Abstract
A topic of recent interest in the retail financial sector has been the growth of credit unions or “pure cooperatives”. Past credit union researchers built mathematical models of credit union operations. These models identified important operating characteristics but were modeled under assumptions of static operating environments. The model presented in this paper departs from the traditional static models and examines dynamic operation for a United States credit union. Its inter-temporal structure clarifies a number of issues—such as optimal equity retention and inter-temporal rate policy—not addressed by earlier studies. Given initial conditions, the model specifies equity retention and inter-temporal deposit and loan rate policies until an equilibrium state is reached.
Similar content being viewed by others
Notes
http://www.woccu.org/memberserv/intlcusystem, accessed on the 14th May, 2012.
http://cuna.org/public/press/credit-union-basics, accessed on the 14th May, 2012.
http://www.cunamutual.com/portal/server.pt/document/2700462/june_2011_cu_trends_report, accessed on the 14th May, 2012.
See Federal Credit Union Act 109.
http://cuna.org/public/press/credit-union-basics, accessed on the 14th May, 2012.
http://www.woccu.org/memberserv/intlcusystem, accessed on the 14th May, 2012.
See the 2001 and 2008 Credit Union Directory, Callahan and Associates, 1.
Smith (1986) denotes an entire paper to the discussion of objective functions.
p can be thought of as the ratio of a constant rate of time preference, q, and the instantaneous probability of credit union survival, s(E,L). As E→0 or L→∞, s→0. As E→∞ or L→0, s→1. Blanchard (1985) uses a similar discount factor form.
Assume, for example, that the loan rate is dropped slightly from such an equilibrium. Borrowers unequivocally benefit from such a change because both loan volume and the credit union/market rate spread increase. If these loans are made by shifting resources from lower-return investments, interest income might also increase, making additional saver allocations possible. Even if their benefit allocations are not identical after the rate change, both borrowers and savers prefer the change.
The loan/investment decision is not a simple function of marginal yields and rates because credit unions often carry positive investments even though the marginal loan rate nearly always exceeds the investment rate. The ultimate rationale for investments is not readily apparent, but probably appeals to risk.
The modern formulation of this technique can be found in Pontryagin et al. (1962).
Blanchard and Fischer (1989) review the use of phase diagrams in optimal control theory.
Our model has no expectations component, but the effect of an anticipated tax change might have different effects. Forward-looking credit unions would probably smooth out the large anticipated drop in \(d_{\mathit{cu}}^{\ast}\) and increase in \(l_{\mathit{cu}}^{\ast}\) by moderately lowering the spreads before the tax change, increasing E, and then supporting these spreads after the tax change by drawing down the enlarged equity reserve.
Concavity of the loan demand schedule ensures that marginal revenue (with respect to the loan rate) will eventually become negative. If the credit union rations credit based on risk, for example, an increase in loan volume is accompanied by higher default. Eventually, the cost of marginal loans will exceed revenue.
The various non-negativity constraints are discussed in the final analysis rather than modeled explicitly for reasons of tractability.
Notice that any non-negative level of equity is feasible, but that the deposit rate is constrained to lie above the market deposit rate.
If steady state deposits and equity are monotonic in the loan rate (analysis in the following section suggests an indeterminant relationship between steady state deposit and loan rates), the loan rate which satisfies this equation is unique. A formal proof is omitted, but the basic logic is that an increase in the loan rate decreases loan demand and investment volume from (A.8), but also increases steady state deposits and equity. Assume that \(l_{\mathit{cu}}'\) satisfies (A.27). A higher rate decreases the left-hand side of (A.27) and increases the right-hand side, so is not admissible. A lower rate is similarly inadmissible since it causes an increase in the left-hand side of (A.27) and decrease in the right-hand side. Continuity ensures the existence of the unique \(l_{\mathit{cu}}'\) which satisfies (A.27) and is denoted \(l_{\mathit{cu}}^{\ast}\).
References
Bauer, K. (2008). Detecting abnormal credit union performance. Journal of Banking & Finance, 32, 573–586.
Bauer, K., Miles, L. L., & Nishikawa, T. (2009). The effect of mergers on credit union performance. Journal of Banking & Finance, 33, 2267–2274.
Blanchard, O. J. (1985). Debt, deficits, and finite horizons. Journal of Political Economy, 21, 87–96.
Blanchard, O. J., & Fischer, S. (1989). Lectures on macroeconomics. Cambridge: MIT Press.
Brown, C., & Davis, K. (2009). Capital management in mutual financial institutions. Journal of Banking & Finance, 33, 443–455.
Burger, A. E., & Kelly, W. A. Jr. (1993). Building high loan/share ratios: challenges and strategies. Madison: Filene Research Institute.
Cargill, T. F. (1977). Recent research on credit unions: a survey. Journal of Economics and Business, 29, 155–162.
Casu, B., Girardone, C., & Molyneux, P. (2004). Productivity change in European banking: a comparison of parametric and non-parametric approaches. Journal of Banking & Finance, 28, 2521–2540.
Chateau, J. P. D. (1980). The demand for and supply of deposits by credit unions: the caisses populaires’ case. Journal of Banking & Finance, 4, 151–173.
Comparative Digest of Credit Union Acts (1989). Madison: Credit Union National Association.
Comparative Digest of Credit Union Acts (1999). Madison: Credit Union National Association.
Comparative Digest of Credit Union Acts (2008). Madison: Credit Union National Association
Croteau, J. T. (1963). The economics of the credit union. Detroit: Wayne State University Press.
Davis, K. (2001). Credit union governance and survival of the cooperative form. Journal of Financial Services Research, 19, 197–210.
Ekman, E. V. (1982). A dynamic model of a managerial firm. In G. Feichtinger (Ed.), Optimal control theory and economic analysis (pp. 79–105). North Holland: Elsevier.
Emmons, W. R., & Schmid, F. A. (1999). Credit unions and the common bond. Federal Reserve Bank of St. Louis Review, 81, 41–64.
Esho, N. (2001). The determinants of cost efficiency in cooperative financial institutions: Australian evidence. Journal of Banking & Finance, 25, 941–964.
Flannery, M. J. (1974). An economic evaluation of credit unions in the United States. Boston: Federal Reserve Bank of Boston.
Forster, B. A. (1977). On a one state variable optimal control problem. In J. D. Pitchford & S. J. Turnovsky (Eds.), Applications of control theory to economic analysis (pp. 229–240). Amsterdam: North Holland.
Frame, W. S., Karels, G. V., & McClatchey, C. A. (1999). Deposit insurance and risk-taking behavior in the credit union industry. Journal of Banking & Finance, 23, 105–134.
Frame, W. S., Karels, G. V., & McClatchey, C. A. (2003). Do credit unions use their tax advantage to benefit members? Evidence from a cost function. Review of Financial Economics, 12, 35–47.
Fried, H. O., Lovell, C. A. K., & Eeckaut, P. V. (1993). Evaluating the performance of U.S. credit unions. Journal of Banking & Finance, 17, 251–265.
Fried, H. O., Lovell, C. A. K., & Yaisawarng, S. (1999). The impact of mergers on credit union service provision. Journal of Banking & Finance, 23, 367–386.
Fry, C. L., Harper, C. P., & Stansell, S. R. (1982). An analysis of credit union costs: a new approach to analyzing costs of financial institutions. Journal of Bank Research, 12, 239–249.
Garden, K., & Ralston, D. (1999). The X-efficiency and allocative efficiency effects of credit union mergers. Journal of International Financial Markets, Institutions & Money, 9, 285–301.
Glass, J. C., & McKillop, D. G. (2006). The impact of differing opportunity environments on U.S. credit union performance. Applied Financial Economics, 16, 1285–1300.
Glass, J. C., McKillop, D. G., & Rasaratnam, S. (2010). Irish credit unions: investigating performance determinants and the opportunity cost of regulatory compliance. Journal of Banking & Finance, 34, 67–76.
Goddard, J., Molyneux, P., & Wilson, J. O. S. (2004). Dynamics of growth and profitability in banking. Journal of Money, Credit, and Banking, 36, 1069–1090.
Goddard, J. A., McKillop, D. G., & Wilson, J. O. S. (2008). The diversification and financial performance of US credit unions. Journal of Banking & Finance, 32, 1836–1849.
Gooptu, S., & Lombra, R. (1987). Aggregation across heterogeneous depository institutions. Financial Review, 22, 369–378.
Hart, O., & Moore, J. (1996). The governance of exchanges: members’ cooperatives versus outside ownership. Oxford Review of Economic Policy, 12, 53–69.
Hart, O., & Moore, J. (1998). Cooperatives vs. outside ownership (Working Paper No. W6421). National Bureau of Economic Research.
Hempel, G. H., & Yawitz, J. B. (1971). Financial management at financial institutions. New York: Prentice Hall.
Jackson, W. E. (2007). Is the U.S. credit union industry overcapitalized? An empirical examination. Madison: Filene Research Institute.
Kane, E. J., & Hendershott, R. (1996). The federal depost insurance fund that didn’t put a bite on U.S. taxpayers. Journal of Banking & Finance, 20, 1305–1327.
Keating, B. P. (1979). Prescriptions for efficiency in nonprofit firms. Applied Economics, 11, 321–332.
Keating, B. P., & Keating, M. (1975). A managerial discretion theory of the nonprofit firm: an economic model of a credit union. Journal of Business Research, 3, 345–354.
Kohers, T., & Mullis, D. (1987). The impact of credit union parent company business on borrower saver treatment. RBER, Review of Business and Economic Research, 23, 38–49.
Koot, R. S. (1976). The demand for credit union shares. Journal of Financial and Quantitative Analysis, 11, 133–141.
Leggett, K. J., & Stewart, Y. H. (1999). Multiple common bond credit unions and the allocation of benefits. Journal of Economics and Finance, 23, 235–245.
Leggett, K. J., & Strand, R. W. (2002). Membership growth, multiple membership groups, and agency control of credit unions. Review of Financial Economics, 11, 37–46.
Mangasarian, O. L. (1966). Sufficient conditions for the optimal control of nonlinear systems. SIAM Journal on Control, 4, 139–152.
Melvin, D. J., Davis, R. N., & Fischer, G. C. (1977). Credit unions and the credit union industry. New York: New York Institute of Finance.
Moody, J. C., & Fite, G. C. (1971). The credit union movement: origin and development 1850–1970. Lincoln: University of Nebraska Press.
Murray, J. D., & White, R. W. (1980). Economies of scale and deposit taking financial institutions in Canada. Journal of Money, Credit, and Banking, 12, 58–70.
Murray, J. D., & White, R. W. (1983). Economies of scale and economies of scope in multiproduct financial institutions: a study of British Columbia credit unions. The Journal of Finance, 38, 887–902.
Navratil, F. J. (1981). An aggregate model of the credit union industry. The Journal of Finance, 36, 539–549.
Overstreet, G. A. Jr., & Rubin, G. M. (1991). Blurred vision: challenges in credit union research and modeling. Madison: Filene Research Institute.
Patin, R. P. Jr., & McNiel, D. W. (1991). Benefit imbalance among credit union member groups: evidence of borrower-dominated, saver-dominated, and neutral behavior? Applied Economics, 23, 769–780.
Pontryagin, L. S., Boltyanskii, V. G., Gamkrelidze, R. V., & Mishchenko, E. F. (1962). The mathematical theory of optimal processes. New York: Interscience.
Prompt Corrective Action Proposal for Reform (2005). National Credit Union Administration.
Ralston, D., Wright, A., & Garden, K. (2001). Can mergers ensure the survival of credit unions in the third millennium? Journal of Banking & Finance, 25, 2277–2304.
Report of the Committee on Credit Union Uniqueness (Philosophy) to the CUNA Executive Committee (1988). Madison: Credit Union National Association.
Smith, P. F. (1971). Economies of financial institutions and markets. Scarborough: Irwin Publishing.
Smith, D. J. (1981). On variant objective functions in credit unions: theory and empirical testing. Unpublished doctoral dissertation, University of California, Berkeley.
Smith, D. J. (1984). A theoretic framework of the analysis of credit union decision making. The Journal of Finance, 39, 1155–1168.
Smith, D. J. (1986). A test for variant objective functions in credit unions. Applied Economics, 18, 959–970.
Smith, D. J. (1988). Credit union rate and earnings retention decisions under uncertainty and taxation. Journal of Money, Credit, and Banking, 20, 119–131.
Smith, D. J., Cargill, T. F., & Meyer, R. A. (1981). Credit unions: an economic theory of a credit union. The Journal of Finance, 36, 519–528.
Sollenberger, H. M. (2005). Building a consensus on credit union capital adequacy: at the appropriate level, risks are covered and competitive capability is enhanced. Bank Accounting & Finance, 20, 3–14.
Sydsaeter, K. (1978). Optimal control theory and economics: some critical remarks on the literature. Scandinavian Journal of Economics, 80, 113–117.
Taylor, R. A. (1971). The credit union as a cooperative institution. Review of the Social Economy, 29, 207–217.
Taylor, R. A. (1977). Credit unions and economic efficiency. Rivista Internazionale Di Scienze Economiche E Commerciali, 24, 239–247.
Taylor, R. A. (1979). Optimal reserve levels for credit unions. Rivista Internazionale Di Scienze Economiche E Commerciali, 26, 971–983.
Thompson, P. (1990). A historical look at the common bond and field of membership. Madison: Credit Union National Association.
Walker, M. C., & Chandler, G. G. (1977). On the allocation of the net monetary benefits of credit union membership. Review of the Social Economy, 35, 159–168.
Walker, M. C., & Chandler, G. G. (1978). Equitable allocation of credit union net revenues: a goal programming approach. Journal of Economics and Business, 31, 63–69.
Ward, A., & McKillop, D. (1997). The relationship between credit union objects and cooperative philosophies (Working Paper). School of Management and Economics, Belfast.
Worthington, A. C. (1998). The determinants of non-bank financial institution efficiency: a stochastic cost frontier approach. Applied Financial Economics, 8, 279–287.
Acknowledgements
The authors are grateful to the anonymous reviewers, who provided many helpful comments and suggestions for improvement. Funding of this research work was, in part, supported by the University of Cape Town Research Office and the Carnegie Research Development Grant.
Author information
Authors and Affiliations
Corresponding author
Appendix
Appendix
This appendix derives the steady state equilibrium solution to the system (1), (2), and (3), and the properties of convergence towards equilibrium.
1.1 A.1 Objective function
The credit union chooses functions d cu (t), l cu (t), and E(t) to maximize
where
-
l(t)=l m −l cu (t)>0; market loan rate less credit union loan rate, which is assumed positive
-
d(t)=d cu (t)−d m >0; credit union deposit rate less market deposit rate, which is assumed positive
-
U(⋅)= borrower utility function; U′(⋅)>0; U″(⋅)<0
-
V(⋅)= saver utility function; V′(⋅)>0; V″(⋅)<0
-
p(⋅,⋅)= discount rate of typical credit union members;
(A.4) -
E(t)= credit union equity; E>0; E(0)=E 0
-
D(t)= credit union deposit volume; D>0
-
L(t)= credit union loan volume; L>0.
1.2 A.2 Constraints
Loan demand is an increasing, concave function of the difference between the market and credit union rates.
Deposit demand is an increasing linear function of the difference between the credit union and market rates.
The balance sheet constraint ensures that credit union loans plus investments, I(t), equal deposits plus equity.
This constraint relates loan and investment volume.
where
r is the investment rate of return with r<d m ,
and τ=1—credit union tax rate; 0<τ≤1.
This is the income statement constraint, which states that addition to equity equals interest income plus investment income less dividends and cost. Operating costs net of fee income, c, is increasing and concave in both loans and deposits, reflecting the increasing returns to scale found by a number of researchers (see, e.g., Fry et al. 1982; Murray and White 1983). For model tractability, we assume cost increases linearly with respect to deposit volume, i.e., \(\frac{\partial^{2} c}{\partial D^{2}}=0\). Presently, τ is unity but for purposes of analysis is allowed to range between zero and one.
1.3 A.3 Solution
The solution method employed is optimal control. First, (A.7) and (A.9) are combined by substituting out I to form the new constraint
E is the state variable and l cu and d cu are the controls. I is solved as a function of l cu via (A.8). The resulting Hamiltonian is
where λ(t) is the costate variable associated with the combined constraint.
The necessary conditions areFootnote 18
Since U(⋅), V(⋅), and c(⋅,⋅) are all strictly concave, the following transversality condition is sufficient to show that (A.13), (A.14), (A.15), and (A.16) yield a unique maximum (see Mangasarian 1966 for proof).
Next we make the following substitutions for notational convenience.
solving (A.14) for λ
which can be differentiated with respect to time
Rearranging terms in (A.13) yields
substitution (A.18) into (A.20), then (A.18), (A.19) and (A.20) into (A.12) gives
Rearranging terms gives
which, along with (A.15), forms the system of two differential equations which characterizes the optimal evolution of equity and the deposit rate.
At this point, one should note the absence of a differential equation in l cu . In the analysis that follows, working with two rather than three differential equations is more revealing because two-dimensional phase diagrams are more graphically palatable than three-dimensional versions. Convenience and pedagogy underlie the selection of equity and the deposit rate. Solving (A.13) for λ and differentiating, and substituting (A.14) into (A.13) would have produced differential equations in equity and the loan rate. Reducing the system to two differential equations mildly restricts the results of later analysis, but the alternative is to solve both (A.13) and (A.14) for λ, differentiate both with respect to time, and then manipulate the three equation system. The significant computational advantage of working with two differential equations justifies the methodology.
1.4 A.4 Steady state analysis
Functions E(t), I(t), d cu (t), and l cu (t) that satisfy (A.15), (A.22), (A.7), (A.8), and (A.16) given the initial condition E(0)=E 0 are optimal policy for the credit union. The phase diagram in Fig. 1, plotted in (E, d cu ) space for a given l cu , illustrates the system.Footnote 19 The two thick curves are two equilibrium loci. The first equilibrium locus (the deposit rate locus) is found by setting \(\dot{d}_{\mathit{cu}} = 0\) and \(\dot{p}=0\) in (A.22) and solving for p(E)
The set of points in (E,d cu ) space which satisfy this equation comprise the equilibrium locus. Slope is given by
(A.24) shows that p increases in the d cu direction along this locus. This, combined with the property of decreasing p in E from (A.4) yields a decreasing deposit rate locus in (E, d cu ) space. The locus approaches the vertical axis asymptotically because as E→0 in (A.23), p→∞ so d cu →∞. By plugging in d cu =d m in (A.23), we see that the locus intersects the horizontal axis at some finite point E a when p is positive.
Similarly, setting \(\dot{E}= 0\) in (A.15) and solving for E gives
which defines the equity equilibrium locus. Slope and curvature are found by differentiating
Thus, the equity locus is increasing and convex in (E, d cu ) space.
Equations (A.23), (A.25), and the constraints (A.7) and (A.8) characterize the optimal solution to the four variable model. For any given credit union loan rate, a phase diagram like in Fig. 1 yields the optimal steady state equity and deposit rate. Thus, we can write E ∗ and \(d_{\mathit{cu}}^{\ast}\) as functions of l cu . Solving (A.7) for I, substituting in E ∗ and \(d_{\mathit{cu}}^{\ast}\), and equating to (A.8) yields
The credit union loan rate and investment volume which satisfy this equation are optimal steady state values and are denoted \(l_{\mathit{cu}}^{\ast}\) and I ∗.Footnote 20 As a result of the reduction to two differential equations, this solution method incorporates only steady state values of the loan rate and investment volume. Given \(l_{\mathit{cu}}^{\ast}\), the phase diagram in Fig. 1 describes the evolution of the deposit rate and equity level.
Behavior of the system in each of the four regions formed by the loci is characterized by the perpendicular arrows. In Eq. (A.15), we see that \(\frac{d \dot{E}}{dE} = \tau r > 0\), so for all points to the right of the equity locus, \(\dot{E} > 0\). The behavior of the deposit rate over time is given by differentiating (A.22) with respect to the deposit rate
(A.28) cannot be signed with certainty.
1.5 A.5 Stability and uniqueness of T ∗
The stability of the equilibrium point (E ∗, \(d_{\mathit{cu}}^{\ast}\)) in Fig. 1 is demonstrated here. The general methodology, taken from Ekman (1982), is to linearize the differential equations (A.15) and (A.22) around a neighborhood of the equilibrium point (E ∗, \(d_{\mathit{cu}}^{\ast}\)) and then find the eigenvalues of the resulting system. Negative real eigenvalues correspond to stable manifolds, while positive real eigenvalues are associated with unstable manifolds. The same is true for the real parts of imaginary eigenvalues, but here the dynamic path towards the equilibrium point has a cyclical component. We restrict our attention to real eigenvalues.
We can rewrite (A.15) and (A.22) as
Then the linear part of the Taylor series expansion about the point (E ∗, \(d_{\mathit{cu}}^{\ast}\)) can be written
where
and \(\varphi_{d_{\mathit{cu}}}\), φ E , \(\psi_{d_{\mathit{cu}}}\), and ψ E are all evaluated at (E ∗, \(d_{\mathit{cu}}^{\ast}\)).
The eigenvalue of the main matrix are
We see that the eigenvalues are real and of opposite signs if and only if
From (A.15)
Thus,
Assuming that (A.29) holds, the speed of convergence along the saddle point path T ∗ in Fig. 1 is given by the absolute value of the negative eigenvalue. The influences of the various parameters on the eigenvalue can then be ascertained. Unfortunately, the only signable differential was the change in the eigenvalue with respect to the discount rate. Higher discount rates, ceteris paribus, lead to faster convergence.
The proof of uniqueness is similar to Blanchard and Fischer (1989). For any E 0, the d cu (0) that places the system on T ∗ is optimal. The transversality condition (A.16) is satisfied because E(t)→E ∗ and λ(t)→−V d e −pt τ −1D−1→0 as t→∞. In Fig. 1, if the credit union starts with E 0 and \(d_{\mathit{cu}}'(0)\) is chosen, the deposit rate and equity will evolve along the trajectory T′ and the vertical axis will be approached (or reached) in finite time. Approaching zero equity in finite time violates the first order conditions (A.13) and (A.14). Since members perceive p→∞ when E→0, benefits would have to be instantaneously unbounded to justify full discounting. If \(d_{\mathit{cu}}''(0)\) is chosen, T″ is followed and the credit union approaches d m in finite time, which prescribes zero saver benefit with only finite equity buildup. The transversality condition is violated because λ(t)→0 as t→∞. Similar arguments hold when E 0>E ∗.
The difficulty with analysis of convergence and the strong assumption made in (A.28) highlight an important fact about optimal control theory—its elegant form can rapidly deteriorate under even the most simple specifications. Sydsaeter (1978) points out a number of pitfalls encountered by economists working with optimal control theory. As illustrated above, copious amounts of algebra and calculus are the price for using this powerful technique.
Rights and permissions
About this article
Cite this article
Rubin, G.M., Overstreet, G.A., Beling, P. et al. A dynamic theory of the credit union. Ann Oper Res 205, 29–53 (2013). https://doi.org/10.1007/s10479-012-1246-7
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10479-012-1246-7