Deposit Activity in
South Dakota Banks and Credit Unions
Leon Korte
The University of South Dakota School of Business
Vermillion, South Dakota
November 2003
Technical and Scale Efficiency Estimates of Industry Classes
Summary and Conclusions
In a recent study of regulatory reports for credit unions and community banks reporting in South Dakota for calendar years 1999, 2000, and 2001, comparing cost of operations including cost of funds and level of deposits, it can be concluded that commercial banks operate more efficiently than credit unions. Community banks reported higher cost of funds when comparing a single aspect of generating deposits, but when including multiple productivity factors community banks reported lower overall costs relative to the levels of deposits generated. The operating reasons underlying these efficiencies may be the result of institutional size, the degree of professional management, or the nature of the regulations under which the two types of financial institutions operate. The comparisons looked at operating costs before income taxes in order to approximate similar operating conditions for the two groups.
Community banks and credit unions reflect similar characteristics in terms of geographic limitations to market area, relatively limited customer base and type of financial products offered to consumers. The two types of institutions operate in a competitive environment seeking deposits which can be converted into earning assets. Although the financial services market is relatively similar for the two types of institutions, the regulatory environment differs between the two segments. As cooperatives, credit unions are deemed not-for-profit institutions and are not subject to federal taxation. The question arises as to whether credit unions operate as efficiently as community banks in operating activities, specifically, in this case, in attracting deposits to the institution.
Customer deposits in financial institutions provide a foundation for lending activities. In the case of credit unions it is a function of members pooling their resources to make such resources available for lending to other members. For community banks deposits from the community are converted into loans to other members of the community. The cost of acquiring deposits reduces either the rate of return depositors can earn or it reduces available resources that might be loaned to borrowers. Minimizing these costs leads to potential benefits for depositors and borrowers. Identifying potential areas of efficiency improvement in financial institutions is a step in economic development.
Data were collected for each community bank reporting a South Dakota headquarters. Deposit variables included transaction accounts [checking accounts and similar demand accounts] and nontransaction accounts [savings accounts and certificates of deposit]. Operating cost variables included interest expense on deposits [the dollar amount banks spend on customer deposits], total noninterest expense [the cost of operating the institution excluding employee compensation and benefits], and employee costs represented by two separate but related variables: compensation and benefits and number of employees.
Data were also collected for each active credit union operating in the state of South Dakota. Deposit variables included share drafts and other demand deposit type accounts which were identified as transaction accounts and other deposits which were classified as nontransaction accounts. Operating cost variables included cost of funds [including dividends on shares and interest on deposits], total other operating expenses, salaries and benefits and an estimate of number of employees. The provision for loan losses and leases were omitted from both groups because it reflects a management decision based on loan performance and does not directly impact the cost of generating deposits in a given year.
Data were collected for sixty-three community banks and fifty-eight credit unions for the years ended December 31, 1999, 2000, and 2001. Commercial banks reflected on average approximately three times the level of nontransaction deposits and total deposits than were reported by credit unions. The average cost of funds as estimated by interest costs divided by nontransaction deposits were higher for commercial banks in all three years of the study period. When total deposits were used as the denominator for calculating the average cost of funds, credit unions reported higher costs of funds. This reflects the proportionately larger transaction deposits reported by commercial banks. The average number of employees reported by commercial banks was approximately twice the number of employees reported by credit unions. The average compensation and benefits cost per employee was approximately fifty percent greater for commercial banks as compared to credit unions.
In terms of absolute numbers, community banks are, on average, about three times larger than credit unions. Transaction accounts represent a relatively larger percentage of total deposits for community banks with a correspondingly higher average cost of funds. Community banks reflected approximately 25 percent greater total deposits per full-time equivalent employee than was reported by credit unions. The average community bank reported on average twice as many full-time equivalent employees and a greater average cost per full-time equivalent than was reported by the average credit union, but the dollar amount of total deposits reported per dollar of compensation and benefit cost was only marginally different when comparing community banks and credit unions. The difference appeared to be greater when comparing the total nontransaction deposits per dollar of compensation and benefit cost was only marginally less for community banks than for credit unions, perhaps reflecting the difference in the relative levels of transaction deposits between the two groups.
Technical and Scale Efficiency Estimates of Industry Classes
Nonparametric models were estimated for each year using virtual observations based on projections of the actual observations to the BCC variable returns to scale efficiency frontier. Virtual observations control for the local variations in management and treat each observation as if it were operating at its benchmark potential. This projection to the benchmark potential allows for an evaluation of the industry segment in addition to the earlier evaluation of the individual observations. After controlling for variations within the individual observations, the modeling results for the industry segments appear consistent with the relative efficiency estimates for the individual firms. Community banks populated the efficiency frontier in greater numbers than did credit unions in each year of the study period. The relative ranking of the observations also reflected greater relative efficiency for community banks than for credit unions in each year and in the pooled data. Consistent with the rankings for the individual firms, the difference in the rankings for the virtual observations were statistically significant in the first and last years of the study period (1999 and 2001) and for the pooled data. The difference in the rankings for 2000, while appearing consistent with the other years, was not significant at the usual confidence levels.
The distribution of observations based on exhibited returns to scale when virtual observations were evaluated was nearly identical to the distribution observed when modeling actual data. The consistency in exhibited returns to scale characteristics suggests no material managerial inefficiencies or resource misallocations in either commercial banks and credit unions on an individual basis.
Summary and Conclusions
Using a multivariate nonparametric modeling process commercial banks reflected greater relative technical efficiency than did credit unions in generating deposits. This relative efficiency advantages holds true when using actual observations to evaluate individual firms or when using virtual observations based on industry segment benchmark estimates to evaluate industry segments. At the same time it appears that there are potential material resource shifts that might be made to make each segment of the industry more efficient in deposit generating activities. Most actual observations reflected decreasing returns to scale when estimated against other firms suggesting increases in resources will yield less than proportionate increases in deposits.
Potential resource shifts might make each segment of the industry more efficient in deposit generating activities. Most actual observations reflected decreasing returns to scale when estimated against other firms suggesting increases in resources will yield less than proportionate increases in deposits.
02/23/2004 04:43:44 PM -0600