opening dates. The age of an account was, according to Dr. Doerfler, "a very critical" factor in his analysis. Dr. Doerfler testified that it was "absolutely imperative" that he "define the age of the account at the time of acquisition." As a statistician, he modeled only the age of the accounts, i.e. the probability of closing as a function of age.
4. Stratification of the Data
In determining the useful life of the deposit base, Dr. Doerfler stratified the data by account type. He further stratified the DDA data by account size into accounts with balances less than or equal to $1,500 and accounts with balances greater than $1,500. Dr. Doerfler used the amount of money in the accounts as of December 14, 1983, the date petitioner actually acquired the assets and assumed the liabilities of the Acquired Bank. He used year-end balances rather than average daily balances in the subsequent years.
Dr. Doerfler testified that it is proper to stratify data if very different decay (or runoff) rates are expected for different groups. The decision to stratify accounts at the $1,500 level was made by petitioner, with the assistance of Robert J. Curtis (Curtis), petitioner's valuation expert. It was Curtis' experience that there is a difference in account behavior (different economics) above and below this point and that smaller accounts tend to run off ( i.e., close) faster. Curtis did not stratify at petitioner's break-even point because, as he explained, he was "interested in the behavior of the accounts, not the behavior of the bank. The bank's break-even point is not even known to the customers. The customer's behavior is what I am trying to determine." Dr. Doerfler thought that the stratification level was sufficient given that the objective was to determine useful life with "reasonable accuracy". Dr. Meeker agreed that this was a sufficient stratification level.
Dr. Doerfler did not have enough information to stratify the accounts by bank branches. The data was not stratified according to business or personal accounts.
5. Results of the Lifing Analysis
Dr. Doerfler's conclusions as to the remaining lives of the acquired accounts are summarized as follows:
6. The Goodness-of-fit Test
Dr. Doerfler characterized the observed account closures with an appropriate probability model. He found that a Weibull survival curve characterized the observed data.
The scale of the two axes has been chosen in such a way that the data points will plot as a reasonably straight line if the observed times to closure follow a specific well-defined probability law. Different graph paper ( i.e., a different scale) exists for different probability distributions which tend to occur frequently in lifing analysis applications. The paper selected * * * is Weibull probability paper which, as it turns out, appears to be a reasonable characterization of the process, since the data points exhibit linearity.
Dr. Doerfler then performed a goodness-of-fit test to assess the adequacy of the chosen distribution. The goodness-of-fit test answers the question: "Does this line explain or adequately characterize the dot points?" Dr. Doerfler applied what is known as the Anderson-Darling test to determine the goodness-of-fit. In determining goodness-of-fit, Dr. Doerfler deviated from the textbook applications of that test. He did not have actual observations and instead used "simulated observations" that were computer-generated.
B. Stephen P. Stuk, Ph.D.
Respondent presented the report and testimony of Dr. Stephen P. Stuk (Dr. Stuk), an expert in applied statistics, to rebut Dr. Doerfler's testimony. Dr. Stuk perceived the following problems in Dr. Doerfler's methodology: (1) Dr. Doerfler showed no relationship between dollars leaving the bank and accounts closing; (2) Dr. Doerfler assumed that account behavior would remain constant for 20 years into the future and there was no proof that it was constant in the past; (3) Dr. Doerfler did not state the assumptions that he used to develop his model; (4) Dr. Doerfler did not perform any analysis of his assumptions; (5) Dr. Doerfler did not evaluate the accuracy of the model; and (6) Dr. Doerfler did not define a confidence interval 8 for a spread of possible future values relative to his prediction. 9
8 A "confidence interval" is a statistical tool used to measure the accuracy of an estimate.
9 Dr. Meeker did not think it necessary to determine a prediction interval in a case like this "where the ultimate goal is to have a best guess" of what the life is.
Dr. Stuk also found that Dr. Doerfler failed to include the equations he used and omitted several interim and final calculations. As a result, Dr. Stuk was unable to recreate all of the data in Dr. Doerfler's report. Finally, during the trial, Dr. Stuk was able to understand some of the things that Dr. Doerfler had done, although he still could not recreate some of Dr. Doerfler's calculations. Dr. Stuk did not believe that Dr. Doerfler analyzed all the relevant data available to him.
Dr. Stuk had no problems with the goodness-of-fit test used by Dr. Doerfler for the savings accounts, but thought that he should have performed some qualitative tests on the predictions for the other types of accounts. In performing a goodness-of-fit test on the accounts, however, Dr. Stuk found that "on the regular [savings accounts] it fits okay and on several of the others it is reasonable." Dr. Stuk did not think the available data was sufficient to make a reasonable estimate of the life of the accounts and that if Dr. Doerfler accurately predicted the useful life of the accounts, it did not necessarily mean that he used a reasonable model.
Dr. Stuk was troubled by Dr. Doerfler's application of the Anderson-Darling test. Dr. Stuk thought that Dr. Doerfler needed to use data from actual closures, not just probabilities of closure, when applying this test.
Dr. Stuk observed that when there is a large difference between median and average balances of accounts, "there is a tremendous skew in how the dollars are distributed among the accounts." Because he "was only looking at behaviors of accounts", Dr. Stuk did not look at whether the total year-end balances of all of the accounts in the bank are a good indication of the total balance on deposit in those accounts throughout the year.
Dr. Stuk thought that there may have been problems with trying to determine the turnover of accounts around the time of acquisition because depositor behavior could have been different from the usual situation.
Some of Dr. Stuk's criticisms of Dr. Doerfler's methodology were based on his misunderstanding of Dr. Doerfler's report. Dr. Stuk conceded at trial that Dr. Doerfler had established an important relationship between dollars and age of accounts. During cross examination of Dr. Stuk, he (Dr.Stuk) learned for the first time that Dr. Doerfler was applying average dollar amounts to different age accounts. Dr. Stuk was "shocked" to learn that Dr. Doerfler segregated the account balances by different age strata and applied different conditional closure probabilities to each age strata. Dr. Stuk admitted that he was "stunned" to learn at trial that Dr. Doerfler "was allowing for this migration of accounts through the age groups." In other words, Dr. Doerfler was not suggesting that when 10 percent of the accounts closed, then 10 percent of the money left the bank.
Respondent did not request Dr. Stuk to compute the useful life of the deposit base. Dr. Stuk has never been engaged to compute the useful life of any asset.
C. Robert J. Curtis
Petitioner engaged Robert J. Curtis (Curtis) to value the core deposit intangible asset. Curtis is a certified public accountant and a certified management accountant. He is an executive senior manager at Ernst & Young and specializes in valuation. He has valued the core deposit intangible assets of approximately 100 financial institutions. Curtis was the taxpayer's valuation expert in Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. 463 (1988), affd. without published opinion 900 F.2d 266 (11th Cir. 1990) and in Peoples Bancorporation v. Commissioner [Dec. 48,226(M)], T.C. Memo. 1992-285. In the instant case Curtis was the only valuation expert who prepared an expert report and testified as to value; respondent did not attempt to value the core deposit intangible asset.
1. Method of Valuation
In valuing the core deposit intangible asset, Curtis used essentially the same methodology as used in the Citizens & Southern case. As in the Citizens & Southern case, Curtis relied on Dr. Doerfler to determine the useful life of the asset.
Curtis valued the deposit base using the cost savings approach. Using this methodology, the value of the core deposit intangible asset is equal to the present value of the difference between the asset's ongoing cost and the cost of the next most favorable market alternative. As Curtis explained it: "In other words, the acquired deposit accounts represent a source of favorable financing to the purchaser."
Curtis chose insured certificates of deposit (CD's) as the next most favorable alternative source of funds for the following reasons:
1. Insured CD's are comparable in terms of insurability, average balance size, and maturity/remaining life.
2. The interest cost of insured CD's is directly observable in the marketplace over a wide range of maturities (rates are published daily in advertisements for both banks and savings and loans).
3. Insured CD's are the most realistic alternative, because no bank would reasonably attempt to replace deposit funds with long-term debt or equity capital.
4. He was able to conservatively estimate the cost of alternative funds based on the rate structure of retail CD's at the appropriate date.
In projecting the surviving account balances, Curtis multiplied the average balance on the date of acquisition by the number of accounts of a certain age that Dr. Doerfler predicted to have survived. The average balance was adjusted for inflation, which was assumed to be 5 percent. Curtis received the data on average balances from petitioner, not from Dr. Doerfler's report. 10
10 However, Dr. Doerfler also received the information as to the dollar amounts in the accounts from petitioner.
In determining the cost savings, Curtis estimated the cost to petitioner in obtaining alternative funds of an amount equal to the "net investable balances of the acquired deposit accounts". "Net investable balances" refer to the actual amount of the projected account balances that petitioner would be able to invest. In other words, net investable balances are the projected account balances reduced by float and reserve requirements. "Float" refers to the amount of deposits received but not available for investment because the funds have not been collected through the banking system. Reserve requirements refer to that portion of account balances that are required to be maintained on deposit at the Federal Reserve Bank.
In determining the cost of alternative funds, Curtis made certain assumptions and estimations. To determine the interest rate expense for alternative funds, he used rates based on the estimated cost of insured CD's with maturity dates not more than the lowest average remaining life of deposits as determined by Dr. Doerfler. Curtis assumed that the interest rate curve would be flat beyond 5 years, and thus did not use any CD's with maturity dates extending more than 5 years in the future. Curtis used the published interest rates of a major competitor of petitioner which ranged from 10.6 percent to 11 percent. He determined a rate of .22 percent for services and other expenses based on information published in the Federal Reserve Functional Cost Analysis 1983 (FCA). The FCA is a Federal Reserve Board publication which compares the cost of products offered by banks. At trial, Curtis acknowledged that if Dr. Doerfler's determination of the useful life of the deposit base is incorrect, then the interest rates on CD's used by Curtis must be changed. This would change the present value of the deposit base.
To determine the cost of deposits, Curtis took the total of the deposit interest expense and the deposit service expense, reduced by the deposit fee income. Curtis based the deposit interest expense on petitioner's stated interest rates at the time of the acquisition. He based the deposit service expense on FCA data. Examples of deposit fee income are service charges, insufficient fund charges, and collection fees on each type of account.
For each type of account, Curtis first determined the pre-tax cost savings, reduced this amount to reflect income taxes, and then reduced this after-tax cost savings to its present value. Curtis considered that the after-tax cost savings is representative of the fair market value that a purchaser would pay. In determining the after-tax cost savings, Curtis reduced cost savings by the product of the tax rate and the taxable cost savings. He defined taxable cost savings as the pre-tax cost savings reduced by the amortization deduction.
Curtis noted that there was a greater decline in the deposit base in the early years. For each year, he determined the amortization deduction to be the net discounted after-tax cost savings for that year. Because of tax preferences in the banking industry, Curtis used an average effective tax rate of 20 percent rather than the maximum statutory rate of 46 percent. 11 He determined a discount rate of 14.3 percent by analyzing the weighted average cost of debt and equity capital.
11 Use of the 46 percent maximum statutory rate would have resulted in a value of $4,552,582 for the deposit base in Curtis' calculation of value set out in the text below.
Curtis determined the following values for each type of account:
D. L. Glenn Harper
L. Glenn Harper (Harper) is the managing partner of the Harper Company, a consulting firm in Jacksonville, Florida. The Harper Company accumulates data from banks for use in core deposit evaluation studies.
Harper observed the actual attrition of the acquired accounts and the loss of the acquired balances from the acquisition date until December 31, 1989. A comparison of the Harper/Doerfler results shows the following:
1 As a result of problems during conversion of the Acquired Bank's data to Trustmark's system, no data survives in magnetic form for the intervening year.
1 As a resutl of problems during conversion of the Acquired Bank's data to Trustmark's system, no data survives in magnetic form for the intervening year.
E. Gary L. French, Ph.D.
Dr. Gary L. French (Dr. French) is a vice president with Nathan Associates, an economic and management consulting firm in Washington, D.C. Dr. French has previously appeared before this Court in Tele-communications, Inc. v. Commissioner [Dec. 47,194(M)], T.C. Memo. 1991-82, where he determined the value of cable television franchises, and in Ithaca Industries, Inc. v. Commissioner [Dec. 47,536], 97 T.C. 253 (1991), affd. [94-1 USTC ╤ 50,100] 17 F.3d 684 (4th Cir. 1994), where he reviewed and evaluated the taxpayer's reports which valued materials supply contracts and an assembled workforce.
In his report, Dr. French states the following:
It is the opinion of this author that the benefits of assuming the core deposit liabilities as part of the acquisition of Canton are not an intangible asset or property for which the Petitioner should be allowed an amortization deduction. Furthermore, if one were to assume the contrary, the value of the claimed asset estimated by the Petitioner, as well as the estimates of useful life, are incorrect for the purpose of asset amortization. Consequently, the amortization deductions claimed by the Petitioner are incorrect and should be disallowed.
1. Critique of Dr. Doerfler's Report
Dr. French advances three basic criticisms of Dr. Doerfler's report. First, the observation period is too short. Second, Dr. Doerfler did not adequately stratify the accounts. Third, Dr. Doerfler assumed that the decay of the account balances matches the decay of the number of accounts.
2. Critique of Curtis' Report
Dr. French perceives the following to be problems with Curtis' report: (1) In estimating the net investable funds, Curtis looked to short-term funds as a means of financing long-term assets; (2) in estimating net investable funds, Curtis started with Dr. Doerfler's estimated balances, which Dr. French asserts are incorrect; (3) Curtis was essentially looking at going concern value, not the core deposit intangible asset; (4) Curtis assumed that the rate differential, or spread, would remain constant over time; and (5) Curtis' use of an inflation adjustment makes the balances larger than they should be.
F. Carl Schkabla
Carl Schkabla (Schkabla) is employed by the Internal Revenue Service. He prepared a list that compares social security numbers from savings accounts closed in 1984 with the social security numbers in petitioner's 1985 CD file. Schkabla found 151 matches. The purpose of the list was not to show that dollars from the savings accounts were used to purchase CD's, but "that even though an account has closed in 1984 that did not necessarily terminate a relationship with the bank subsequent to 1984".
G. Donald M. Roberts, Ph.D.
Dr. Donald M. Roberts (Dr. Roberts) is a professor of decision and information sciences at the University of Illinois at Urbana/Champaign, Illinois. Dr. Roberts was engaged by respondent to critique the report of Dr. Doerfler.
Dr. Roberts perceived several problems with Dr. Doerfler's report. Dr. Roberts first criticized the statistical techniques used by Dr. Doerfler because those techniques were first developed for analyzing product-life data. Specifically, Dr. Roberts found the following problems in adapting those techniques in determining the useful life of the core deposit intangible asset. First, Dr. Doerfler used observational data rather than data from a controlled experiment and must show that the same techniques apply. Dr. Roberts noted that, when using observational data, there may not be homogeneous units and that, if this is the case, the data must be stratified into homogeneous groups. Second, the accounts did not all begin at the same time. Third, the customers of the Acquired Bank were not a homogeneous group. Fourth, there was a wide variation in the factors affecting the behavior of the customers. Fifth, the observation period was relatively short. All of these situations, however, are addressed in statistical literature and are not necessarily fatal to Dr. Doerfler's analysis.
Dr. Roberts concluded that Dr. Doerfler's adaptation of this methodology would work only if the following three factors were true. First, one must be able to determine the rate of account closings only by knowing the age of the account. Dr. Roberts thought that Dr. Doerfler should have done a study to find the relationship between rates of attrition for dollars and the closure rates of accounts. Dr. Roberts did not, however, attempt to determine this relationship because he thought it was apparent that there was not a close relationship. Second, the rate of closure must not change over time. Dr. Doerfler assumed that accounts of a certain age would close at a certain rate, regardless of the year the account was opened, otherwise known as time homogeneity. Dr. Roberts believed that one could test for time homogeneity by using Dr. Doerfler's model to estimate the number of accounts opened at a certain date in prior years. Third, the data must be free of major flaws. Dr. Roberts found the following problems in the data:
One problem is the problem of the accuracy of the date of account opening. There may be some problems with the date of account closings. There are problems that pertain to the treatment of dormant accounts; for instance, the definition of what is a closed account? ≈ whether a dormant account with zero dollars in it is open?
Dr. Roberts, however, assumed that Dr. Doerfler and Harper did not check the validity of the data. He did not know if there actually were dormant accounts with zero balances.
Dr. Roberts has never performed studies of product failure or performed a survival or lifing analysis. He has never computed the useful life of a core deposit intangible asset.
G. Robert E. Litan, Ph.D.
Dr. Robert E. Litan (Dr. Litan) works at the Brookings Institution, performing economic research in the areas of financial services, international trade, and legal liability. Dr. Litan is of the opinion that a core deposit intangible asset does not have a limited useful life and can never be separate and distinct from goodwill. He believes, however, that, all other things being equal, a bank would be worth more with a greater proportion of core deposits.
In addition to Dr. Litan's legal conclusions and disapproval of prior core deposit cases decided by this Court, he also criticizes the methods used by petitioner to determine the useful life and value of the core deposit intangible.
1. Critique of Dr. Doerfler's Analysis
Dr. Litan bases his conclusion that petitioner could not measure the useful life of the deposit base with reasonable accuracy on the reports of Dr. Stuk and Dr. Roberts, submitted on behalf of respondent.
2. Critique of Curtis' Analysis
Dr. Litan's initial criticism of Curtis' valuation is that "Curtis' analysis assumes that any percentage `spread' between the costs of core deposits and alternative funds on the date of acquisition remains constant throughout time. "
Dr. Litan also concluded that Curtis overstated the value of the core deposit intangible asset because he understated the cost of the core deposits by failing to take into account implicit interest. Lower interest rates on loans to depositors and expedited loans to a bank's customers are examples of implicit interest. Although, theoretically, it would be appropriate to include implicit interest in calculating the value of the deposit base, Dr. Litan admitted that, from a practical standpoint, it was impossible to include it.
Dr. Litan concluded that, because Curtis could not have estimated either the implicit interest or the spread, he could not estimate with reasonable accuracy the value of the core deposit intangible asset.
Dr. Litan stated that Curtis should have used the 46-percent marginal corporate tax rate rather than the 20-percent effective tax rate. He did not, however, explain why he believed this would be more appropriate.
H. Wayne Nelson, Ph.D.
Dr. Wayne Nelson (Dr. Nelson) is a statistical consultant and an adjunct professor of applied statistics at Union College in Schenectady, New York. He is the author of the book Applied Life Data Analysis.
Dr. Nelson generally agreed with the criticisms of Dr. Doerfler's work that were made by Dr. Roberts and Dr. Stuk. Dr. Doerfler used some methods set out in Dr. Nelson's book. Dr. Nelson believes that Dr. Doerfler also used methods not set out in Dr. Nelson's book, Applied Life Data Analysis, which were questionable. Unlike respondent's other experts, Dr. Nelson, while criticizing Dr. Doerfler's poor technical writing style, testified that he (Dr. Nelson) could, with one exception that will be discussed below, understand what Dr. Doerfler was saying in his report.
Aside from the criticisms mentioned above, Dr. Nelson found the following problems with the data. First, the data covers only 1 year and should cover a longer period of time. Second, Dr. Doerfler rounded the age of the accounts to the nearest year and then used a mid-year convention. Dr. Nelson believes it would have been better, rather than using a mid-year convention, if Dr. Doerfler had rounded them to the nearest day. Third, Dr. Doerfler used account balances as of December 14, 1983. Dr. Nelson thought that: "It would be much better if he had used complete histories of accounts or at least monthly balances so he could follow the history of individual accounts or a population of accounts to see how the dollars behave."
Although Dr. Nelson initially believed that Dr. Doerfler's analysis of the relationship between dollar balances and numbers of accounts was flawed, upon further study and after the cross-examination of Dr. Stuk he concluded that Dr. Doerfler was correct:
Q Are you aware of any statistical practice where one can assume that one can apply the results of account decay and predict decline in account balances?
A No. I think you're talking about the interpretation of Dr. Doerfler's report where it sounded as if 10 percent of the accounts are closed and 10 percent of the money leaves the bank.
This morning I discovered that his analysis is more realistic than that. So I must say that the original impression I had of his method is incorrect and now that I better understand his method, what he is doing is different from assuming that if 10 percent of the accounts leave the bank, then 10 percent of the money leaves the bank.
* * *
Let me say that I entirely misread what he [Dr. Doerfler] wrote in the report, misunderstood it until this morning when I overheard the cross examination of Dr. Stuk.
* * *
I think I was able to read almost everything else in that report when few people could, but that's one thing I missed.
Dr. Nelson thought that further stratification beyond the $1,500 level "could be beneficial".
He noted the problems inherent in Dr. Doerfler's pooling of accounts opened before 1969.
What he really wants to calculate here, and what he purports to calculate here, is the fraction of accounts that will survive from age 13 and a half years to age 14 and a half years. Well, to do that, he should be using accounts in that age range, and only accounts in that age range. But he's using accounts that are much older. The result of this is that he will have fewer accounts closing because the accounts are older than 13 and a half to 14 and a half years, so the number he calculates on the far right called the "peak survival," it's .926; that number should really be smaller.
Q So that's an artificial number, isn't it?
A Yes. This number is entirely incorrect. And he goes and plots it on the plot and uses that point which is in the upper tail, so it's a very influential point in determining the Weibull distribution and extrapolating beyond that point. It's the last point of data before you extrapolate.
* * *
So that's just plain wrong, and biased towards making these look longer lived than they are.
Dr. Nelson agreed that the Weibull curve provided a reasonable fit between zero and 14 1/2 years. The problems were in the years beyond that.
Dr. Nelson did not perform his own analysis of the useful life of the core deposit intangible asset.
Petitioner's Treatment of the Deposit Base
For financial, regulatory, and income tax accounting purposes, petitioner has been depreciating the amount it allocated to the deposit base of the Acquired Bank. For Federal income tax purposes, petitioner amortized the amount of the purchase price allocated to deposit base intangible using an accelerated method of amortization in accordance with the following schedule:
For financial and regulatory purposes, petitioner has amortized the amount allocated to deposit base on a straight-line basis over a period of 10 years.
Petitioner filed applications for tentative carryback allowances for 1973 and 1974. Respondent paid tentative refunds to petitioner in the amounts of $2,664,974 for 1973 and $2,530,793 for 1974. In the notice of deficiency, respondent disallowed in their entirety the depreciation deductions claimed by petitioner on its Federal income tax returns for amortization of the amount allocated to deposit base intangible in the amounts of $50,880 for 1983 and $1,079,551 for 1984.
Section 167(a) allows "as a depreciation deduction a reasonable allowance for the exhaustion, wear and tear * * * of property used in the trade or business".
If an intangible asset is known from experience or other factors to be of use in the business or in the production of income for only a limited period, the length of which can be estimated with reasonable accuracy, such an intangible asset may be the subject of a depreciation allowance. * * * intangible asset, the useful life of which is not limited, is not subject to the allowance for depreciation. * * * No deduction for depreciation is allowable with respect to good will. * * *
Sec. 1.167(a)-3, Income Tax Regs.
In Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. 463 (1988), affd. without published opinion 900 F.2d 266 (11th Cir. 1990), affd. per curiam [91-1 USTC ╤ 50,043] 919 F.2d 1492 (11th Cir. 1990), we expressly noted the factual nature of the core deposits inquiry. We stated:
To qualify under the foregoing regulation for a depreciation deduction with respect to the deposit base, petitioner must show that the deposit base (1) had an ascertainable cost basis separate and distinct from the goodwill and going-concern value of the Acquired Banks, and (2) had a limited useful life, the duration of which could be ascertained with reasonable accuracy. Whether these requirements have been satisfied is essentially a factual question. * * *
Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. at 479 (emphasis added) (citing Houston Chronicle Pub. Co. v. United States [73-2 USTC ╤ 9537], 481 F.2d 1240, 1245-1251 (5th Cir. 1973); Banc One Corp. v. Commissioner [Dec. 41,985], 84 T.C. 476, 492 (1985), affd. without published opinion 815 F.2d 75 (6th Cir. 1987); Los Angeles Central Animal Hospital v. Commissioner [Dec. 34,425], 68 T.C. 269, 273 (1977)).
Overtaken by events, respondent's arguments in this case are too little and too late. This Court has already decided a number of bank core deposit cases 12 and does not choose to revisit, let alone overrule, its Court-reviewed opinion in Citizens & Southern Corp. v. Commissioner, supra. 13 Also the Supreme Court has cited Citizens & Southern Corp. with approval and has rejected respondent's underlying legal argument that as a matter of law core deposits (and other "customer-based intangibles") are inseparable from goodwill/going concern value and thus nondepreciable. Newark Morning Ledger Co. v. United States [93-1 USTC ╤ 50,228], 507 U.S. ___, 113 S.Ct. 1670, 1676 (1993). 14
12 IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. 496 (1991); Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. 463 (1988), affd. without published opinion 900 F.2d 266 (11th Cir. 1990), affd. per curiam [91-1 USTC ╤ 50,043] 919 F.2d 1492 (11th Cir. 1990); Peoples Bancorporation v. Commissioner [Dec. 48,226(M)], T.C. Memo. 1992-285; Colorado National Bankshares, Inc. v. Commissioner [Dec. 46,875(M)], T.C. Memo. 1990-495, affd. [93-1 USTC ╤ 50,077] 984 F.2d 383 (10th Cir. 1993).
13 Respondent's first argument is that we should overrule our Court-reviewed opinion in Citizens & Southern Corp. v. Commissioner, supra. Respondent claims that this Court was misled in Citizens & Southern Corp. by expert statisticians whose assumptions and methodology went unchallenged by the Commissioner in that case. We will decide the instant case on its own facts. We decline, however, to overrule our opinion in Citizens & Southern Corp. on the basis of respondent's current claims and will continue to apply the reasoning of that case.
14 Respondent also argued on brief that we should revisit Citizens & Southern Corp. in light of the Third Circuit's opinion in Newark Morning Ledger Co. v. United States [91-2 USTC ╤ 50,451], 945 F.2d 555 (3d Cir. 1991), revd. [93-1 USTC ╤ 50,228] 507 U.S. ___, 113 S.Ct. 1670 (1993), and in light of our opinion in Ithaca Industries, Inc. v. Commissioner [Dec. 47,536], 97 T.C. 253 (1991), affd. [94-1 USTC ╤ 50,100] 17 F.3d 684 (4th Cir. 1994). The Supreme Court has now reversed the Third Circuit, and our Ithaca Industries opinion is wholly distinguishable on its facts.
In Newark Morning Ledger, the taxpayer purchased eight newspapers and attempted to depreciate the portion of the purchase price that related to "paid subscribers". The Third Circuit held that generally customer lists are not depreciable in connection with the sale of a going concern because it is too difficult to separate them from their goodwill value. Newark Morning Ledger Co. v. United States [91-2 USTC ╤ 50,451], 945 F.2d 555, 568 (3d Cir. 1991), revd. [93-1 USTC ╤ 50,228] 507 U.S. ___, 113 S.Ct. 1670 (1993). The view of the Third Circuit was that "for tax purposes, there are some intangible assets which, notwithstanding that they have wasting lives that can be estimated with reasonable accuracy and ascertainable values, are nonetheless goodwill and nondepreciable." Id. The Supreme Court rejected that approach.
In Newark Morning Ledger Co. v. United States [93-1 USTC ╤ 50,228], 507 U.S. at ___, 113 S.Ct. at 1672, the Supreme Court framed the issue as whether under section 167 the Internal Revenue Service (IRS) may "treat as nondepreciable an intangible asset proved to have an ascertainable value and a limited useful life, the duration of which can be ascertained with reasonable accuracy, solely because the IRS considers the asset to be goodwill as a matter of law." The Supreme Court accepted the short-hand description of goodwill as "the expectancy of continued patronage", but noted that the definition is of "little assistance" to a taxpayer trying to evaluate which of its intangible assets is subject to a depreciation allowance. Addressing the broad issue of "customer-based intangibles", the Supreme Court said:
The category of intangibles that has given the IRS and the courts difficulty is that group of assets sometimes denominated "customer-based intangibles." This group includes customer lists, insurance expirations, subscriber lists, bank deposits, cleaning-service accounts, drugstore-prescription files, and any other identifiable asset the value of which obviously depends on the continued and voluntary patronage of customers. The question has been whether these intangibles can be depreciated notwithstanding their relationship to "the expectancy of continued patronage."
Newark Morning Ledger Co. v. United States [93-1 USTC ╤ 50,228], 507 U.S. at ___, 113 S.Ct. at 1676. After reviewing the case law, including three of our core deposit cases, 15 the Supreme Court rejected the Government's legal position, saying:
15 Citizens & Southern Corp. v. Commissioner, supra; IT&S of Iowa, Inc. v. Commissioner, supra; and Colorado National Bankshares, Inc. v. Commissioner, supra.
The significant question for purposes of depreciation is not whether the asset falls "within the core of the concept of goodwill," Brief for United States 19, but whether that asset is capable of being valued and whether that value diminishes over time. * * *
Newark Morning Ledger Co. v. United States [93-1 USTC ╤ 50,228], 507 U.S. at ___, 113 S.Ct. at 1681.
The Supreme Court expressly held that a taxpayer able to prove that a particular asset can be valued and that it has a limited useful life may depreciate its value over its useful life "regardless of how much the asset appears to reflect the expectancy of continued patronage". Id. The Supreme Court recognized the heavy evidentiary burden placed on the taxpayer, but reaffirmed, as has this Court in its core deposit cases, that this is a question of fact.
Respondent made no attempt to value the core deposit intangible in this case or to determine its useful life. While criticizing the quality of the statistical evidence and Dr. Doerfler's methodology in Citizens & Southern Corp. v. Commissioner, supra (see supra note 13), respondent in this case presented statistical experts who criticized Dr. Doerfler's methodology but offered none of their own. The Court must point out that respondent's well-qualified statistical experts, Dr. Stuk and Dr. Nelson, were not asked to determine the useful life of the core deposits intangible asset. Having taken the position that as a matter of law the core deposits intangible can have no ascertainable value or useful life separate and distinct from goodwill and going concern value, respondent was consistent, but not particularly helpful to this Court, in refusing to try to determine its value and useful life.
We have analyzed the issue of separability of the deposit base and goodwill on several occasions in this Court. IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. 496 (1991); Citizens & Southern Corp. v. Commissioner, supra; Peoples Bancorporation v. Commissioner [Dec. 48,226(M)], T.C. Memo. 1992-285; Colorado National Bankshares, Inc. v. Commissioner [Dec. 46,875(M)], T.C. Memo. 1990-495, affd. [93-1 USTC ╤ 50,077] 984 F.2d 383 (10th Cir. 1993). Nonetheless, respondent tenaciously adhered to the argument that, as a matter of law, core deposits can never give rise to an intangible asset that is separate and distinct from goodwill and going concern value. Respondent continued to argue that the premium associated with these deposits cannot be distinguished from goodwill. We disagreed with respondent's argument. As the United States Court of Appeals for the Fifth Circuit noted in Houston Chronicle as regards customer lists:
Our view ≈ that amortizability for tax purposes must turn on factual bases ≈ is more in accord with the realities of modern business technology in a day when lists are bartered and sold as discrete vendible assets. * * *
Houston Chronicle Pub. Co. v. United States [73-2 USTC ╤ 9537], 481 F.2d 1240, 1253 (5th Cir. 1973). The Supreme Court adopted this modern approach in Newark Morning Ledger Co. v. United States, supra. Respondent's position has now been rejected by the Supreme Court, by this Court, and by mounting evidence to the contrary in the marketplace.
Core deposits are bought and sold in the marketplace. Beginning in the late 1980s, banks have bid for and acquired the deposit liabilities of failed thrift institutions through the Resolution Trust Corporation (RTC). In these transactions, the bank pays money to assume the deposit liabilities. In such circumstances, the bank may or may not purchase any other assets or properties of the thrift institution. These marketplace developments were merely hinted at in our prior core deposit cases, but the facts are undisputed in this case.
Clearly, in the above RTC sales of core deposits, the acquiring bank is not purchasing goodwill or going concern value of a failed financial institution because there is little or no goodwill or going concern value to be purchased. The acquiring bank may later generate its own goodwill as a result of the purchase of the deposits, but one could say that about any asset purchased by a business. The use of that asset in the acquirer's line of business may create future goodwill in the acquirer, but this is not the equivalent of goodwill or going concern value of a defunct, failed financial institution. These separate sales of only core deposits are persuasive proof that core deposits have a value and cost basis separate and distinct from goodwill or going concern value.
The core deposit intangible asset may be amortized upon a proper showing by petitioner of its cost basis and a reasonably accurate estimate of its useful life.
1. Determining Useful Life of the Deposit Base
a. Admissibility of Dr. Doerfler's report
Respondent argues that the report of Dr. Doerfler, and the Curtis and Harper reports which rely on Dr. Doerfler's report, should be stricken from the record in this case. We do not find, and respondent does not suggest, that any data was altered in going from McLeod to Harper to Doerfler. See supra note 7.
We initially note that Dr. Doerfler's report is not a model of clarity. To be helpful to the Court, the expert must assist the Court to understand the area of expertise and thus be able to judge its usefulness in deciding the issue before the Court. Otherwise litigation becomes a battle of the experts, with the Court merely choosing between experts rather than weighing all expert opinion as it contributes or fails to contribute to resolution of the issue.
Dr. Doerfler agreed that he was "probably guilty of assuming everybody in the world is a statistician." Here the Doerfler report was so lacking in explanation of his methodology, assumptions, and computations that respondent's statistical experts, Dr. Stuk and Dr. Nelson, could not follow his reasoning or duplicate his results. See supra note 6. The education of respondent's experts as to Dr. Doerfler's methodology was accomplished during the lengthy trial rather than during trial preparation, contrary to our Rule 143(f). Under our expert rule, the expert's report is exchanged 30 days before trial and is to be wholly self-explanatory and self-contained, since it normally constitutes the expert's direct examination. Here Dr. Doerfler's workpapers had to be subpoenaed and produced at trial in order for respondent's experts to understand his methodology, assumptions, and computations. See supra note 6.
Dr. Stuk, in frustration, opined that "I believe there's probably only two people in the world who could read that report and recreate the work with any hope, including all existing statisticians." Dr. Nelson, the most qualified of the statistical experts to testify in this case, was more charitable. He indicated that while Dr. Doerfler's report was "very difficult reading" and Dr. Doerfler was not a good technical writer, nonetheless he (Dr. Nelson) was able to understand most of the report. However, one key item became clear to Dr. Nelson only near the end of the trial, during petitioner's cross-examination of Dr. Stuk. Dr. Nelson admitted that "I think I was able to read almost everything else in that report when few people could, but that's one thing I missed." The one thing Dr. Nelson had missed up to that point was that Dr. Doerfler was not saying that when 10 percent of the accounts closed that 10 percent of the money left the bank.
If the Court chose to enforce the strict requirements of its Rule 143(f), as respondent vigorously urges, the Court could well agree to strike Dr. Doerfler's report from the record. Petitioner would then lose this case for failure of proof. However, at this point, the Court, counsel for both parties, the litigants, and the witnesses have invested too much time and effort in this litigation to warrant such a draconian result. Also, the difficulties encountered by respondent's experts could easily have been averted by taking depositions of the experts, as the Court on numerous occasions before trial had strenuously urged both parties to do. Counsel and the experts having been educated during the course of the protracted trial, the Court can perceive no compelling reason to disregard what has been learned. The Court will not exclude any of the expert reports from evidence and will instead decide the case on the basis of the record as a whole, including Dr. Doerfler's report.
We also note that Dr. Doerfler's report is the only attempt we have in this case to determine the useful life of the core deposit intangible asset. Neither Dr. Stuk nor Dr. Nelson was asked to make a statistical analysis of the useful life of the core deposit intangible asset. Viewing their testimony as a whole, the Court does not think they were saying it could not be done, only that they would have done it differently. Respondent's insisting, contrary to the decided cases and the marketplace realities, that it cannot be done is a futile attempt to refight a battle that was lost as long ago as Houston Chronicle Pub. Co. v. United States [73-2 USTC ╤ 9537], 481 F.2d 1240 (5th Cir. 1973). With the recent Supreme Court opinion in Newark Morning Ledger Co. v. United States, supra, the issue now is clearly just a question of fact. 16 The task is to determine the useful life of groups of accounts, not of any particular account as respondent seems to argue, and the standard is one of reasonableness.
16 That could presage years of fierce litigation ahead on such factual issues, but with Congress' prospective legislation on amortization of intangibles, one can hope that the "customer-based intangibles" battles can be permitted to recede into the mists of history.
Extreme exactitude in ascertaining the duration of an asset is a paradigm that the law does not demand. All that the law and regulations require is reasonable accuracy in forecasting the asset's useful life.
Houston Chronicle Pub. Co. v. United States [73-2 USTC ╤ 9537], 481 F.2d at 1253-1254.
Dr. Doerfler's report, though flawed, is admissible and has been admitted by the Court. We will now address whether Dr. Doerfler's report properly applies the methodology we approved in Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. 463 (1988), affd. without published opinion 900 F.2d 266 (11th Cir. 1990), affd. per curiam [91-1 USTC ╤ 50,043] 919 F.2d 1492 (11th Cir. 1990), and whether it provides an adequate basis for determining the useful life with reasonable accuracy.
b. The lifing analysis
We have the benefit of subsequent studies to corroborate the reasonable accuracy of Dr. Doerfler's predictions. Although the determination of the useful life of an asset must be based upon facts in existence at the close of the taxable year in issue, evidence of petitioner's subsequent experience is acceptable to corroborate its projections. Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. at 500; Banc One Corp. v. Commissioner [Dec. 41,985], 84 T.C. 476, 499, 501 (1985), affd. without published opinion 815 F.2d 75 (6th Cir. 1987). We conclude, as will be discussed later, that the money market and NOW accounts are not properly included in the deposit core. We thus have the following comparison of Dr. Doerfler's predictions and Harper's analysis of the actual deposits for the remaining DDA's:
We have the following composite comparison for the SDA's:
The predictions are reasonably close to the actual, particularly for the SDA's. While there is greater variance for the remaining DDA's, we think this is generally consistent with our conclusion that the money market and NOW accounts should be excluded from the deposit core. As we noted in the Citizens & Southern Corp. case, after obtaining results roughly similar to the ones herein:
This comparison shows that petitioner's projections were substantially the same as Doerfler's projections which were calculated using statistical analysis. This comparison also shows that the validity of petitioner's projections was corroborated by actual closures.
Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. at 502.
Dr. Doerfler compared the accounts that were open on December 31, 1982, with those accounts remaining open on December 14, 1983. Thus, the observation period was approximately 1 year. Respondent argues that this observation period is an inadequate basis for predicting deposit runoff 20 years into the future. In our prior core deposit cases, we have approved observation periods ranging from 3 months to 3 years. In IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 505, we approved the use of a 1-year observation period ending around the time of acquisition. In Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. at 470, the taxpayer looked at the number of accounts closed each month during observation periods ranging from 3 to 13 months. See also Peoples Bancorporation v. Commissioner [Dec. 48,226(M)], T.C. Memo. 1992-285 (observation periods ranging from 2 to 11 months); Colorado National Bankshares, Inc. v. Commissioner [Dec. 46,875(M)], T.C. Memo. 1990-495, affd. [93-1 USTC ╤ 50,077] 984 F.2d 383 (10th Cir. 1993) (taxpayer selected two points in time approximately 1 year apart).
Dr. Doerfler stratified the accounts based upon account type, age of accounts, and, in some cases, account size. We find Dr. Doerfler's stratification levels to be adequate for the purpose of calculating a reasonably accurate useful life.
The taxpayer in the Citizens & Southern Corp. case stratified only by account type and age. As we stated therein:
Petitioner's stratification of the accounts resulted in accurate projections. It appears that petitioner stratified the accounts based upon the factors which it considered reasonable and relevant to determine the lives of the accounts. We will not disregard petitioner's projections because respondent thinks they should have been stratified differently.
Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. at 505, see also IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 513.
We also adhere to our prior position that account closures can be used to measure the decline in account balances, where, as here, adjustments were made for the aging of the accounts. Petitioner's methodology has been previously approved by this Court. See Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. at 469-471. Petitioner in this case has shown the reasonable accuracy of its method by the follow-up study prepared by Harper. This is an acceptable means by which to show a correlation between account closures and decline in account balances. IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 512; see also Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. at 502-503.
2. Valuation of the Deposit Base
We first consider whether petitioner used a proper method to value the deposit base. We have previously approved the cost savings methodology as a proper measure of the value of the deposit base. Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. at 506-507, 510. While we disagree with respondent that the valuation methodology is conceptually flawed, we do find some flaws in petitioner's application of the cost savings analysis.
a. What constitutes the deposit core?
In IT&S of Iowa, Inc. v. Commissioner, supra, we noted two essential features of core deposits ≈ interest rate insensitivity and stability of deposits. IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 516-517. We excluded from the composition of the core deposits those accounts which were hypersensitive to interest rate changes; i.e., "hot money". Id. at 517. We also excluded those accounts which were sensitive, but not hypersensitive, to interest rate changes, because if deposits "are sensitive to interest rate fluctuations, they offer no potential for extra-normal profitability and there is no potential for cost savings associated with them." Id. at 516.
We noted therein:
Adjustable rate deposit accounts, such as certificates of deposit, money market deposit accounts, and super NOW accounts, were expressly created in order to allow banks to compete successfully for deposits against other financial intermediaries not bound by the interest rate ceilings applicable to banks. * * * Inasmuch as such accounts were designed to be the interest rate sensitive and rate competitive products of banking, it is illogical to include them automatically in core deposits, a term defined to mean deposits which are not sensitive to interest rate changes.
Id. at 517
We did not foreclose the possibility of including money market accounts and super NOW accounts in the deposit core in situations where the taxpayer could show that such deposits were in fact insensitive to interest rate changes. In the instant case, petitioner did not include certificates of deposit or super NOW accounts in its valuation of the deposit core. As for the NOW accounts and the money market accounts, however, petitioner has not carried its burden of proof on the issue of interest rate insensitivity, and they may not be considered in valuing the deposit core. Rule 142(a); IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 519; Colorado National Bankshares, Inc. v. Commissioner, supra. 17
17 We would also exclude the NOW accounts for a further reason. Over 45 percent of those accounts had unknown opening dates. The age of the accounts was a "very critical" factor in Dr. Doerfler's methodology. Since he was modeling, in his statistical analysis, the age of the accounts and not dollar balances, we have no confidence in his useful life projections for the NOW accounts.
b. Accounting for implicit interest
In valuing the deposit core, Curtis took into account the deposit interest expense and the deposit service expense. Respondent, relying on Dr. Litan's testimony and our opinion in IT&S of Iowa, Inc. v. Commissioner, supra, criticizes Curtis' failure to take implicit interest into account. Dr. Litan admitted, however, that from a practical standpoint, it is impossible to include this in calculating the value of the deposit base. The failure to include implicit interest in the calculations is not fatal to petitioner's valuation.
Respondent's reliance on IT&S of Iowa, Inc. v. Commissioner, supra, is misplaced. In that case, what was called "implicit interest" is the equivalent of what Curtis in this case termed "deposit service expense".
Once the deposit core has been defined, its cost to the bank, both in terms of explicit interest and the cost of services provided to depositors, called "implicit" interest, must be ascertained. * * *
Petitioner calculated the cost of implicit interest furnished depositors by allocating a portion of each category of expense incurred by the What Cheer bank to the deposit function. * * *
IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 521.
In the present case, Dr. Litan cited as examples of implicit interest immeasurable benefits to depositors such as lower interest rates on loans to depositors and expedited loans to a bank's customers. There is no evidence that these supposed benefits even existed in this case. Any expense to the bank resulting from these supposed depositor benefits was not even considered in IT&S of Iowa, Inc. v. Commissioner, supra. We note that petitioner is only required to furnish a reasonably accurate determination of value. As Dr. Litan noted, these "expenses" are, on a practical level, impossible to measure. We have no evidence that petitioner's failure to take into account what respondent here has labeled "implicit interest" distorts the value of the deposit core beyond what is reasonably accurate. Petitioner has properly accounted for its expenses in maintaining the accounts.
c. The effect of deregulation
As in IT&S of Iowa, Inc. v. Commissioner, supra, we have defined the deposit core as including only accounts that are insensitive to interest rate changes. We reject respondent's broad argument, however, that there is a "likelihood that deregulation will be expected to cause turmoil in the shifting of funds from lower to higher yielding accounts." To the extent that deregulation had any effect in this case, we have accounted for it by excluding the NOW and money market accounts from the deposit core. Also the record indicates that the major changes caused by deregulation had taken place by 1982 and the markets had settled by 1983.
d. Cost of alternative funds
The cost savings methodology measures the present value of the difference between the cost of an asset and the cost of the next most favorable market alternative. IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 514, 525; Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. at 498, 506. Curtis chose insured certificates of deposit (CD's) as the alternative source of funds because of their similarity to core deposits in terms of insurability, average balance size, and maturity or remaining life. This is a reasonable basis for comparison. IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 525; Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. at 507. In prior cases, we have approved the use of insured CD's as an alternative funding source. IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 525; Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. at 507. We do so again here.
Curtis assumed that the yield curve on CD's would be flat beyond 5 years and so applied the 5-year rate. We held in IT&S of Iowa, Inc. that "the term to maturity of a financial instrument is not decisive of comparability." IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 528. As we explained in that case:
Consequently, in deciding comparability, the actual life of a group of deposits, not its term to maturity, is the key characteristic. In deciding comparability, we therefore focus on the time over which the bank can be expected to have the use of the funds based on attrition data, rather than on the time when the instruments mature. * * * Accordingly, where, based on attrition rates, funds in certificates of deposit can be expected to remain with a bank for a period similar to that of core deposits, the funds should not be disqualifed as an alternative funding source simply because the individual certificates do not have the same time to maturity as the life of the core deposit intangible.
IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 528-529. Respondent relies on nothing but speculation in the argument that the 5-year yield is inappropriate as an alternative cost.
Respondent also takes issue with petitioner's use of its competitors' interest rates on CD's. In IT&S of Iowa, Inc., we suggested that the best source of comparison are the rates paid by the taxpayer on interest rate sensitive deposits. IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 529. There is no evidence in the record of the interest rates paid by petitioner on CD's. Given general market considerations, however, petitioner would have been unable to attract CD investments if it had paid a significantly lower interest rate than its competitors. Therefore, we find the competitors' CD interest rates to be a reasonable cost of alternative funds to petitioner.
e. Assuming a constant spread between costs
Curtis assumed that the difference between the cost of core deposits and the cost of alternative funds would remain constant throughout the useful life of the core deposits. Respondent argues against this approach, adhering to the underlying position that the deposit base, even if it exists, cannot possibly be valued. We disagree with respondent's position. There is no evidence to suggest that rate spreads could not be predicted or that Curtis' predicted rate spreads were unreasonable.
f. Computing after-tax cost savings
In computing the after-tax cost savings, Curtis first defined taxable cost savings as the pre-tax cost savings reduced by the amortization deduction. He then applied the effective tax rate of 20 percent rather than the statutory rate of 46 percent. For the following reasons, we conclude that this method is improper.
We approved the use of the effective tax rate in our opinion in Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. at 506-507, 510 and n. 31 (effective tax rate of 20 percent used "because that rate conformed to banking industry experience in the relevant period"). See also Colorado National Bankshares, Inc. v. Commissioner [Dec. 46,875(M)], T.C. Memo. 1990-495, affd. [93-1 USTC ╤ 50,077] 984 F.2d 383 (10th Cir. 1993). However, in both of those cases the effective tax rate was applied to cost savings unreduced by the amortization deduction. In IT&S of Iowa, Inc. v. Commissioner, supra, however, the taxpayer used the maximum statutory rate of 48.7 percent and also used the amortization deduction, but in a somewhat different method. While our prior opinions may not have been wholly clear on this point, we follow our reasoning in IT&S of Iowa, Inc. and hold that petitioner may not use both the effective tax rate and the amortization deduction in determining the value of the deposit base. We analyzed this issue in IT&S of Iowa, Inc. as follows:
Petitioner's calculation of the savings in the instant case varies from that used by the taxpayer in Citizens & Southern. In that case, the after-tax value of the deposit core was figured by applying a 20-percent effective tax rate to its projected income stream. [Dec. 45,036] 91 T.C. at 506-507. In the instant case, petitioner adopted a two-step approach to calculating the effect of taxes on the deposit core. In the first step, the cost savings generated by the deposit core were reduced by a marginal tax rate of 48.7 percent in each year of its life. In the second step, the tax savings produced by the deposit core depreciation deductions were added to the value of the cost savings generated by the core to yield the overall value for the core. Petitioner's method takes account of the tax effect of depreciation in reducing the effective rate of tax below the marginal rate by adding back some of the tax taken out in the first part of the calculation. Petitioner's calculations therefore accomplish the same end as the application of an effective tax rate to the deposit core cost savings approved in Citizens & Southern. Accordingly, we hold that petitioner's method is a proper calculation of the tax savings in the instant case.
IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 532-533.
The tax benefit of the amortization deduction is already included in the effective tax rate used by Curtis. Including this benefit again in the valuation, as Curtis did here, artificially inflates the value of the deposit base. We hold that, because petitioner has used the effective tax rate of 20 percent, it may not take into account the amortization deduction in its valuation calculation. However, we agree with petitioner that the use of the 46-percent statutory rate as opposed to the 20-percent effective rate does not decrease the value of the deposit base as much as respondent seems to think. This is because, with the use of the 46-percent statutory rate, the tax benefit of the amortization deduction is greater. 18
18 Using the figures in petitioner's example in its requested finding 94 (p. 37 of Petr. Opening Br.):
Petitioner's computation is as follows:
x - .20 (x - y) = $487,588 after-tax cost savings
Under the methodology of our IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. 496 (1991), opinion, the computation should be:
x - .46x + .46y = $477,344 after-tax cost savings
The $10,244 difference between the two methods is attributable to the .26 difference in tax rate (.46 - .20) applied to taxable cost savings (x - y):
x - .46 (x - y) = $477,344 after-tax cost savings
g. The discount rate
Curtis determined a discount rate of 14.3 percent by analyzing the weighted average cost of debt and equity capital. Respondent relies on our opinion in IT&S of Iowa, Inc. in support of the argument that this method of determining a discount rate is improper. Respondent does not refer us to any part of the evidentiary record in this case in support of this argument.
In IT&S of Iowa, Inc., the taxpayer used a weighted average cost of capital of 7.667 percent as the discount rate. "Such figure was calculated as the sum of the after-tax cost of longterm debt capital and after-tax desired rate of return on equity, weighted according to the relative percentages of debt, in the form of deposit liabilities, and equity funding petitioner's assets on the acquisition." IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 530 (emphasis added; fn. ref. omitted). We concluded that the appropriate discount rate, based on the facts in that case, was the taxpayer's "after-tax return on equity as of the time of the acquisition." IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 531. In reaching this conclusion, we relied upon the testimony of respondent's expert, Dr. Kane:
Respondent's expert, Dr. Kane, concluded that, because the purpose of the valuation was to calculate the payment made by petitioner's stockholders for the stream of cost savings generated by the What Cheer bank deposit base and because the cost savings of the core deposits accrue to petitioner's equity holders, the rate of interest payable to a hypothetical class of debtholders was conceptually irrelevant. Furthermore, long-term debt like the hypothetical alternative funding source did not form part of petitioner's capital at the time of the acquisition, and so the interest rate payable on such debt did not affect petitioner's cost of capital at such time.
Dr. Kane elaborated that, under corporate finance theory, the appropriate discount rate to use would be the marginal capitalization rate, which equals the weighted average of the returns paid on the mix of debt and equity actually used to finance the acquisition of the What Cheer bank. In the instant case, the cash used to acquire the What Cheer bank came from petitioner's equity capital. No debt was issued to finance the acquisition, although petitioner subsequently issued preferred stock after the acquisition in order to satisfy the minimum capital requirements of Iowa banking regulators.
IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 531.
This portion of our analysis in IT&S of Iowa, Inc. was very fact specific. The relative sizes of the taxpayer in IT&S of Iowa, Inc. and petitioner in this case suggest that the same analysis does not apply here. We noted in IT&S of Iowa, Inc. that
Dr. Kane also testified that it was very difficult for small banks, such as petitioner, to determine an appropriate capitalization rate. Consequently, we hold, under the facts of the instant case, that the 20-percent after-tax rate of return on equity used in petitioner's study is the best available information * * *
IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 531-532 (emphasis added). On the record in the instant case, we are unable to conclude that a discount rate of 14.3 percent is too low. Nothing in our record suggests that a higher rate is warranted.
h. The inflation factor
To reflect long-term inflation, Curtis assumed that the average balance of all types of accounts would increase at a rate of 5 percent per annum. On brief respondent objects to this. However, respondent's expert, Dr. Litan, agreed on cross-examination that an inflation factor should be included and agreed that 5 percent was in "the right ballpark". Another of respondent's experts, Dr. French, in his report and on direct examination, criticized the use of an inflation factor. However, on cross-examination, it became evident that Dr. French's criticism was based on a misreading of this Court's opinion in Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. 463 (1988), affd. without published opinion 900 F.2d 266 (11th Cir. 1990), affd. per curiam [91-1 USTC ╤ 50,043] 919 F.2d 1492 (11th Cir. 1990). We approved the use of an inflation factor in that case, and we do the same here.
3. Amortization Method
The final issue is whether petitioner may amortize the core deposit intangible asset on an accelerated basis. "Although the straight-line method generally is used to depreciate intangibles, petitioner may use an accelerated method if it can show that such method results in a more reasonable depreciation allowance." IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 533; Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. at 512. We disagree with respondent's arguments that petitioner has failed to sustain its burden of proof on this issue.
"A method other than straight-line is appropriate when, considering the facts of the case, it `results in a fair allocation of the basis of the asset to periods in which income is realized.'" Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. at 512 (quoting Schneider v. Commissioner [Dec. 33,456], 65 T.C. 18, 32 (1975)); see also IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 533. As the taxpayer in Citizens & Southern Corp. did, petitioner in this case has established that the deposit base declines more quickly in the years immediately following acquisition. See Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. at 514. We conclude that petitioner's method of amortization of the deposit base results in a fair allocation of the basis of the asset to the periods in which its benefit is realized. We thus hold that petitioner may amortize its cost basis in the core deposit intangible asset on an accelerated basis.
To reflect the foregoing,
Decision will be entered under Rule 155.