Peoples Bancorporation and Subsidiaries v. Commissioner., United States Tax Court - Memorandum Decision, T.C. Memo. 1992-285, Docket No. 29058-89., Filed May 18, 1992
Peoples Bancorporation and Subsidiaries v. Commissioner.
United States Tax Court - Memorandum Decision
T.C. Memo. 1992-285
Docket No. 29058-89.
Filed May 18, 1992.
Philip C. Cook, Terence J. Greene, and Timothy James Peadon, 1201 W. Peachtree St., Atlanta, Ga., for the petitioners. Anne Hintermeister and William Stoddard, for the respondent.
Memorandum Findings of Fact and Opinion
HAMBLEN, Judge: Respondent determined deficiencies in Federal income tax as follows:
The issues remaining for decision involve petitioners' entitlement under section 167 to depreciate the value of core deposit intangibles 1 relating to two banks acquired by petitioners. 2 The issues are: (1) Whether petitioners have established that the core deposit intangibles of two acquired banks have an ascertainable value separate and distinct from goodwill; (2) whether petitioners have proved that the core deposit intangibles have a limited useful life; (3) whether the values and amortization schedules used by petitioners in calculating the depreciation deductions for the core deposit intangibles are reasonable; and (4) whether petitioners may amortize the core deposit intangibles using an accelerated method. 3
1 For convenience, the terms "core deposit" and "core deposit intangible" are used in the findings of fact section to refer to categories of deposits which petitioners or the witnesses consider amortizable. These references are not intended to indicate our conclusion that all of these deposits are amortizable; the extent to which we agree with the petitioners' contentions is discussed in the opinion section.
2 A Rule 155 computation will be necessary because several adjustments in the notice of deficiency were not raised as issues in the petition or amended petition filed. Unless indicated otherwise, all Rule references are to the Tax Court Rules of Practice and Procedure and all section references are to the Internal Revenue Code in effect for the years in issue.
3 In its amended petition, petitioners alleged that they incurred certain costs relating to the acquisitions of the banks. The parties stipulated to the amounts of these costs and that they should be allocated to the assets acquired in proportion to their respective fair market values. Petitioners proposed findings of fact allocating part of these costs to the core deposit intangibles. Respondent objected stating that the allocation should be to goodwill. Respondent did not argue on brief that these costs were not allocable based on fair market value to the core deposit intangibles if we concluded that core deposits were an asset separate and distinct from goodwill. Thus, the allocation is controlled by our conclusion regarding that issue.
Findings of Fact
Some of the facts have been stipulated and are found accordingly. The stipulation of facts and attached exhibits are incorporated herein by this reference.
Petitioners, Peoples Bancorporation (Bancorporation) and its subsidiaries, including Peoples Bank and Trust Company (Peoples) 4 are North Carolina corporations whose principal business is commercial banking. Petitioners' principal place of business was in Rocky Mount, North Carolina, at the time the petition was filed in this case.
4 Petitioners filed consolidated Federal income tax returns for the taxable years 1983, 1984, and 1985. Petitioners filed Corporation Applications for Tentative Refund, Forms 1139, relating to carryback allowances for Peoples for the taxable years 1971, 1974, and 1975; tentative refunds were paid by respondent in the amounts of $24,697, $186,549, and $377,155, respectively.
Peoples was incorporated and chartered by the State of North Carolina in 1931. Bancorporation was incorporated as a bank holding company on July 22, 1982, and acquired the stock of Peoples on January 1, 1983.
During the 1982-84 time period, Peoples operated primarily in northeastern North Carolina, with about 50 branch offices and $350 million in total assets.
Petitioners have never acquired or sold, or offered to acquire or sell, accounts or deposit relationships independent from the sale or purchase of a bank or a branch of a bank.
I. Banking Industry Background
Core deposits are a relatively low-cost source of funds, reasonably stable over time, and relatively insensitive to interest rate changes. The profitability of a bank depends on the difference between the cost of its funds and the interest earned on the loans and other investments into which the funds are invested, or the "spread". For example, before making a loan, Peoples computes the likely profit, and requires a profit or spread of about 5 percent in order to make the loan. As a lower cost source of funds, core deposits permit a larger spread, and thus higher profits, than other sources. There is attrition of core deposits for many reasons, for example, business closings or customer deaths.
For financial statement purposes, core deposits are considered liabilities, and goodwill is an amortizable asset. Before 1982, the Federal Deposit Insurance Corporation (FDIC) did not allow an acquiring bank to capitalize intangibles, including core deposits, in accounting for the purchase. In a letter dated March 5, 1982, the FDIC announced a change in this position to permit banks, on a case-by-case basis, to record as an asset and amortize acquired core deposit intangibles. In order to be permitted to do so, the bank had to submit with its merger application a detailed analysis to enable the FDIC to evaluate the account balances, the life thereof, the future net income stream therefrom, and the present value of the income stream.
In 1985, the FDIC issued regulations prohibiting banks from including core deposit intangibles in equity capital for purposes of assessing capital adequacy; this was not a change in position. The FDIC stated that: "Interest rate deregulation has been undermining the concept of the low cost core deposit base and assumptions about the average remaining lives of such deposits when acquired and the interest rate spreads projected over these lives have made the valuation of purchased core deposit intangibles increasingly subjective."
In an acquisition of a branch bank, the purchaser acquires assets, the customer base, established locations, and branch employees, and is able to avoid start-up costs.
Deposits are sometimes purchased separately from any other assets; the FDIC has done this in resolving insolvencies of troubled banks.
All deposits at commercial banks during the years in issue were insured up to $100,000 per account by the FDIC.
By 1982, deregulation of the banking industry was almost completed, which resulted in rising interest costs, increased competition, elimination of interest ceilings on certificates of deposit, and the creation of new types of accounts, such as money market and "super NOW" 5 accounts. As a result of deregulation, there was a shift to higher cost types of deposit accounts (i.e., a shift from demand deposits to certificates of deposit). During 1983 and 1984, the increased competition from deregulation was beginning to create pressure towards higher costs on transactional accounts.
5 The acronym "NOW" stands for "negotiable order of withdrawal".
In general, most bank deposits are held in a small number of high balance accounts, which tend to have longer lives than lower balance accounts.
II. NCNB Branch Acquisition
National Bank of North Carolina (NCNB) 6 entered into a plan of reorganization and merger agreement with the Bank of North Carolina (BNC) on March 26, 1982. As a condition to approval of the merger, NCNB was required by the regulatory authorities to divest two NCNB branches in Jacksonville, North Carolina, and six BNC branches in Wilmington, North Carolina (collectively sometimes referred to as the NCNB branches). The NCNB branches were full-service banks which offered the normal types of banking services, and were historically profitable.
6 The parties stipulated that the name of NCNB was National Bank of North Carolina.
NCNB's first step in this divestiture was to identify potential bidders, which the regulatory authorities required be banks not then operating in the Jacksonville or Wilmington markets. NCNB identified about 11 potential bidders, and prepared a package of information to forward to them.
NCNB estimated a selling price for the branches by using methodologies based on a multiple of book value and a multiple of earnings, and on computing a premium on certain deposits 7 of the branches and appraised values for fixed assets. By August 1982, NCNB had estimated a price of about $7 million, including estimated premiums of $544,850 and $4,802,600 for the core deposits at the Jacksonville and Wilmington branches, respectively. The greater value estimated for the deposits of the Wilmington branches was in part based on the strong market position of these branches. In September of 1982, NCNB estimated that the premium on the deposits at the branches would be $2.6 million.
7 Mr. David Watterworth, who was an NCNB department manager in charge of marketing research and branch planning in Charlotte, North Carolina, testified that the deposits considered were those less than $100,000. We infer based on the amounts of the premiums he computed that he only excluded certificates of deposit of $100,000 or more.
In November of 1982, NCNB delivered its information package soliciting bids to a number of potential bidders, including petitioners. The information package invited potential bidders to bid on either the Jacksonville or Wilmington branches or both. All offers were to state the total amount of cash to be paid. The information package stressed that the locations of the branches were in desirable areas with high growth, high median income, low unemployment, and other favorable economic characteristics. The package included income statements and balance sheets for the NCNB branches for 1978 through October of 1982, and a schedule of types of deposits as of October 31, 1982.
Petitioners retained Grant L. Hamrick (Hamrick), a managing partner in the Charlotte, North Carolina, office of Price Waterhouse, to assist in formulating a bid for the NCNB branches. Hamrick worked with petitioners' employees to analyze the impact of different bid prices on earnings and capital. His projections, which were presented at petitioners' December 3, 1982, executive committee meeting, assumed premiums for bidding ranging from $3 million to $6 million. In analyzing the impact on cash flow of acquiring the NCNB branches at a $4 million premium, Hamrick allocated $2.125 million to deposits and $1.875 million to goodwill.
By November 30, 1982, letter, NCNB furnished Peoples a draft purchase agreement for the NCNB branches, which neither provided for an allocation of the purchase price to, nor defined bank assets as including a category identified as, core deposits. The draft purchase agreement provided for a price adjustment based on the actual amount of deposits existing not later than 15 days after closing as compared with the amount shortly before closing.
By December 6, 1982, letter, Peoples made a bid of $6,175,000 for the NCNB branches (which was subject to regulatory authority and board of directors' approval, but preapproved by the executive committee). On December 8, 1982, NCNB accepted Peoples' bid subject to certain conditions, including regulatory and board approval.
By December 9, 1982, letter Hamrick forwarded a schedule of deposits as of October 31, 1982 for the NCNB branches for review by Dr. Bruce Morgan (Morgan) of Golembe Associates, Inc. (Golembe), an economist and consultant specializing in banking. Morgan had no input in determining the price Peoples offered in the bid. The schedule included demand deposits (including NOW and related deposits), savings accounts, and certificates of deposit. In a December 12, 1982, letter, Morgan provided the following preliminary estimates of the value of all deposits of the NCNB branches reflected on the schedule, excluding certificates of deposit of $100,000 or more: (1) $3.22 million before tax; (2) $4.82 million after tax; and (3) $4.39 million if only part of the tax benefits were considered. 8
8 The $4.39 million value allowed the tax benefit from the amortization of the deposits, but excluded the value of amortizing the additional value attributable to tax benefits. Morgan described this as a conservative approach, which would reflect the uncertainty at the time regarding amortization of deposits for tax purposes.
The board of directors of Peoples approved the purchase of the NCNB branches at a December 15, 1982, meeting. On December 20, 1982, Peoples and NCNB executed a purchase agreement for the NCNB branches, which included the following purchase price allocation proposed by Peoples and prepared by Hamrick:
The purchase agreement provided for a price adjustment if there was a variation of 7.5 percent or more between the amount of deposits (excluding certificates of deposit of $100,000 or more) at closing and the amount thereof reflected on a schedule dated October 31, 1982. The adjustment was equal to 13.4 percent of the amount of the deposit variation; the 13.4 percent factor was based upon dividing the purchase price by the total amount of such deposits as of October 31, 1982. The adjustment was triggered by changes in deposit balances, not numbers of accounts, and did not distinguish between accounts which existed on the October 31, 1982, measuring date and those which were opened thereafter. The purchase agreement required NCNB to use reasonable efforts to preserve for Peoples the banking business and goodwill of existing customers at the NCNB branches.
On December 20, 1982, Peoples submitted to the FDIC its application for approval of its acquisition of the assets and assumption of the liabilities of the NCNB branches. For financial statement purposes, Peoples indicated in the application that $4,340,000 of the premium paid for the NCNB branches would be allocated to the value of core deposits. 9 Peoples did not submit copies of either valuation report described below to the FDIC, but only a letter with preliminary estimates prepared by Morgan. The FDIC approved the purchase of the NCNB branches on February 7, 1983.
9 In the December 12, 1982, letter, Morgan estimated the value of the core deposit intangible for the NCNB branches to be $4.34 million for purposes of the FDIC requirements, which excluded all time certificates.
On March 25, 1983, Peoples purchased the assets and assumed the liabilities (except the individual retirement accounts) of the NCNB branches. Core deposit intangibles were not identified as a separate category of assets in the closing documents in the record. There was no adjustment to the purchase price for deposit variation.
Before the closing, NCNB monitored deposit balances at the NCNB branches in connection with the price adjustment clause. Overall balances of deposits increased by about 6.7 percent during the period between the October 31, 1982, measuring date and the March 25, 1983, closing. However, demand deposits and savings account balances decreased, while the balances of money market and NOW accounts, which had then become available because of deregulation, increased by over $6 million.
During May of 1983, Peoples was offering rates ranging from 8.14 to 9.25 percent on certificates of deposit with maturities from 91 days to 3 1/2 years. The maximum maturity of certificates of deposit offered by Peoples at that time was 3 1/2 years; certificates of deposit with maturities of 5 years were rare and not used as a significant source of funding. For the week ending March 25, 1983, the "Federal Reserve Bulletin" reflected rates in the secondary market ranging from 8.79 to 9.08 percent for certificates of deposit with maturities of 1 to 6 months. Comparable rates in the national market ranged from 8.20 to 9.20 percent.
At the time of the NCNB acquisition, petitioners relied upon deposits, capital notes, Federal funds, and equity as funding sources at a cost of capital of about 14 percent.
After the closing, the banking business conducted at the eight NCNB branches acquired continued without change using most of the same preacquisition employees. 10 There was no interruption of the banking business conducted at the NCNB branches during 1983 through 1985. There was no unusual or abnormal run-off of deposit accounts or deposit dollars at least during 1983. NCNB continued to operate in both Wilmington and Jacksonville after the closing.
10 Peoples advertised that "WE'VE CHANGED THE NAME, BUT THE SMILES ARE THE SAME" and that customers would "still receive the same great service from the same friendly people." The NCNB manager in Jacksonville stayed with NCNB and the BNC manager in Wilmington left Peoples within a year.
Many of the customers of the NCNB branches had more than one type of account (for example, demand deposit and money market accounts). A significant number of the loans of the NCNB branches were made to deposit customers of the NCNB branches.
On the FDIC application, Peoples indicated that after the purchase it would have only 8.8 and 6.2 percent of the total bank deposits in the Wilmington and Jacksonville markets, respectively. Peoples listed six bank competitors with two to six branches each in the Wilmington market and three bank competitors with two to seven branches each in the Jacksonville market. Petitioners also reported that there was competition from four savings and loan associations, four credit unions, and 12 finance companies in the Wilmington market, and two savings and loan associations, one credit union, and seven finance companies in the Jacksonville market.
Under North Carolina law, Peoples could have opened new branches in the Wilmington and Jacksonville markets, but considered this to be more expensive and a slower way to build a "deposit base" and had never taken steps to do so.
For a bank the size of Peoples, industry data from the 1982 Federal Reserve System "Functional Cost Analysis" survey indicates that 44 percent of the balances are held in slightly less than .4 percent of the accounts.
B. Expert Valuation Evidence
Morgan was a principal and director of Golembe, a consulting firm specializing in part in banking, financial, and economic issues for banks, at the time he was retained by Peoples to value the acquired deposits of the NCNB branches. He has bachelor's, master's, and doctoral degrees in economics from Yale University. His doctoral thesis and postgraduate work were in financial theory and macroeconomics, and his doctoral dissertation was on the cost of capital. Morgan has taught on the faculties of Yale University, the University of Illinois, and the Wharton School of Finance and Commerce at the University of Pennsylvania.
Morgan was retained to estimate value coefficients, a value per dollar for each type of deposit (e.g., demand deposits), which could be multiplied by deposit balances to provide values for each. His valuation method used the present value of the future earnings from the deposits. The method involved estimating and combining four components (i) the accounts and associated balances that were purchased, (ii) the attrition patterns of the accounts, (iii) the after-tax net earnings relating thereto, and (iv) the capitalization rate by which to discount the earnings.
First, Morgan estimated the deposit balances acquired. He included all deposits, except certificates of deposit of $100,000 or more, for this purpose.
Second, he computed attrition patterns and survival rates for the deposits. 11 He examined all account closings for the period from January through May of 1983 for the former BNC branches, and February through May of 1983 for the former NCNB branches, as provided from data prepared by Peoples employees. These closed accounts were narrowed to severances where the customer neither had nor promptly opened another account. The accounts were stratified by age, but not balance size; business and personal accounts were considered separately 12 for the Wilmington branches, but not for the Jacksonville branches because they involved too few accounts. The data for the three customer groups (personal and business accounts for Wilmington and all accounts for Jacksonville) was used to formulate survival rates for each, which represented the proportions of the acquired deposits projected to remain at the end of each calendar year. The survival rates were based on patterns for demand deposit accounts only. In deriving the survival rates, Morgan assumed that existing deposits would leave Peoples in the future at the same rate that they had terminated in the past. Using this information, he computed survival coefficients for the personal and business accounts for Wilmington and all accounts for Jacksonville.
11 Morgan did not make a specific adjustment to the historic pattern of closings to reflect the effect of deregulation, and assumed that neither competition, economic changes, nor changes in management would cause a significant change in the pattern of closings from those observed in the period during which the data was collected.
12 Personal and business accounts were considered separately where the data was available because different factors affect the respective closings of such accounts. For personal customers, such factors include relocations to other communities, job changes, marriages and divorces, and mortality. For business customers, such factors include business reorganizations and failures, and retirement of owners and key personnel.
Third, the net earnings margins on the deposits were estimated and applied to remaining average annual balances (i.e., deposit balances combined with survival rates) to obtain the projected future income stream for the deposits. For this purpose, the valuation report assumed a hypothetical bank which does not establish new accounts, but only services existing customers and invests those deposits. Morgan indicated that the net after-tax 13 earnings margins on deposits have generally averaged about 1.124 percent 14 per year in commercial banking. He adjusted the 1.124 percent average to 1.13 percent to reflect the actual composition of the deposits of the NCNB branches by comparing the weighted average margin of their actual deposit mix with that of comparably sized banks, using data from the 1981 Federal Reserve System "Functional Cost Analysis" survey. 15 Morgan then assigned the 1.13 percent weighted average net after-tax earnings margins to the following four types of accounts included in the deposit mix of the NCNB branches: demand deposit (DDA); NOW; savings (SAV); and other time deposits (OTD), excluding certificates of deposit of $100,000 or more. Morgan then computed a value coefficient for each type of account included in each of the three customer groups by multiplying the net after-tax earnings margin of the account by the survival coefficient for the customer group.
13 For this purpose, Morgan assumed a tax rate of 46 percent.
14 In reaching this initial net after-tax earnings margin, Morgan excluded data from 1979 forward, which would lead to higher margins, on the basis that this was not a clear long-term trend.
15 The "Functional Cost Analysis" computes average earnings margins which are the net profit which the average bank earns from assets funded by deposits. For this purpose, it is assumed that the deposits are invested in an average bank portfolio.
Fourth, to determine the present value of the earnings stream Morgan used a capitalization rate of 4 percent based on the average yields on tax-exempt State and local government bonds for the period from 1959 through 1979 and adjusted to eliminate the effect of an estimated inflation rate.
Finally, Morgan estimated amortization patterns based on the discounted value at each year-end of the remaining projected earnings stream from the acquired core deposits over a 20-year life. The percentage of the original value thereof expected to remain at the end of each of the years 1983 through 1985 was estimated as reflected in Table 1: 16
16 Although Morgan's valuation report does not make it clear which of two tables of amortization patterns is the appropriate one, respondent did not object to the above information propounded by petitioners based upon one of the tables.
The Golembe value coefficients, 17 multiplied by the acquisition date deposit balances determined by Price Waterhouse representatives, resulted in the initial values as shown in Table 2:
17 In a July 15, 1983, letter, Morgan indicated preliminary values of the NCNB branch deposits of $4,234,798 including the tax benefits of amortization and $2,591,808 without considering the benefits to Peoples. He explained that the values were below average because of an unusually rapid attrition pattern, mainly in the Jacksonville branches. He also stated that the values from applying the value coefficients were likely to be understated because the data used probably omitted some customer accounts.
Hamrick advised Peoples to use for tax purposes the preliminary $4,820,000 value determined by Morgan and included in the purchase agreement. Price Waterhouse allocated this amount by grossing up the values shown in Table 2. 18
18 A September 19, 1983, memorandum by an employee of Price Waterhouse states that the formula for increasing these values is based on the ratio of the $4,820,000 initial value to the $3,728,003 value. However, the percentage used for the gross-up does not equal that ratio.
Dr. Thomas E. Doerfler (Doerfler) was the principal statistician for Arthur D. Little at the time he prepared his study to determine the remaining useful life of the acquired deposits of the NCNB branches. Doerfler has a bachelor's degree in mathematics from the University of Dayton and master's and doctorate degrees in statistics from Iowa State University.
Doerfler relied on data relating to existing accounts and individual account severances 19 provided by Peoples employees. The data for the Wilmington branches covered periods of 3 to 8 months, and for the Jacksonville branches 2 to 5 months, in 1983. 20 He analyzed separately for each market five types of accounts: demand deposit accounts for personal customers (DDA≈personal), demand deposit accounts for business customers (DDA≈business), interest bearing accounts, money market accounts, and regular savings accounts.
19 According to Doerfler, a severance involves a customer closing all accounts with the bank, not just one account. Severance data was used for all accounts except savings accounts, for which closure data was used because severance data was unavailable.
20 As a result of the limited period of observing closings, older accounts, which close at slower rates than newer accounts, tend to be underrepresented in the experience band collected.
From the closure data, Doerfler estimated the conditional probability of existing accounts remaining open, which he used to develop a cumulative probability or "survival curve" (of percentages of accounts of certain ages remaining open), and finally, a model and arithmetic averages for predicting the remaining life of existing accounts given their respective ages at acquisition. In developing the survival curve and model, Doerfler assumed that the existing accounts would leave Peoples in the future at the same rate that they had in the past. He predicted the remaining life of each of the types of accounts at the Wilmington and Jacksonville branches, respectively, and the number of accounts of each type which would remain open as of yearend 1983 through 2000; Table 3 shows for each type the remaining mean and median life, and the number predicted to remain open as of the end of the years 1983 through 1985.
Doerfler did not make any assumptions regarding management in his valuation, and did not consider the effect of deregulation of the industry. He did not determine a life for individual accounts.
Doerfler stated that the determination of useful life is a statistical estimation task which involves making inferences from data; and as such, it is important that all aspects of data collection, analysis, and interpretation be carried out in accordance with sound statistical principles. He acknowledged that extrapolations of data for periods of nearly 20 years cannot be performed with great precision. Doerfler computed confidence intervals for his point estimates and calculated a margin of error in his results, but did not include them in his report. Although he testified that it is recommended that the measurement of linearity of the data be quantified, Doerfler did not make such a measurement.
Robert J. Curtis (Curtis) was an executive senior manager with Ernst & Young at the time he prepared his report valuing the acquired deposits of the NCNB branches. He has a B.B.A. in accounting and economics from the University of Miami and a master's degree in accounting and finance from the Wharton School of Finance and Commerce at the University of Pennsylvania. Before joining Ernst & Young, he was a vice president of valuation services at Arthur D. Little Valuation, Inc. and a director of financial valuation at American Valuation Consultants.
Curtis used a cost savings approach, which estimated the value of the deposit accounts as the present value of the difference between the ongoing cost of servicing the deposits and the cost of the market alternative source of funds. Under this approach, the deposits are valued as a source of favorable financing to the purchaser or a below market liability. The underlying theory is that by obtaining these deposits banks reduce their cost of funds. He applied this approach to four major types of accounts, demand deposit accounts, NOW accounts, money market accounts, and regular savings accounts. The demand deposit accounts were stratified into personal and business accounts. Certificates of deposit were excluded because they have little value. The total beginning average balances of the acquired accounts from data furnished by Peoples were as shown in Table 4:
Curtis projected the net available balances of acquired deposit accounts in future years by multiplying the accounts predicted by Doerfler to survive in each age group by the acquisition date average balance 21 thereof, adjusted for inflation at an annual rate of 5 percent. The resulting amounts were then reduced for reserve requirements and float 22 to determine net investable funds.
21 Curtis calculated the average balance by dividing the number of accounts into the aggregate balance of the respective account age groups.
22 Reserve requirements are the balances required to be held against deposits at the Federal Reserve; float is an allowance for the time between when deposits are made and when they are collected.
Curtis then computed the cost savings derived from the net available balances by first determining the cost of the alternative source of funds, which in his opinion was insured certificates of deposit of comparable balance size and maturity (5 years or more). He stated that the interest cost of such comparable certificates of deposit for the five types of accounts ranged from 10.20 to 11.20 percent. Based on information in the 1983 Federal Reserve System "Functional Cost Analysis" survey (1983 FCA), he estimated that the cost of servicing and other associated expenses would be .22 percent of the funds.
Second, he projected the ongoing cost (interest based on rates paid by Peoples, and service expenses and deposit fee income, such as nonsufficient funds charges, based on 1983 FCA data) of the deposits which he valued (the total deposit cost). 23 He then subtracted the total deposit cost from the cost of alternative funds to reach pretax cost savings. 24 Assuming a tax rate of 20 percent, 25 Curtis converted the pretax cost savings to an after-tax cost savings.
23 The service expense and fee income percentages for money market accounts were assumed to be the same as for NOW accounts.
24 At trial, Curtis acknowledged that cost savings and thus value would decline with a shift of customers from demand deposits to NOW or money market accounts. For purposes of his valuation, he assumed that there would be no change in the deposit mix during the projected lives of the deposits. He also testified that the 1982 Federal Reserve System "Functional Cost Analysis" survey reflected an average earnings margin for commercial checking accounts for banks comparably sized to Peoples of 5.234 percent, as compared to his estimate that cost savings associated with such accounts would be 8.41 percent.
25 Curtis' report states he used this rate because the historical effective tax rate reflected in "Survey of U.S. Effective Income Tax Rates for the Banking Industry" published by the Bank Administrative Institute was less than 20 percent.
Curtis next discounted the cost savings stream to present value. He chose a 16-percent discount rate by estimating the weighted average cost of capital (the pretax cost of debt based on average yields on long-term medium grade (Baa) corporate bonds, which was tax-affected, and the estimated cost of equity, assuming a debt-equity ratio of 22:78, which he said was that of a "hypothetical typical buyer" bank). The resulting total discounted net cost savings value of each type account in the Wilmington and Jacksonville branches, respectively, was as shown in Table 5:
1 Totals rounded to the nearest thousand dollars.
Finally, Curtis estimated the amortization method based on net discounted after-tax cost savings for the acquired deposits for each year over an 18-year life; the results for the years 1983 through 1985 are shown in Table 6:
d. Tracking System
Price Waterhouse created an account tracking system to monitor the loss of the accounts acquired in the purchase of the NCNB branches. The tracking system was designed to search the relevant customer information file or CIF (a computer file containing the account number, customer name, and other data for each account of the NCNB branches) and identify closed accounts. The closed accounts were identified by account number, customer name, and value assigned to each as of the acquisition date. An account was not considered closed if the customer opened another account. The results of the tracking system for the period when it was in place, 1984 through 1986, were as follows:
The system was discontinued after 1986.
Edward J. Kane (Kane) was a professor at Ohio State University who held the Reese chair of banking and monetary economics at the time he prepared his report. Kane has a bachelor's degree from Georgetown University and a doctorate degree from the Massachusetts Institute of Technology, where he majored in economics and statistics and minored in mathematics. He has studied deposit interest rates, written a textbook on econometrics, and authored or coauthored several other textbooks. He has published dozens of articles regarding bank regulation and several papers regarding bank valuation and pricing in the United States and Japan.
Most of Kane's report and testimony were restricted to specific criticism of parts of the methodology of petitioners' expert witnesses, which is addressed in the opinion, infra.
III. County Bank Acquisition
County Bank and Trust Company of Morehead City, North Carolina (County Bank), was incorporated in North Carolina on October 17, 1978, and opened for business on January 2, 1979. County bank's primary market area was Carteret County, North Carolina. Commercial activity in Carteret County is seasonal because of the predominance of the summer resort business, and deposits tend to increase during the summer.
County Bank was a full service, retail bank offering the normal types of banking services, with three offices located in Morehead City, Beaufort, and Atlantic Beach, North Carolina. The main office was located in Morehead City; the Beaufort and Atlantic Beach offices were located in temporary mobile homes.
County Bank operated at a deficit until 1983 when a profit of $29,624 was earned. The 1983 operating results were distorted because County Bank had not paid the chief executive officer's salary in December (as a result of his resignation) and certain normal operating expenses were booked as prepaid expenses.
On January 9, 1984, the North Carolina State Banking Commission completed an examination of County Bank. On February 2, 1984, representatives of the FDIC and the North Carolina commissioner of banks met with County Bank's board of directors to discuss the results of the examination. A representative of the commissioner of banks reported that the overall financial condition of County Bank was unsatisfactory, noting inexperienced senior management, a serious liquidity problem, unacceptable investment policies and loan management, a declining trend in the ratio of capital to total assets, very low growth in deposits compared to loans, insufficient deposits to support an adequate earning asset base, and various violations of law and regulations.
As of March 5, 1984, County Bank, the North Carolina commissioner of banks, and the FDIC entered into a memorandum of understanding which required County Bank to undertake certain steps including hiring a competent chief executive officer, reducing its loan to deposit ratio, establishing loan and investment policies, and notifying the FDIC and the commissioner (who had the right to object) if County Bank's use of "brokered deposits" 26 reached 5 percent of total deposits.
26 Brokered deposits are deposits funded by third party agents or nominees for depositors, including deposits managed by a trustee or custodian when the individual beneficial interest is entitled to or asserts a right to Federal deposit insurance.
Faced with the cost of complying with the memorandum of understanding, County Bank decided to find a buyer or another bank with which to merge. County Bank was asking $13 per share and wanted to obtain at least $12 per share inasmuch as its directors and original shareholders paid $11 per share.
By the spring of 1984, Calvin Wellons (Wellons), who founded County Bank and was acting as chairman of the board of directors and chief executive officer, contacted Robert Mauldin, who was then president and chief executive officer of Peoples, about buying County Bank. Wellons was asking about $13 per share. Peoples was aware of County Bank's management and liquidity problems, and offered $11 per share.
Peoples did not hire any outside advisers in evaluating the purchase price of County Bank. Peoples prepared projections of the effect of purchase prices ranging from $7.83 to $12 per share on earnings and capital. In projections assuming prices of $11 and $11.14 per share, $389,430 of the total purchase price was allocated to deposits and $254,570 and $271,051, respectively, was allocated to goodwill. In another projection assuming a price of $11 per share, $624,000 was allocated to deposits and $20,000 to goodwill. In these projections, deposits were amortized over 10 years and goodwill over 25 years, both on a straight-line method.
Peoples considered Carteret County to be an attractive banking market because demographic studies indicated good growth in population and family median income, and other favorable economic indicators existed. Peoples considered County Bank not to be well managed. Peoples had higher lending limits than County Bank, and expected to improve management of the operation and customer service and satisfaction, and to offer additional services to customers. As of June 1, 1984, County Bank had 19 full-time and 5 part-time employees.
On June 5, 1984, Peoples and County Bank entered into an Agreement and Plan of Merger. The purchase price was $11 cash per share for the 120,000 shares of common stock outstanding, for a total price of $1,320,000. The amount paid by Peoples in excess of the book value of the County Bank assets (other than goodwill, going-concern value, and core deposits) was $672,363.21. The board of directors of Bancorporation and the shareholders of Peoples approved the merger agreement on June 20, 1984.
The Statement of Condition attached to the June 25, 1984, application submitted to the FDIC for approval of the merger, showed debit adjustments of $625,000 for the "Value of core deposits" and $20,000 for "Goodwill charge off". Peoples did not submit copies of either valuation report described below with the application. The merger of County Bank into Peoples was approved by the County Bank shareholders, the North Carolina commissioner of banks, and the FDIC in August 1984.
After the close of business on Friday, September 28, 1984, County Bank merged into Peoples. On Monday, October 1, 1984, the former County Bank branches opened for business as offices of Peoples. All of the County Bank employees were retained. Peoples acquired all of the assets and assumed all of the liabilities of the County Bank pursuant to the merger.
As of September 27, 1984, Peoples was offering rates ranging from 9.25 to 11 percent on certificates of deposit with maturities ranging from 31 days to 4 years. 27 Certificates of deposit of 5 years or more were not a significant source of funding at that time. Rates for the week ending September 28, 1984, quoted in the Federal Reserve Bulletin ranged from 10.90 to 11.40 percent for certificates of deposit with maturities ranging from 1 to 6 months.
27 An indication of petitioners' cost of capital is an amendment to a registration statement issued on July 11, 1985, whereby petitioners offered $13 million in 20-year debentures convertible into common stock at a rate of 8 3/4 percent. The prospectus attached to the amendment also shows that for the years 1980 through 1984, the tax rates for petitioners were at most 22 percent.
At the time of the County Bank acquisition, petitioners relied upon deposits, capital notes, Federal funds, and equity as funding sources at a cost of capital of about 15 percent.
On the FDIC application, Peoples listed three bank competitors each with 3 to 10 branches in Carteret County. Peoples also indicated that there was competition from three savings and loan associations, one credit union, and three finance companies in the Carteret County market.
There was no interruption in the banking business conducted at the locations of County Bank during 1984 and 1985. There was no unusual or abnormal run-off of deposit accounts or deposit dollars for County Bank in 1985. No attempt was made to monitor the closing of County Bank accounts existing at the time of the merger. In 1986, petitioners relocated the Atlantic Beach branch; in 1988, petitioners closed the Beaufort branch.
B. Expert Evidence
In November 1984, a representative of Peoples met with Dr. John Danforth (Danforth), an economist with Golembe, about valuing the deposits acquired in the County Bank merger. Morgan supervised Danforth in preparing the valuation.
Danforth used the same valuation method used by Morgan to value the deposits acquired from the NCNB branches. The amount of acquired deposits, excluding certificates of deposit of $100,000 or more, was $7,823,000. To estimate the life of the deposits, Danforth obtained from Peoples computer printouts reflecting a survey of demand deposit and NOW account closings, along with related opening dates, closing dates, and average balances, for County Bank for the first 11 months of 1984. This data was stratified by age of account, and there was an adjustment because of the absence of older accounts 28 based on data from the Wilmington NCNB branches. 29 From the data, Danforth projected the percentages of acquired deposits which could be expected to remain as of September 30 of the years 1985 through 2004.
28 The absence of older accounts was because County Bank was a relatively new bank.
29 A document prepared by Golembe entitled "Appraisal Study Project Data Sheet" has a handwritten notation that the Wilmington NCNB branches are similar to County Bank's offices in Morehead City in terms of type of customer base; no other explanation is provided.
Danforth adjusted the average net after-tax earnings margin of 1.12 percent to reflect the composite mix of the acquired County Bank accounts as compared with those shown in 1983 FCA data for banks with deposits of less than $50 million, to an average after-tax earnings margin of 1.2564 percent. This margin was multiplied by the average deposits in the accounts (taken from the beginning and the end of each year) to result in a projected after-tax earnings stream which was discounted to a value of $1,053,075 using a capitalization rate of 4 percent, based on the after-tax rate on State and local government bonds.
Finally, the amortization pattern was estimated from the discounted value of the remaining earnings at each year-end over a 20-year life; the percentages remaining at the end of the years 1984 and 1985 were 97.62 percent and 88.25 percent, respectively.
Curtis used the same cost savings methodology to value the acquired County Bank deposits which he used to value those of the NCNB branches. For this purpose, he considered demand deposit accounts (DDA), regular savings accounts, and money market accounts separately, but did not value any certificates of deposit. The total beginning average balances of the acquired County Bank accounts were as shown in Table 7:
Curtis multiplied the account survival percentages determined by Danforth by the average account balances, adjusted them for 5-percent annual inflation, and reduced them for reserve requirements and float to estimate net investable funds. In computing the cost of alternate funds, he assumed an 11.4 percent interest expense for the alternate funds (using interest rates on certificates of deposit for less than $100,000 and a .22-percent rate for service and other expenses (based on data in the 1984 Federal Reserve System "Functional Cost Analysis" survey (1984 FCA)). He projected the ongoing cost (interest based on rates paid by Peoples, and service expenses and deposit fee income based on 1984 FCA data) of the deposits which he valued (the total deposit cost). He subtracted the total deposit cost from the cost of alternate funds, and then converted this to a net after-tax cost, again using a 20-percent tax rate.
Using a 17-percent discount rate which he stated was based on "prevailing rates at the time of the acquisition", Curtis next discounted the cost savings stream to present value as shown in Table 8:
1 Total rounded to the nearest thousand dollars.
The amortization method was again based on net discounted after-tax cost savings for the acquired deposits for each year; the results as of the end of the years 1984 and 1985 are reflected in Table 9:
As with the NCNB evidence, Kane, respondent's expert witness, restricted his opinions to criticism of petitioners' experts, which is discussed in the opinion, infra.
IV. Financial Accounting, Regulatory, and Tax Treatment
Petitioners, NCNB, and County Bank expended and deducted amounts for advertising, marketing and business development, promotion, salaries, and other expenses relating to attracting and retaining customers during the taxable years before the acquisitions.
Peoples did not value deposit accounts as an asset in its ongoing business, and would not do so absent a sale. For financial statement purposes, Peoples does not reflect postacquisition deposits separately from deposits acquired.
For tax purposes, $4,820,000 of the purchase price for the NCNB branches was allocated to "core deposit" and $1,534,698.17, to goodwill, 30 and $672,636.21 of the purchase price for the County Bank was allocated to "Deposit Intangible" and no amount to goodwill. 31 For regulatory and financial statement purposes, the amounts allocated to core deposit intangibles were amortized over a 10-year life using the straight-line method, as contrasted with amortization deductions over a 20-year life using an accelerated method for tax purposes.
30 The parties stipulated that this allocation applied for regulatory and financial statement purposes as well; however, the pro forma balance sheets attached to the FDIC application for merger record "core deposits" as $4,340,000 and goodwill as $1,626,000. Additionally, in a December 12, 1982, letter, Morgan indicated that the FDIC excluded all time certificates from core deposits, and on that basis estimated a value thereof of $4.34 million as contrasted with $4,820,000 if only certificates of deposit of $100,000 or more were excluded. In view of this, we assume that the stipulation is in error. Kirchner, Moore & Co. v. Commissioner [Dec. 30,098], 54 T.C. 940 (1970), affd. [71-2 USTC ╤ 9669] 448 F.2d 1281 (10th Cir. 1971).
31 The parties stipulated that this allocation also applied for regulatory and financial statement purposes, but the Statement of Condition attached to the FDIC application for merger reflects a debit adjustment of $625,000 to "Value of `core' deposits" and a debit adjustment of $20,000 to "Goodwill charge off". Thus, the stipulation appears to be in error. Kirchner, Moore & Co. v. Commissioner, id.
Respondent has not raised an issue regarding petitioners' right to step up the basis of the assets of the County Bank.
In the statutory notice of deficiency, respondent disallowed the deductions claimed by petitioners for amortization of the amounts allocated to core deposit intangible in the amounts of $636,562.14, $664,073.80, and $622,336.99 for the tax years 1983, 1984, and 1985, respectively. 32 Additionally, the parties stipulated that petitioners incurred $188,862 and $35,298 in costs relating to the acquisition of the NCNB branches and County Bank, respectively, which the parties agreed should be allocated to the assets acquired in proportion to their fair market values, resulting in allocations of $49,882 and $10,120 to land and to depreciable tangible assets, respectively.
32 Petitioners conceded that under the amortization method which they used they should have claimed amortization deductions in the amounts of $527,835 and $655,035 for the tax years 1983 and 1984, respectively, rather than the amounts claimed.
Petitioners contend that the core deposits acquired are amortizable because they are separate from goodwill, petitioners established a reasonable estimate of their value and useful life, and the amortization method used is reasonable. Respondent maintains that the deposits are inseparable from goodwill and are part of the banks' going-concern value, and that petitioners failed to establish a limited life and value with reasonable accuracy and to show that its amortization method is reasonable.
We agree with petitioners that some of the deposits are amortizable because they are separable from goodwill and going concern, but based on all of the evidence we have reached our own conclusions as to which deposits these are, their value, their life, and the proper amortization method.
Section 167(a) allows as a depreciation deduction a reasonable allowance for the exhaustion of property used in a trade or business. Pursuant to the rules stated in section 1.167(a)-3, Income Tax Regs., intangibles, such as core deposits, qualify for a depreciation allowance if they have both an ascertainable value, separate and distinct from the goodwill and going-concern value of the business acquired, and a limited useful life, the duration of which can be established with reasonable accuracy. Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. 463, 479 (1988), affd. without published opinion 900 F.2d 266 (11th Cir. 1990); see Southern Bancorporation, Inc. v. Commissioner [88-1 USTC ╤ 9344], 847 F.2d 131, 136-137 (4th Cir. 1988), affg. [Dec. 43,562(M)] T.C. Memo. 1986-601. Whether these requirements are met is essentially a factual question. Citizens & Southern Corp. v. Commissioner, supra at 479.
At the outset, it is appropriate for us to define which accounts we consider to be "core deposits". In IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. 496, 516 (1991), we emphasized that interest rate insensitivity is an important feature of core deposits (as found in Citizens & Southern Corp. v. Commissioner, supra at 465). We recognized that "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". IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 517. We concluded that such accounts were designed to be interest rate sensitive, and thus would not be treated as core deposits absent proof that they were in fact insensitive to interest changes under the facts of the case.
Petitioners bear the burden of showing which deposits are truly core deposits (Rule 142(a); Welch v. Helvering [3 USTC ╤ 1164], 290 U.S. 111 (1933)) and have not shown that any of the certificates of deposit, money market accounts, or NOW accounts were interest rate insensitive. The record reflects that, as reported to the FDIC, Peoples had significant competition in the relevant markets involved in the NCNB branch and County Bank acquisitions, which undoubtedly would have put competitive pressure on the rates for adjustable rate deposit accounts. IT&S of Iowa, Inc. v. Commissioner, supra at 518. The record does not enable us to determine the extent to which the NOW accounts offered were not sensitive to rate changes. Thus, we conclude that petitioners have failed to show that any accounts other than the demand deposit and savings accounts were in fact core deposits.
In presenting their respective positions concerning the separate existence, the value, the life, and the amortization method regarding the core deposits, both parties have relied on expert opinions and testimony. We are not bound by the opinion of an expert witness, and may reject all or a portion of an expert's opinion which is contrary to our sound judgment. Parker v. Commissioner [Dec. 42,966], 86 T.C. 547, 561 (1986).
II. Ascertainable Value Separate and Distinct from Goodwill and Going Concern
Goodwill is "the expectancy of continued patronage". Pridemark, Inc. v. Commissioner [65-1 USTC ╤ 9388], 345 F.2d 35, 44 (4th Cir. 1965), affg. in part and revg. in part [Dec. 26,840] 42 T.C. 510 (1964), overruled on another issue by Of Course, Inc. v. Commissioner [74-2 USTC ╤ 9546], 499 F.2d 754 (4th Cir. 1974), revg. [Dec. 31,581] 59 T.C. 146 (1974) (quoting Boe v. Commissioner [62-2 USTC ╤ 9699], 307 F.2d 339, 343 (9th Cir. 1962), affg. [Dec. 24,641] 35 T.C. 720 (1961)). Going-concern value has been distinguished from goodwill as "the additional element of value which attaches to property by reason of its existence as an integral part of a going concern", and "the ability of a business to continue to function and generate income without interruption as a consequence of the change in ownership". VGS Corp. v. Commissioner [Dec. 34,519], 68 T.C. 563, 591-592 (1977). Neither goodwill nor going-concern value of a business is subject to an allowance for depreciation. Computing & Software, Inc. v. Commissioner [Dec. 33,197], 64 T.C. 223, 232, n.7 (1975); sec. 1.167(a)-3, Income Tax Regs.
We have held that core deposits may be proven to have an ascertainable value separate and distinct from the goodwill and going-concern value of the bank acquired. Citizens & Southern Corp. v. Commissioner, supra; IT&S of Iowa, Inc. v. Commissioner, supra. As we stated in Citizens & Southern:
The evidence in the instant case establishes that the acquisition of core deposits was the primary reason petitioner purchased the Acquired Banks and that petitioner paid a premium in order to obtain the core deposits. * * * These core deposits are a low-cost source of funds and are an important factor contributing to the profitability of a commercial bank. * * * the value of deposit base does not depend upon a vague hope that customers will patronize the bank for some unspecified length of time in the future. * * * The value of deposit base rests upon the ascertainable probability that inertia will cause depositors to leave their funds on deposit for predictable periods of time. * * * [91 T.C. at 498-500; fn. refs. omitted.]
We analogized these core deposits to the agreements to service existing mortgages acquired with a loan servicing business, which agreements were cancelable upon written notice and were conceded to be amortizable, and the right to use escrow funds associated therewith, which we held to be amortizable, involved in First Pennsylvania Banking & Trust Co. v. Commissioner [Dec. 30,859], 56 T.C. 677 (1971). Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. at 488-490. We emphasized that:
The lives of the escrows, therefore, depended upon the lenders not withdrawing their business, just as the life of deposit base depends upon depositors not closing their accounts. The escrows were purchased in a transaction in which First Pennsylvania acquired a going concern which expected to obtain new servicing and new escrows in the future. Similarly, the deposit base was acquired in the purchase of going concerns which expected to obtain new deposit accounts in the future. Our holding that the opportunity to use the existing escrow funds was an amortizable intangible asset supports petitioner's position that deposit base is an asset separate and distinct from goodwill. [91 T.C. at 489-490.]
Core deposits are an essential part of the lifeblood of the commercial bank enterprise. Such deposits constitute a significant portion of the inventory the bank draws upon to fund its loans and investments that generate its profits≈its taxable income. Consequently, core deposits are an essential ingredient of a bank's functional life. Without them, the bank could not function beyond its paid-in capital and high cost funds which would greatly diminish its profit potential or vitality. It follows that when one bank considers acquiring another bank, core deposits are a target that the acquiring bank focuses its sights on. Core deposits are a significant, distinct property with an inherent value that provides an inexpensive means for generating the business of the bank≈loans. Those deposits differ from goodwill due to their foreseeable turnover or useful life and their unique value to the function of the bank. Accordingly, core deposits are the means by which an acquiring bank gauges its assurance of low cost funds to prognosticate profitability in order to recover and form capital for a predictable period.
In the acquisition of the NCNB branches, Peoples did not acquire the NCNB name or reputation and NCNB remained a competitor in the markets. Additionally, NCNB recognized the core deposits as a separate element of value when it estimated an appropriate selling price for the branches. We recognize that Peoples did acquire most of an existing work force and established, existing locations; however, we consider the core deposits to be separate and distinct from these intangibles. Absent the element of inertia associated with the core deposits, we believe that business might very well have followed the NCNB reputation and name.
The record regarding the County Bank acquisition is less complete and convincing. However, we conclude that there was no substantial goodwill or going-concern value at the time of the acquisition, and that to the extent there were any such assets the core deposits had a separate and distinct value. At that time, County Bank was operating under a memorandum of understanding with Federal and State regulators because of liquidity problems, unacceptable investment and loan management, a declining trend in the ratio of capital to total assets, lack of experienced management, and various violations of law and regulations. Additionally, two of the offices were in temporary facilities. Finally, the price at which the merger was consummated was the same as that paid by the original shareholders.
This case is factually distinguishable from Newark Morning Ledger Co. v. United States [91-2 USTC ╤ 50,451], 945 F.2d 555 (3d Cir. 1991), cert. granted ___ U.S. ___ (April 6, 1992), in which the Third Circuit concluded that the taxpayer failed to show that the at-will newspaper subscriber relationships were an asset separate and distinct from goodwill. In Newark Morning Ledger, the court emphasized that the newspaper trade name was acquired too, and that the taxpayer used the income projection methodology to value the asset. Id. at 562, 568. Neither the NCNB branches nor the County Bank were operated under their prior names after they were acquired by Peoples. Peoples presented both the income projection methodology and the cost savings methodology for valuation purposes here. However, the Third Circuit has indicated disagreement with our core deposit line of cases, stating that they represent a minority view. Newark Morning Ledger Co. v. United States, supra at 564 n.7a, 565. In this regard we follow our Court-reviewed opinion in Citizens & Southern Corp. v. Commissioner, supra, which was affirmed by the Eleventh Circuit.
Based on these facts, we conclude that petitioners established that the core deposits acquired in both the NCNB branch and County Bank acquisitions were separate and distinct from goodwill or going-concern value.
Respondent contends that the value of the core deposits presented by petitioners is the value (future profits) from continued customer patronage, which is an inseparable component of goodwill. In support thereof, respondent cites Southern Bancorporation, Inc. v. Commissioner [88-1 USTC ╤ 9344], 847 F.2d 131, 138 n.8 (4th Cir. 1988), which concluded that the taxpayer had not established a useful life for the deposits and stated also:
Given the very nature of the deposit relationship as an important point of contact with customers for selling a bank's other income producing services, we question the ultimate success of any attempt to separate the value of the deposit base from goodwill. * * * While some taxpayers find it difficult to accept, there are some things that may not be subject to amortization or depreciation. [Citation omitted.]
This statement by the Fourth Circuit, to which this case is appealable, is dicta, and thus is not dispositive of the issue under Golsen v. Commissioner [71-2 USTC ╤ 9497], 445 F.2d 985 (10th Cir. 1971), affg. [Dec. 30,049] 54 T.C. 742 (1970). We, therefore, follow our own recent Court-reviewed opinion regarding this issue, Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. 463 (1988); see also IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. 496 (1991). In IT&S of Iowa, Inc. v. Commissioner, supra at 509-510, we rejected the same contention respondent makes here on the basis that the taxpayer's cost savings methodology did not include the value of the sale of all banking services to the bank's established customers, but only the separate value of the existing core deposits. The same reasoning is true here if the cost savings methodology of Curtis is used to measure value. Future profits would only be the measure of value if we were to accept the Golembe methodology, and as explained below, we do not.
In making this contention, respondent also cites Dodge Brothers, Inc. v. United States [41-1 USTC ╤ 9309], 118 F.2d 95 (4th Cir. 1941), in which the Fourth Circuit concluded that the taxpayer failed to show that the model and design of a "proved" car, which had been sold to the public in the same form for many years, was a depreciable intangible asset separate from goodwill. Specifically, respondent refers to the Fourth Circuit's statement, id. at 101, that: "A valuation based upon the capitalization of the average of a company's earnings over a number of years smacks peculiarly of a method used in evaluating good will." However, we continue to believe that core deposits are more properly analogized to the escrow funds considered in First Pennsylvania Banking & Trust Co. v. Commissioner [Dec. 30,859], 56 T.C. 677 (1971).
Respondent also asserts that petitioners are attempting to fragment an intangible asset which is inseparable from goodwill because such an asset's indefinite useful life and self-regenerating ability make it "inextricably linked to goodwill". This Court has considered and rejected a similar argument and most of the cases cited by respondent as not establishing a per se "mass asset" rule, but requiring a factual examination of the depreciation of each type of intangible asset. Citizens & Southern Corp. v. Commissioner, supra at 482-487. We adopt that reasoning here. The cases cited by respondent which were not previously considered, Decker v. Commissioner [89-1 USTC ╤ 9113], 864 F.2d 51 (7th Cir. 1988), affg. [Dec. 44,099(M)] T.C. Memo. 1987-388; United Finance & Thrift Corp. of Tulsa v. Commissioner [60-2 USTC ╤ 9723], 282 F.2d 919 (4th Cir. 1960), affg. [Dec. 23,232] 31 T.C. 278 (1958); Dodge Brothers, Inc. v. Commissioner, supra ; and Finoli v. Commissioner [Dec. 42,994], 86 T.C. 697 (1986), similarly do not establish a per se rule or convince us to change our prior conclusion.
Respondent contends that this case is indistinguishable from AmSouth Bancorporation v. United States [88-1 USTC ╤ 9232], 681 F. Supp. 698 (N.D. Ala. 1988), which held that the core deposit base was not depreciable because the taxpayer did not show that it had a value separate and distinct from goodwill. However, the District Court in AmSouth was faced with a different record than that existing here, including the fact that the taxpayer continued to use the acquired bank's name for several years after the acquisition.
Respondent maintains that the recognition of the core deposit base as an asset separate from goodwill for financial statement and regulatory accounting purposes should be irrelevant for tax purposes. A similar argument was considered by this Court in Citizens & Southern Corp. v. Commissioner, supra at 496-498; see also IT&S of Iowa, Inc. v. Commissioner, supra at 510-511. We have not relied on financial statement or regulatory accounting rules in reaching our conclusion.
Respondent asserts that the fact that core deposits are not considered a separate asset in the capital adequacy rules is "additional substantial evidence of their inseparability from tax goodwill." This assertion was previously considered and rejected in Citizen & Southern Corp. v. Commissioner, supra at 497-498, and IT&S of Iowa, Inc. v. Commissioner, supra at 510-511, because the regulatory concept of capital adequacy is irrelevant to the determination of whether an asset is amortizable. We follow the prior analysis and apply it here.
Additionally, respondent maintains that petitioners failed to separately value the core deposits in determining the price paid to NCNB and County Bank, which should cast "considerable doubt" on its attempt to do so for tax purposes. This again is an argument which has been rejected in Citizens & Southern Corp. v. Commissioner, supra at 494-495, and IT&S of Iowa, Inc. v. Commissioner, supra at 510. As in those cases, the record here demonstrates that Peoples considered the value of core deposits in deciding to acquire the NCNB branches and County Bank, and NCNB considered them in determining a selling price. The fact that Peoples did not obtain a detailed valuation before making the acquisition does not negate the separate existence or value of core deposits.
Respondent contends that treating the core deposits as an amortizable capital asset is inconsistent with the deduction of expenses relating to attracting customers by NCNB and County Bank. Respondent expounds that expenses which create an asset with a life extending beyond the taxable year must be capitalized, so that the expenses incurred by NCNB and County Bank are deductible only if they relate to the general establishment of goodwill and not a separate capital asset; a double deduction occurs if amortization is permitted. We rejected the same contention in Citizens & Southern Corp. v. Commissioner, supra at 490-491, and distinguished the cases cited by respondent as not dealing with a purchased intangible asset, the proper tax treatment of which should not depend on whether the seller deducted or capitalized the expenses which created it. We adopt the reasoning of Citizens & Southern here.
In support of this argument, respondent cites several cases not addressed in Citizens & Southern: Battelstein Investment Co. v. United States [71-1 USTC ╤ 9227], 442 F.2d 87 (5th Cir. 1971); Dodge Brothers, Inc. v. United States, supra ; and Meredith Publishing Co. v. Commissioner [3 USTC ╤ 1096], 64 F.2d 890 (8th Cir. 1933), affg. Successful Farming Publishing Co. v. Commissioner [Dec. 6949], 23 B.T.A. 150 (1931). Battelstein Investment and Meredith Publishing do not involve purchased assets and are inapposite to the identification of core deposits as a separate asset from goodwill. In Dodge Brothers, the Fourth Circuit did not create a per se rule that the seller's deduction of expenses relating to development of an asset negates the existence of an amortizable intangible asset in the hands of the purchaser; this was merely one factor considered in denying the separate existence of the "proved" car. Dodge Brothers, Inc. v. United States [41-1 USTC ╤ 9309], 118 F.2d 95, 100 (4th Cir. 1941). This analysis is not binding on us under Golsen v. Commissioner [71-2 USTC ╤ 9497], 445 F.2d 985 (10th Cir. 1971), and does not persuade us to depart from our prior decisions.
Respondent asserts that core deposits are liabilities rather than "property" for purposes of section 167 and the regulations thereunder, and that their transfer other than with the sale of a bank or an entire branch thereof was unusual. Similar arguments were considered in Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. at 490 and 492. These arguments simply fail in the face of Citizens & Southern Corp. v. Commissioner, supra, and IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. 496 (1991).
Additionally, respondent contends that neither petitioners' efforts to predict a useful life for the core deposits nor the allocation in the purchase agreement for the NCNB branches establish separability of the core deposits as an asset. We agree and have not relied thereon.
Respondent maintains also that the core deposits are part of the going-concern value of the NCNB branches and the County Bank because of the ability to continue these businesses without interruption. Respondent's argument appears to be a nonsequitor; the existence of going-concern value does not make the core deposits inseparable therefrom, and respondent has not cited us to any cases which conclude otherwise.
III. Useful Life
Petitioners contend that the method for determining useful life used by Doerfler and Morgan was similar to that accepted by this Court in Citizens & Southern Corp. v. Commissioner, supra at 469-471, 500-505, and as to the NCNB branches was supported by the run-off data collected from the tracking system developed by Price Waterhouse.
Respondent asserts that the core deposits do not have a determinable useful life because they are self-regenerating through postacquisition deposits and can be extended by customer satisfaction. Additionally, respondent maintains that petitioners' experts have not established a reasonably accurate estimate of the useful life because: (i) they restricted the observation to account severances, not all account closings, (ii) they measured the decline in numbers of accounts rather than account balances and did not present evidence of a correlation between the two, (iii) they did not stratify the accounts based on size, (iv) they did not make an adjustment for the effect of deregulation, (v) they did not compute separate rates for each branch, 33 (vi) they used only a few months' data relating to severances, and (vii) they failed to include confidence intervals and margins of error in the reports.
33 Respondent criticizes Morgan's report on the basis that he did not compute separate rates for personal and business accounts at the Jacksonville branch of the NCNB branches. We have not addressed this criticism because we have not relied on Morgan's conclusions or report in this regard.
We follow our opinion in Citizens & Southern Corp. v. Commissioner, supra, and agree that petitioners have shown a useful life with a reasonable degree of accuracy.
Petitioners' estimates of useful lives relating to the core deposits of the NCNB branches and County Bank are derived from historical severances for periods of 2 to 8 months in 1983 and 11 months in 1984, respectively. Doerfler's methodology regarding the NCNB branches does not differ appreciably from that accepted by this Court from him in Citizens & Southern Corp. v. Commissioner, supra. By comparison, Morgan's analysis regarding the NCNB branches used survival rates based on patterns for demand deposits only, and combined personal and business accounts at the Jacksonville branch, and Danforth's analysis regarding the County Bank adjusted the data to correct for the absence of older accounts using data from the Wilmington NCNB branches without establishing an adequate basis for this as a comparable. Although their analyses may therefore be less accurate than that accepted by this Court in Citizens & Southern, we do not consider either to be beyond the bounds of reasonable accuracy set therein. Consequently, we conclude that petitioners have established that the core deposits acquired from the NCNB branches and County Bank had limited useful lives, which were estimated by Doerfler to be 18 years and by Danforth and Curtis to be 20 years, respectively.
Respondent's first assertion, that the self-regenerative nature of core deposits prevents them from having a determinable useful life, is similar to one we rejected in Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. at 499-500, on the basis that new accounts opened are not included in the core deposits valued. The nuances added by respondent here are that balances in the acquired accounts fluctuate and are affected by customer satisfaction. However, the facts that individual core deposit balances change from time to time in the ordinary course of business, and that account closure rates are influenced in part by customer satisfaction, do not negate the basic point that acquired accounts do terminate over a demonstrable useful life.
Respondent's criticism of the use of account severances rather than closings is to be contrasted with her position in Citizens & Southern Corp. v. Commissioner, supra at 504, where respondent argued that data on closings is inaccurate because it includes situations where funds were moved between branches of the same bank. Petitioners' use of severances appears to be in response to respondent's position in Citizens & Southern and to be acceptable in estimating useful life with reasonable accuracy under the bounds established therein.
In IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 511-512, we rejected respondent's criticism of determining useful life by measuring the decline in numbers of accounts rather than balances, even where there were no impairment studies done or other independent corroboration of accuracy. We concluded that the historical rate of account closures provides the most accurate and reasonable estimate of the useful life of the core deposits. Respondent has not persuaded us to alter that conclusion here.
Similarly, we refused to disregard the projections of petitioners' experts because accounts were not stratified by size in Citizens & Southern Corp. v. Commissioner, supra at 505, and IT&S of Iowa, Inc. v. Commissioner, supra at 513. Perfection is not demanded, only reasonable accuracy.
Inasmuch as we have concluded that petitioners failed to show that any deposits other than demand deposits and savings accounts are properly treated as core deposits, any adjustment to useful life for the effects of deregulation would result in a less conservative estimate thereof (i.e., a shorter useful life). Such an adjustment would insert an uncertain variable and its absence should not serve to prevent petitioners from establishing a useful life.
Additionally, respondent criticizes the useful lives estimated by petitioners' experts as not reflecting separately computed attrition rates for each of the branches. We are convinced that the estimates of Doerfler and Danforth establish the useful lives with reasonable accuracy, and the computation of separate attrition rates for each branch is a level of precision which is not required given the facts in this case.
In Citizens & Southern Corp. v. Commissioner, supra, we accepted a 3-month observation period for account closures of three of the banks acquired. The observation periods here are equivalent thereto and on that basis we reject respondent's criticism that the account severance observation periods are too short to provide a reasonably accurate estimation of useful life.
We also reject respondent's criticism that it is "fatal" to the estimations of useful life by petitioners' experts that margins of error and confidence intervals regarding point estimates were not included in the reports, a view which respondent contends is supported by Burlington Northern, Inc. v. United States [82-1 USTC ╤ 9250], 230 Ct. Cl. 102, 676 F.2d 566 (1982). Doerfler testified that such statistical testing would involve assumptions that would not apply here, and would thus be meaningless. Doerfler, a well qualified statistical expert, stated in his report: "In conclusion, it is my opinion that the analytical methods described here support the contention that these deposit bases have a well-defined life, and exhibit predictable decay rates, both of which can be estimated in an objective manner that is consistent with sound statistical practices." Based on this opinion on his part and the absence of any indication in the record that the margins of error or confidence intervals were vital to the opinion or out of line, we conclude that Doerfler's estimate reflects a reasonably accurate measure of useful life.
Burlington Northern is inapposite because the taxpayer there attempted to establish useful lives through general mortality statistics for types of property very different from that involved in the case. Doerfler used actual data from the NCNB branches, not general data. In fact, the court recognized that in Burlington Northern, Inc. v. United States [82-1 USTC ╤ 9250], 676 F.2d at 578: "This is not to say that results obtained by the proper application of the actuarial method must be verified through the use of specific principles of mathematical statistics."
Danforth's estimate of useful life was similarly based on actual data, not general mortality statistics, and is not negated by the absence of margins of error and confidence interval computations.
A. NCNB Branches
Petitioners contend that Morgan's original $4,820,000 value for the core deposits included in the NCNB branch purchase agreement is the fair market value as corroborated by Curtis' valuation and because an adequate amount was allocated to goodwill. Alternatively, petitioners suggest that Curtis' valuation using the cost savings methodology reflects fair market value.
Respondent asserts that the $4,820,000 value was not bargained between parties with adverse tax positions relating thereto. Respondent maintains also that Morgan's $4,820,000 value was unreliable and inaccurate because there was no conclusion regarding value in the report, valuation conclusions in Morgan's correspondence varied widely, Price Waterhouse arrived at a significantly lower valuation initially using the coefficients in his report, and the $4,820,000 value was computed as of October 31, 1982, not the closing date, and included the value of tax benefits associated with amortization of the deposits. Respondent further suggests that the income method used by Morgan "includes the value of the federal deposit insurance guarantee." Respondent contends that Curtis' valuation is erroneous because: (i) in adopting Doerfler's assumptions, it assumes that deposit balances will decline at the same rate as account severances, (ii) the accounts were not stratified by size, (iii) the cost of alternative funds assumed is unsupported and so high that it would not have been profitable to borrow at such rates, (iv) there was no adjustment to the core deposit cost for the resulting shift from demand deposits and savings accounts to higher cost deposits caused by deregulation, (v) the 5-percent inflation factor assumed encompasses postacquisition deposits and was unverified, and (vi) it assumes the full tax benefit of the amortization of the deposits. Additionally, respondent suggests that the true cost of core deposits cannot be calculated because the interrelated nature of banking services provided and associated costs create certain costs which cannot be measured, such as "implicit interest" (e.g., when a bank provides a favorable line of credit to a significant depositor).
We agree with respondent to the extent that we will not follow Morgan's $4,820,000 valuation. While we conclude that petitioners may calculate the value of the core deposits using the cost savings methodology, we conclude that certain modifications must be made to arrive at the fair market value. See IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 515. We have only analyzed the methodology used to the extent objections were raised thereto.
Initially, we do not consider the $4,820,000 value allocated in the purchase agreement to core deposits to be the result of bargaining between parties with adverse tax positions. UFE, Inc. v. Commissioner [Dec. 45,793], 92 T.C. 1314, 1324 (1989). We reject Morgan's valuation for two reasons. First, the valuation is long on complicated formulae and short on valuation conclusions and the sources for various underlying assumptions. Restricting the conclusion to coefficients at best led to a confused record regarding Morgan's "final" opinion, and suggests that the opinion was orchestrated and result-oriented so as to support the $4,820,000 preliminary valuation. Second, we have implicitly indicated a preference for the cost savings method as contrasted with the income projection method. Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. 463 (1988).
We note that Kane, respondent's expert, recognized the vitality of core deposits in our holding in Midlantic National Bank v. Commissioner [Dec. 40,477(M)], T.C. Memo. 1983-581, but failed to provide us with any estimate of the value of the core deposits before us in this case. His report and testimony were focused only on a critique of the value determination of petitioners' experts. Although we understand that respondent's primary position is that core deposits are not an asset separable from goodwill and going-concern value, we believe that given our holding in Citizen & Southern a secondary position encompassing a valuation would have been appropriate and helpful. This failure handicapped the Court in reaching a valuation in the case at hand.
1. Net Investable Funds
In computing the "net investable funds" for which cost savings are determined, Curtis multiplied Doerfler's survival rates by the average balances of demand deposits and NOW, money market, and savings accounts, adjusted by a 5-percent inflation factor, and reduced for reserve requirements and float. First, we concluded above that petitioners had only shown that demand deposits and savings accounts meet the definition of core deposits, which will require a modification to Curtis' valuation. Second, the 5-percent inflation factor appears to us to be inappropriate given the likely effects of deregulation (i.e., a shift from demand deposits and savings accounts) which to some extent are reflected in the record vis-a-vis NCNB's monitoring of deposits before closing; additionally, we were not provided a satisfactory explanation or source for this assumption.
We have previously rejected respondent's arguments that account severances are not sufficiently correlated to the decline in deposit balances, and that account severances must be stratified by size, and do so in this context as well.
2. Cost of Core Deposits
Curtis projected the ongoing cost of the core deposits using actual interest rates paid by Peoples, and service expenses, reduced by deposit fee income, based on 1983 FCA data. Although actual service costs and deposit fee income of Peoples might have been preferable (see IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. 496, 521-523 (1991)), respondent has not pointed to nor do we perceive any flaws in this basic methodology. Except for the modifications necessary because only demand deposits and savings accounts were shown to be core deposits, we conclude that Curtis' projection of the cost of the core deposits is reasonable.
Respondent contends that Curtis did not make an adjustment for increased costs because of deregulation. However, with only demand deposits and savings accounts included as core deposits, we do not believe that such an adjustment is necessary. See IT&S of Iowa, Inc. v. Commissioner, supra at 524-525.
As in IT&S of Iowa, Inc. v. Commissioner, supra at 521-522, we will not reject the cost savings methodology because of respondent's suggested argument that the interrelated nature of banking services and associated costs make it impossible to measure the core deposit costs. We have concluded similarly that Curtis' projection of core deposit costs is a reasonable one.
3. Cost of Alternative Source of Funds
Curtis opined that the most appropriate alternative source of funds for Peoples would be certificates of deposit with maturity of 5 years or more and assumed rates ranging from 10.20 to 11.20 percent as their cost. However, he failed to provide a source for these rates, and 5-year certificates of deposit were rare at that time. In IT&S of Iowa, Inc. v. Commissioner, supra at 528-529, we concluded that the most appropriate cost of the alternative source would be certificates of deposit issued by the taxpayer. We rejected the argument that comparability to core deposits necessitated that the certificates have lengthy maturities, and concluded that if the certificates had comparable attrition rates, 34 then they would be an alternative source of funding.
34 Both Morgan and Danforth included certificates of deposit in estimating a useful life for the deposits valued, and thus we conclude the certificates of deposit have comparable attrition rates to the core deposits.
The record indicates that Peoples was offering rates ranging from 8.14 to 9.25 percent for certificates of deposit during May of 1983. For the week ending March 25, 1983, the rates in the national market ranged from 8.20 to 9.20 percent and in the secondary market from 8.79 to 9.08 percent, respectively. Petitioners' expert, Morgan, testified that the most favorable alternative source of funds would be certificates of deposit issued to institutional investors. From the record, we conclude that the appropriate cost of alternative funds here is 9 percent.
To the extent indicated above, we agree with respondent's criticism that the cost of alternative funding selected by Curtis is unsupported and too high.
4. Tax Benefit of Amortization as Part of Value
In valuing the core deposits, Curtis assumed a 20-percent tax rate and considered the effect of taxes in selecting the capitalization rate applied to the stream of net cost savings. In Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. at 506-507, and IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 532-533, we allowed the taxpayers to include as part of the value the full tax benefit of amortization of the core deposits using a reasonable method. The methodology used by Curtis appears to be similar to that which he used in Citizens & Southern. Accordingly, we will follow our prior opinions.
Respondent contends that the fair market value of the core deposits would not have included the "full" tax benefit of the amortization because there was no case law allowing amortization of core deposits at the time of the acquisition of the NCNB branches and the Treasury position was clearly to the contrary. Respondent emphasizes that petitioners' expert Morgan opined that including the "full" tax benefit would not reflect the uncertainty at the time regarding this issue. The recent decisions of this Court, however, obviate any need to give any weight either to respondent's contention or Morgan's opinion.
B. County Bank
Petitioners contend that both Danforth and Curtis used acceptable methods of valuing the core deposits and their valuations support petitioners' allocation of the $672,636.21 premium to the core deposits. Petitioners reason that the $1,053,075 value found by Danforth, and the $809,000 value determined by Curtis were so much higher than the $672,636.21 premium paid for the County Bank because it was a "bargain purchase" based on its bad financial condition. Petitioners assert that using the residual method of allocating goodwill or going-concern value is appropriate, even if the value allocated thereto is zero.
Respondent maintains that the valuations of Danforth and Curtis are so out of line as to cast doubt on their methodologies and would either result in negative goodwill or a write-down of all other assets. Respondent asserts that petitioners' contention that this was a bargain purchase is unsupported by the record and contrary to the parties' stipulation that the price paid was fair market value.
We agree with petitioners to the extent that Curtis' cost savings methodology with certain modifications supports the allocation. We have embraced the cost savings methodology in Citizens & Southern Corp. v. Commissioner, supra at 506-507, and IT&S of Iowa, Inc. v. Commissioner, supra at 514-515, and continue to do so here. The income projection here is premised on some unsupported assumptions and suffers from some of the same infirmities as existed with the Morgan valuation.
First, in computing "net investible funds", there must be adjustments to reflect that petitioners have only shown that demand deposits and savings accounts are core deposits and that the 5-percent inflation factor is inappropriate for the reasons stated above.
Second, only costs relating to demand deposits and savings accounts should be considered in determining costs of core deposits.
Third, based on IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 528-529, the appropriate cost of alternative funds is Peoples' cost of certificates of deposit at the time of the County Bank acquisition. The record indicates that Peoples was offering rates ranging from 9.25 to 11 percent as of September 27, 1984; rates quoted in the Federal Reserve Bulletin ranged from 10.90 to 11.40 percent. We conclude that the appropriate cost of alternative funds is 10.75 percent.
We note that the Curtis valuation with the above modifications will be closer to the $672,636.21 premium. Moreover, Peoples' allocation does not result in negative goodwill. We agree that the assertion that this was a bargain purchase appears contrary to the stipulation that the price reflects fair market value. However, the value of County Bank on the market apparently suffered from the financial problems with the bank found by the Federal and State regulators, which were the subject of the memorandum of understanding and are in the record. Additionally, the selling price equaled the amount originally paid by the County Bank shareholders, and there is evidence in the record of other attempts to sell which apparently fell through. Thus, the fact that the fair market value was low is supported in the record.
V. Amortization Method
Petitioners contend that the amortization method used was the same as the one accepted as reasonable in Citizens & Southern Corp. v. Commissioner [Dec. 45,036], 91 T.C. at 512-514.
Respondent asserts that petitioners have not shown that the straight-line method is not a reasonable allowance or that another method provides a reasonable allowance because of (i) the flaws in the useful life analysis of petitioners' expert (see 50-51 supra ), particularly that petitioners have not established a correlation between account severances and decline in deposit balances, and (ii) petitioners' use of the straight-line method for financial statement purposes.
We agree with petitioners to the extent that the amortization method described by Curtis subject to appropriate modifications described above results in a reasonable allowance for depreciation as to the NCNB branches and the County Bank under Citizens & Southern Corp. v. Commissioner, supra.
The depreciation of intangibles is not restricted to the straight-line method if another method provides a more reasonable allowance. Citizens & Southern Corp. v. Commissioner, supra at 512 (quoting Computing and Software, Inc. v. Commissioner [Dec. 33,197], 64 T.C. 223, 237 (1975)). In Citizens & Southern Corp. v. Commissioner, supra at 514, we described the taxpayer-bank's depreciation method as "based upon the contribution the current year's cost savings made to the value of" the core deposits, and concluded that it resulted "in a fair allocation of the basis of the asset to periods in which benefit is realized and [provided the taxpayer] a reasonable allowance for depreciation." We conclude that this is controlling here in allowing petitioners to use the amortization method reflected in Curtis' report subject to the modifications described above.
We concluded above that none of respondent's contentions regarding flaws in the useful life analysis of petitioners' experts negated petitioners' showing of reasonable estimates of useful life. We conclude similarly here regarding the Curtis amortization method.
Respondent suggests that there is a 40-percent discrepancy between the deposit balance decline under the NCNB branch tracking system and the amortization deducted by petitioners. Petitioners counter on brief that there were numerous errors in the tracking system, but in terms of account severances it is not significantly different from Doerfler's projections. Respondent answers that petitioners failed to present witnesses to explain the errors, and that if the tracking system is erroneous from the deposit balance standpoint then it undoubtedly is from the account severance standpoint. First, we are not allowing petitioners the amount of amortization deducted, but allowing the amount under Curtis' method. Second, although we are troubled by the insufficiently explained errors in its tracking system alleged by petitioners and its discontinuance, in IT&S of Iowa, Inc. v. Commissioner [Dec. 47,735], 97 T.C. at 512, we accepted reliance on account closings to establish useful life, even absent an independent effort to show a correlation to deposit shrinkage. Therefore, we will not penalize petitioners for the attempt to provide a tracking system given this record.
Respondent's contention that the fact that the straight-line method was used for financial statement purposes precludes petitioners from arguing that it does not provide a reasonable allowance appears to be controlled by Citizens & Southern Corp. v. Commissioner, supra at 514.
Accordingly, based on Citizens & Southern and IT&S of Iowa, we conclude that petitioners have shown that the demand deposits and savings accounts relating to the acquisitions of the NCNB branches and County Bank were core deposits, the core deposits were an asset separate and distinct from goodwill and going-concern value, the core deposits acquired had useful lives of 18 and 20 years, respectively, the values thereof are properly computed using Curtis' cost savings methodology as modified herein, and the amortization method of Curtis as modified is a reasonable allowance and may be used. Additionally, the acquisition costs may be allocated to the core deposits of the NCNB branches and County Bank, respectively, in proportion to the values of the assets acquired, including the values of the core deposits so computed (see note 3 at 3 and the text at 37 supra ).
To reflect the foregoing,
Decision will be entered under Rule 155.