The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
TECHNICAL CHANGE, REGULATION, A N D ECONOMIES OF SCALE F O R L A R G E C O M M E R C I A L BANKS: A N APPLICATION O F A M ODIFIED VERSION O F SHEPHARD'S L E M M A Douglas D. Evanoff, Philip R. Israilevich and Randall C. Merris Working Paper Series Issues in Financial Regulation Research Department Federal Reserve Bank of Chicago June, 1989 (WP-89-11) DRAFT Technical Change, Regulation, and Economies of Scale For Large Commercial Banks: An Application of a Modified Version of Shephard's Lemma by Douglas D. Evanoff * Philip R. Israilevich and Randall C. Merris *The authors are economists at the Federal Reserve Bank of Chicago. Individually they are also affiliated with DePaul University, the University of Illinois-Urbana, and the Asian Development Bank, respectively. Helpful comments and suggestions were provided by Herbert Baer, Jeffrey Brown, and Larry Mote. Betsy Dale provided helpful assistance in data collection. The views expressed are those of the authors and may not be shared by others, including the Federal Reserve Bank of Chicago or the Federal Reserve System. R e l a t i v e Price Efficiency, T e c h n i c a l Change, and Economies of Scale For Large C o m m e r c i a l Banks Recent additions to the c o m m e r c i a l b a n k i n g have functional forms scale bank of the economies role resu l t studies re c e n t bank neoclassical that this Banking facets firms ma y not find that levels. The should be decisions the basis conce r n i n g decis i o n s the for w e i g h i n g may grant Pot e n t i a l a more problems perhaps duality theory Mor e to g e n e r a t e e x t e nsive evidence to exten s i v e regulation the p o s s i b i l i t y subject a in exists p r a c t i c e d by r e g u l a t e d 1 as a a s s u m p t i o n of However, occur. exists specifically, input pric e s cost m i n i m i z a t i o n s i g nificant the m a j o r p r o b l e m with to m a rket raising cost e x c l u s i o n of r e l e v a n t resp e c t subj e c t of operations, consistent with as is m e t h o d o l o g i c a l . the b e h a v i o r are cost f u n c t i o n b a s e d on the m a i n t a i n e d markets. is not generally economies However, have u t i l i z e d cost m i n i m i z a t i o n wit h competitive studies of r e s trictions are s e l e c t i o n and cost costs low output literature cost studies studies and scope, at r e l a t i v e l y than merited. e x p l a n a t o r y variables. previous of scale on b a n k that in and expansion. existing of sample effect on p r o d u c t i o n t h e o r y and flexible in b a n k r e g u l a t o r y in b a n k r e g u l a t o r y to past d uality empirical implications acquisitions Use the The consideration literature e c onomies are e x h a u s t e d resulting policy little and expansion. economies given incorporated to e v a l u a t e c o m p e t i t i v e viability, on g e o g r a p h i c ex t e n s i v e s u ggesting firms. in n e a r l y all that b e h a v i o r to m a r k e t input prices 2 The p u r p o s e large b anks of this utilizing standard neoclassical g e n e r a l i z e d model shadow cost is s u p e r i o r there is also input p r ices w h i c h may a statistical could r e sult an d other inappropriate banks and the more suggest that for m o d e l i n g The to alter Thus, the estim a t e s results traditional found neoclassical We altered. the to see on shadow factor Each of role these of shares, found w h e n m i n i m i z i n g cost of aspects of b o t h of the n e o c l a s s i c a l model. the n e o c l a s s i c a l cost the p r o d u c t i o n b e h a v i o r framework of the imposed by r e g ulation, in p r e v i o u s if different measures the co n t e x t to shad o w of the various to generalized Behavior based restrictive subject the The re s p e c t test generally neglected constraints cost m i n i m i z a t i o n shown The those subsumes in the p r o d u c t i o n of substitution, is e v a l u a t e d w i t h i n analyzed. resulting than for prices or economic v i e w p o i n t c ould also be g e n e r a l i z e d m odel results distortions elasticities to m a r k e t prices. bank production that case. from m a r k e t difference. e conomic m e a s u r e s The deviate in s i g n i f i c a n t l y of scale, change function to the n e o c l a s s i c a l model. economies technical e s t i mates for cost m i n i m i z a t i o n w i t h F rom a t h e o r e t i c a l model cost f u n c t i o n as a s pecial of r e g u l a t i o n - i n d u c e d process. subject is to generate a g e n e r a l i z e d cost allows input prices because study b ank cost large and the input prices, e l e ments of b a n k studies f r a m e w o r k may be biased. is are costs. employing the 3 The next s tudies and section b r i e f l y introduces reviews the l i t e r a t u r e the m o t i v a t i o n for the introduces the g e n e r a l i z e d cost model, app l y are discussed this detailed in S e ction 4. discusses 1. the in S e c t i o n The implic a t i o n s final of the study. and the 3. The on b a n k cost data Sect i o n 2 to w h i c h we empirical sect i o n draws results conclusions are and findings. B A C K G R O U N D AND M O T I V A T I O N Many of the Cobb-Douglas a ra t h e r 1965, cost Wit h of recent account studies range the development the p o t e n t i a l m u l t i -p r o d u c t production, effects. scale These output are (Benston, Humphrey 1987, and M a r s h a l l found studies economies laws These Hanweck, 1984). studies data Analysis (FCA) Program. include a ver y from the average of scope significant to p r o d u c e Federal W hile 1982, 1968, cost forms a into curves, as scale any b e n e f i t s low levels Berger, from of Hanweck, and Gilligan, diseconomies at single over Benston taking as well found that 1984, Smirlock have been offices by unit and Humphrey). analyzed relatively utilized samples for U - s h a p e d and H u m p h r e y Hanweck, economies functional e x h a u s t e d at r e l a t i v e l y constrained (Berger, scale r e e v a l u a t e d b a n k costs and e c o nomies In fact, the and M u r p h y of flexible g e n e r a l l y have fully utilized found (Bell G i l l i g a n and S m i r l o c k for banks banking of output studies have directly of b a n k costs f u n c t i o n and g e n e r a l l y substantial 1972). number early Rese r v e rich small banks System's in cost l i m i t e d numb e r and in most Functional allocation of banks which cases Cost detail, are not FCA 4 representative of insufficient of the size multicollinearity forms, which ( H e g gestad and Mingo, to g e n e r a t e resulting and exclude are most industry results the use from not of f l e x i b l e the v e r y banks likely to be 1978), (those involved may be d e t e r i o r a t e d by over function $1 b i l l i o n in a c q u i s i t i o n s in assets) and m a rket e x p a n s i o n .1 Analyses pr e v a l e n t . of costs Add i t i o n a l l y , the role the significant of technical incorporated (1984) and H u n t e r co u l d be and Timme and find s i g n i f i c a n t of technical important Most (1932) change of the b a n k cost in d u a l i t y and S h e p h a r d functional bank there costs, still be too conditions the function. the shares also large evaluated the it to be scale 1980s hav e duality have p r o d u c e d that Duality the effects. utilized derived in D i e w e r t theory and the advances functional implies is e q u i v a l e n t Additionally, can be consider and s u m m a r i z e d f u n c t i o n pro v i d e s production HT Shaffer as d e v e l o p e d and r e f i n e d by H oteling Wh i l e forms that HT) and to e n h a n c e in the 1970), is e v i d e n c e cost p r o d u c t i o n process fact o r theory restrictive. economies. studies (1986). of f l e x i b l e Stud i e s by hereafter costs In light of the p o t e n t i a l on p r o d u c t i o n and f o u n d (1953, and J o r g e n s o n in recent y ears (1986; scale costs. of t e c h n o l o g y w h i c h have been substantial.2 in d e t e r m i n i n g b a n k developments 1974) in e l e c t r o n i c are not nea r l y as total n e g l e c t on b a n k p r o d u c t i o n into b a n k o p e r a t i o n s on b a n k costs effect c o m m e r c i a l b anks there has b e e n a n ear change advances effect banks for large that u n d e r u s e d may certain of the to that p r o v i d e d by given that dire c t l y via the adopt i o n in m o d e l i n g forms a description (1971, the c o n d i t i o n s hold, differentiation from 5 the cost function--Shephard's generate the dual cost c o m p e t i t i v e markets output level, behaved It cost still trend is h e a v i l y geographic limitations, perspective alter beca u s e the illustrative the purposes that as to compete additional magnitude Thus, in m a r g i n a l Other For The of the d i s t o r t i o n s laws, loan accounts, are likely from the firm's an a d d i tional from w hat example, it w ould assume for in a m a r k e t where on d e p o s i t s role are limi t e d by of p h y s i c a l capital Office p r o l i f e r a t i o n may occur and the p e r c e i v e d the effective in d e t e r m i n i n g While the w ill r e t u r n on p rice true account rest r i c t i o n s behavior. inputs different the b a n k m a n a g e r will constraints. adjustments of of u t i l i z i n g and prices or c o n v e r s e l y capital usuary constraints a b a n k operates affected. input prices cost for deposits is o b v i o u s l y on c e r t a i n d e p o s i t o p t i m u m choice expanded. in requirements, regulatory levels. hold the b a n k i n g limitations, constraints.5 significantly effective in to p r o d u c e well conditions reserve significantly m a y now be will be The is a l l o w e d expenditures, these interest the m a r g i n a l to b e l o w - m a r k e t capital, to the p r e d e t e r m i n e d conditions and p r o d u c t regulation these that include additional limited branching a means by only s u p e r v i s e d and b e h a v i o r allowable of an input may be be w i t h o u t required firm cost m i n i m i z a t i o n toward d e r e gulation, entry barriers. significantly unit the Restrictions requirements, and m a r k e t conditions functions.3 restrained.4 to constrained The and c e r t a i n r e g u l a r i t y Despite industry size f u n c t i o n are is s e riously d o u b t f u l banking. Lemma. of the or for similarly direction p r o d u c e d by b a n k i n g the lead to and restrictions are 6 not obvious, there is little A t k i n s o n and H a l v o r s e n r e a s o n to b e l i e v e (1980; regulation-induced distortions neoclassical electric produce cost utility factor HT d i r e c t l y structure cost the on the role in b a n k i n g found by AH, we b e l i e v e of technology. then H T 's use This may p a r t i a l l y to be c ounter regulation In the next in s l i g h t l y (1971), more Lovell g eneral is shown Lemma (1983), is use d to be a special on the cost effect of this if cost m i n i m i z a t i o n constraints of the neoclassical role of their is n e e d e d in the of results which is a more for d i s t o r t i o n s i n duced by in cost m i n i m i z i n g behavior. a model w h i c h has b e e n u t i l i z e d (1980, 1984), and Toda La u and Yo t o p o u l o s (1976, 1977). It is a modified version in w h i c h cost m i n i m i z i n g b e h a v i o r These prices c r e a t e d by regulation. is not binding. the However, What allows forms by AH and Sickles scale of o r g a n i z a t i o n e x p l a i n some change on shadow input prices. distortions effect change than the n e o c l a s s i c a l model b e c a u s e Shephard's based analyzed estimates se c t i o n we p r e s e n t different case of shown to of cost para m e t e r s , of the r e s t r i c t i v e intuitive. and the r e s u l t i n g the costs cost m odel was is a l t e r e d by r e g u l a t o r y g e n e r a l b a n k cost model w h i c h of the f u n c t i o n may p r o d u c e b i a s e d technology. Evaluating of technical They and tr a d i t i o n a l elasticities. the role elasticities offsetting. f ound that of the t r a d i tional estim a t e s of c o m m e r c i a l banks. structure behavior r e n d e r e d use and p rice evaluated on scale manner companies, shares, h e r e a f t e r AH) inappropriate. significantly biased effects, chan g e function they are of the account for The n e o c l a s s i c a l general model is cost whe n f unction regulation 7 2. The A G E N E R A L I Z E D M O D E L OF B ANK C O S T S - - T H E Theoretical Model AH d e r i v e d the p r o d u c t i o n under industry. directly sha d o w pri c e regulatory In m o d e l i n g apply pertinent their model constraint SP model, is s u m m a r i z e d to of the in this cost m odel production output; From function; the rate equ a l the we include v a r i a b l e s inclu s i v e of our section.6 firms are assumed function to m i n i m i z e g iven by: (1) of input p r ices and quantities, is a w e l l - b e h a v e d n e o c l a s s i c a l Z is a v e c t o r of exoge n o u s variables; Q is and \ is a L a g r a n g i a n multiplier. i marginal to /(X,Z) ele c t r i c power - H[/(X,Z)-Q] P and X are m x 1 v e c t o r s respectively; it to of large banks, SP m o d e l in the L a g r a n g i a n - c o n s t r a i n e d cost L - P'X and a p p l i e d in the U.S. augmented Derivation In the n e o c l a s s i c a l where (SP) the p r o d u c t i o n b e h a v i o r to banking. modifications costs SHADOW-PRICE MODEL first-order conditions of tec h n i c a l ratio for cost m i n i m i z a t i o n , substitution between of prices of the two inputs. inputs That the i and j is is, 8 i / j - 1 , . . .m, where is the m a r g i n a l the price level of input of output as inputs can be Assume now i. Gi v e n the only derived L = P'X conditions exist. of the two / M - f > hRh(P,X) hl = /j (3) aris i n g the m a r g i n a l inputs. S from regulation, From i and j is equal P1 + = The f u n c t i o n to be m i n i m i z e d becomes: for cost m i n i m i z a t i o n , prices c o m b i n a t i o n of costs. constraints inputs and Pj is the o p t i m a l are L a g r a n g i a n multip l i e r s . substitution between effective are i, and the p r e d e t e r m i n e d constraints - |»[/(X,Z)-Q] Rh (h— l,...,n) Kh (h-l,...,n) constraint, that a d d i t i o n a l cost of input input prices, to m i n i m i z e Lagrangian-constrained where product (2) W That 3*! ___________ h = l __________________ the rate to the and first-order of technical ratio of is, P * = Pj + i ^ R j / a x j M PJ* (4) i / j = where P } is the * effective or shadow price of input i. 9 In the reduces absence of b i n d i n g to the n e o c l a s s i c a l technical regulatory constraints, condition whereby s u b s t i t u t i o n equals the ratio of equation the m a r g i n a l market prices (4) rate of of inputs; Pi* i This and special case is r e f e r r e d is n e s t e d w i t h i n The the more sh a d o w prices Th e r e f o r e , of the kj is an input- proportionality. all general inputs Shadow-Price are not and f i r m - s p e c i f i c As n o t e d by AH, first-order shadow-price s h a d o w prices kj — 1 for all neoclassical to as the M a r k e t - P r i c e (SP) directly (1971) (MP) Model Model. observable. and AH (1984), the i * 1, . . . ,m, i n t e r p r e t e d as arbitrary (5) are a p p r o x i m a t e d by Pi* - k t Pi can be .m 1, J fol l o w i n g Lau and Y o t o p o l o u s s h a d o w prices where / i, cost equal functions. the and the the factor of shadow-price Taylor's series When regulation approximations expansions of is nonbinding, r e s p e c t i v e m a r k e t prices, s h a d o w cost function. (6) f u n c t i o n redu c e s i mplying to the 10 Differing from the n e o c l a s s i c a l i n p u t - p r i c e variables, the cost s h a d o w cost function function only in the is given by CS - CS(kP,Q,Z), w h e r e kP is a v e c t o r Applying Shephard's functions of s h a d o w prices Lemma (1970), of inputs. the set of d e r i v e d input demand can be obtained: 3C5 3(k,P,)- L 1 Using (7) equation (8), the firm's (8) total actual cost is 3Cr P'X - I Pi i, M 3 ( W t The shadow differentiation M 1 Rearranging factor-cost shares (9) are o b t a i n e d by logarithmic of Cs: 31nCiS ainCkiP,) equation M,SCS k,Pi kjPjX, 1 .....m . (10) (10), l,...,m, (ID 11 and s u b s t i t u t i n g e q u a t i o n (11) into e q u a t i o n (9) gives, (12) 11 = Taking logarithms, A c “ M,S lnCA - l n C S + I n S il K i = Using also be equations (11) PiX, 1 Comparison combination of the equivalent to are optimal in response factor-cost ,m shares can that the observed and optimal and are d e t e r m i n e d by the observed and o p t i m a l of the firm. the w h i c h may cause The to differ. Thus, to the e f f e c t i v e function of costs input or s h adow is a more to be m i n i m i z e d The the SP model the o bserved costs and o b s e r v e d b e h a v i o r by a l l o w i n g constraints shadow cost (14) Models the b e h a v i o r the p r o d u c t i o n process. Cost equal A c c o r dingly, input prices representation M P it is a s s u m e d of a d d i t i o n a l resulting and and d e t e r m i n e differentiates effective S P of inputs input prices. adopted actual 1, • M , V In the MP model effect (12), obtained, Mi Intuitive and (13) are SP model for the o b s e r v e d and combination input prices. is The comprehensive and is also the appropriate allows one to dual 12 calculate prices, the the optimal P. cost This combination d i f f e rences conditions the procedure resulting differences inputs combination is r e l e v a n t the are regulatory represent SP mod e l constraints. produces C , and s h a d o w different. and are us e f u l optimizing behavior cost, directly an example. eff e c t increases, determined to that ma y the behavior. to be found a c t ually but this shadow in other Bankers may share consider lower than the market in illustrated observe that as Regulation However, is a result viewing may be misleading. on c a p i t a l in the actually true, the more in e v a l u a t i n g in b anking; obsc u r e d the can be and ma y to economize are are interested and shares which relevant increases. using are Cs estimates interested industries. effect ma y be all This shares quantities, at t e m p t is may be capi t a l input prices significantly the capi t a l input prices, increases, to be Ban k r e g u l a t o r s if one of r e g u l a t i o n on factor regulation therefore expressions of Shephard's on o b s e r v e d data observable, or i n f l u e n c i n g b a n k behavior. is this of results. o b s e r v e d behavior. evaluating It empl o y sets are b a s e d of the Bankers We C , estimates incorporating a l t h o u g h not two two meaningful observed constraints. a modified version CA results for f o r e c a s t i n g i n f o r m a t i o n and, similar co m p e t i t i v e in CA c o m p u t e d w i t h P and P* d e r i v e d by u s i n g actual, conceptually the for m e a s u r i n g in S e ction 4. Although both with given observed from p r o d u c t i o n u n d e r cost of the b i n d i n g E s t i m a t i o n of the Lemma, of input and those u n d e r b i n d i n g Correspondingly, measure (unobserved) observed data. influencing effective price as r e g u l a t i o n as cost of capital a direct result 13 of regulation. and the 3. saving.7 policies, EMPIRICAL following Z is based the specified H - Holding on the T - A time Banks 3. The 4. The to AND DATA SOURCES include the are trend the fol l o w i n g vari a b l e s : struc t u r e (H — 1 for b a n k s companies; to acco u n t a single (L) in H * 0 for ban k s and c apital function in o n e - b a n k and kK w o u l d be is s p e c i f i e d the services of two Total form, a continuous shadow cost is spe c i f i e d in shadow prices. i d entical is b e y o n d change. in t r a n s l o g to an a r b i t r a r y factors, input, technological (K). function. linearly homogeneous shad o w price for output by u t i l i z i n g approximation s p e c i f i c but as kj (i * L,K), firm, the are across b a n k i n g and time scope specifed firms. specific, of our e m p i r i c a l The sh a d o w price factor for the s h adow price factor for capital, banks, companies), s h adow cost generality o b s e r v e d data may be misguided. adopted: twice - d i f f e r e n t i a b l e as a c t ually be or an e v a l u a t i o n of specifications comp a n y labor convenient r e g u l a t i o n may used of b a n k offices, prod u c e inputs: of capital to large U.S. multi-bank holding holding i.e., quantity S h a d o w - P r i c e Model empirical B — Number 2. of the Bank r e g u l a t o r y policies, SPECIFICATIONS In a p p l y i n g 1. the p r o d u c t shadow price may decline, capital those Thus, labor, kL, is set as input Ideally, but kL such estimation.8 equal to unity; kK , is a p a r a m e t e r to be 14 estimated. The estimated, given factor kK . cost absolute values that shares Therefore, but can test The total we e q u a tions are h o m o g e n e o u s ca n n o t test for r e l a t i v e shadow cost lnCs -• pQ + PQlnQ + + the of kL and kR cannot be for total of degree for a bsolute price efficiency function in tra n s l o g actual zero p rice cost and in kL and efficiency, (kL kK) . form is (1/2) PQq ( I n Q ) 2 + S ^ l n Q l n d c ^ ) ( l / 2 ) 2 i j i l n ( k 1 i)ln(kj I y j P Pj) + $TlnT + + 0QTlnQlnT + S ^ l n (k,P,) InT + pglnB + + X^ln^P,) ( 1/2 )<> (InT )2 tu (1/2 )PBB(I n B )2 + y ^ l n Q l n B + y ^ l n T l n B + S i i ^n (ki y B Pj) InB + PhH + yHQH l n Q + yHTH l n T + y ^ H l n B + S ^ l n (k,P,) H i - L ,K where y implies — y the . j - L ,K Linear h o m o g e n e i t y following adding-up (15) in shadow p r ices restrictions on p a r a m eters: SjPi ” 1 > S,y1 ” S,y1 ” Sj^iT ~ ^iy i “ ® Q B H = SjVtj - S 1 j 1 “ 0 S yJ Shadow by cost logarithmic shares for the differentiation i >j “ L ,K . translog of C c (16) specification in e q u a t i o n are de ri v e d (15): 31nCs M,S ■ ain(k1 1 " P ) = P + I j i ln(kj i y J PJ) + y iQlnQ + y iTlnT + y i lnB + y ( H B H i,j - L,K. (17) 15 F r o m equations (13), (15), and (17), total actual lnCA - lnCS + I n { ^[P. + ^ ^ ^ ^ ( k j P j ) + S v iTlnT + I V i lnB + B ln C s is given in e q u a t i o n Using are e q u ations (11), } (18) - L,K, (15). (12), (14), and (17), actual - [Pt + I jY ijl n ( k JP j) 2,[P, + + y iQl n Q ZjVijinCkjPj) + + V 1Tl n T Y 1Ql n Q + + Y^lnB Y 1Tl n T + Equation with (18) classical equations because err o r a dditive to be j o i n t l y of the terms adding-up for the co n d i t i o n s and actual to share terms, estimated. One shares equation (19), constitute share share equations. equation equations. of share the We esti m a t e The (19) appended the set of is dropped matrix from the and of the the total empirical shadow-price y 1HH ^ arbitrarily equation deleted labor + “ L,K. system r e s u l t i n g and j o i n t l y labor-share for V ^ H l k , -1/ of the v a r i a n c e - c o v a r i a n c e on the kL — 1, equ a t i o n s disturbance three - e q u a t i o n the choice normalization, of the singularity c a p i t a l -share invariant and one + y iBl n B i,j cost cost g i v e n by M t A the are + V 1QlnQ i,j where costs actual results to the factor.9 drop are 16 Data To estimate the years the banks collected in the U.S. 164 banks The data additional recent years of data in results could be consequences e s t i m a t i o n methods, The Costs, Fede r a l Re s e r v e C, were capital. of and loans of w h i c h banking was are were accounted price for also PL, was and a s s i g n e d to each b a n k Call Report measured as level wage data as the trend, ratio to p lant sample to the capital, data source. on labor and the dollar value sheet results, The n u m b e r some of Call R e p o r t as was Technical the progress The input of Labor collected for each year l o c a t i o n of its home PR, was of phys i c a l and specification. similar H. and five samples T - 1,2,...,16. trends were is differences balance from the B u r e a u according of p h y s i c a l additions Alternative organization, obtained State The price de f i n e d as taken from the a time Statis t i c s . office. was in the a p p e n d i x . 1 0 company for w i t h labor, Q, c o n s i d e r e d and p r o d u c e d B, was of b a n k h o l d i n g the m a j o r some c o n s i s t e d of Therefore, sum of e x p e n d i t u r e s securities. summarized offices, This of d i f f e r e n t the eliminating the d i f f e r e n t m o d e l B a n k output, inv e s t m e n t output measures as the after set for company are more banks included. Call R e p o r t was d e f i n e d as physical type as well set, there for of a h o l d i n g observations. to that used by H T , a l t h o u g h data of data in i n d i vidual years, of 2,624 the f u n c t i o n a pan e l that c o n s i s t e d final of data p r o b l e m s for a total cost w h i c h were m embers entire period. because similar generalized 1972-87 was l a r g e s t banks over the approximated c a pital and equipment, from expenditures furniture, and 17 physical premises, furniture 4. r e p o r t e d b o o k value of net b a n k premises, equipment. E M P I R I C A L RESULTS section, shadow-price technical Results Basic model change, and Table iterated sample.1 1 for in the and OBHCs, the c o e f f i c i e n t and m a r k e t - p r i c e sample e s timates (MP) and for for b o t h models imposed. are met While for all estim a t e s that the general SP m odel are binding. likelihood-ratio the test strong r e j e c t i o n quite not the the p a r a m e t e r globally stable, in our similar. are not suggest The different that b a n k m a n a g e r s than the and g e n e r a l i z e d R 2 two models less the e s t i m a t e d by data po i n t s reveals is s i g n i f i c a n t l y to be indicates (16) appear constraints suggests of scale, respe c t i v e l y . SP and MP m o dels from the more which elasticities Tests condi t i o n s the the de m a n d and s u b s t i t u t i o n are presented. in MBHCs The p a r a m e t e r production capital and e s t i m a t e d for all banks in e q u a t i o n scrutiny sense, input nested within seemingly unrelated regression method with stability .562, hypotheses tested, Hypothesis (SP) restrictions Results and 1 contains shadow-price closer are of banks Results results various are r e p o r t e d subsamples the the and p h y s i c a l In this the to that equivalent. the a d d i tional e s t i m a t e d va l u e from one per c e i v e price of the MP model for kK, in a st a t i s t i c a l the price o b s e r v e d m a rket price. of r elative However, Additio n a l l y , efficiency relative of phys i c a l to (kL k K-l) the more 18 general SP model, T h e r efore, as except sho w n by for dir e c t i n d i c a t i o n of the extent c o m m o n l y u s e d MP model, resu l t s from the test results c o m p a r i s o n of the of the p o t e n t i a l bias our d i s c u s s i o n will in Table 2. two m o dels p r o d u c e d by concentrate to get an the on the the SP model. S cale E c o n o m i e s To for returns actual total test cost w i t h elasticity--was to scale respect in the to o u t p u t --i .e ., d e r i v e d by d i f f e r e n t i a t i o n the the actual of e q u a t i o n ztr'y i Q + 0TQlnT + GgglnB + 9HqH sha d o w cost m e a s u r e equation (20) with the is given by last scale, equal model, Table then PQ * 1 and all zero. the r i g h t - s i d e 2. Table 2 also 1 i k e 1 i h o o d -ratio of h o m o g e n e i t y eve r y w h e r e other and w i t h o u t The h y p o t h e s i s shows the test results these expression constant in based res t r i c t i o n s , returns for the returns in e q u a t i o n to scale r elevant p a r a m e t e r stati s t i c s and h o m o t h e t i c i t y exhib i t s coefficients of c o n s t a n t test (20) (21) Likelihood-ratio e s t i m a t e d wit h rejected. and (18). term deleted, If the p r o d u c t i o n p r o c e s s (20) scale i-L,K. SE s _ 31nCs 01nQ to e l a s t i c i t y of 31nCA ' 'Q P ) 31nQ = Pq + PQQl n Q + 2 1> tl n (k 1 i sea The SP model, testing of the p r o d u c t i o n on the are SP shown in is s t r ongly restrictions and r e j e c t i o n function. 19 Estimated scale dat a points. elasticities Estimates the sample, (2) and (3) at the means For the SP model, elasticities, As shown Tab l e g r a n d means of the results in Panel less returns 3 shows for all derived data (a) observations 3, on OBHCs and for cases, displaying decreasing the m a j o r i t y of them were diseconomies For in states comparative located coefficient estimates P anel (c) of Table the means of the data How e v e r , the were for all to the a test of the two mo d e l s (SE vs. statistically indicate by H T . 0.69 close The m e a n value for OBHCs. scale and OBHCs. were of than unity three p e r c e n t that (b) of the to scale studies for Panel less of the r e v e a l e d that states. This f o u n d greater elasticities scale corresponding for MBHCs, e s timates All of these economies that of SE r e p o r t e d by HT was in at the and for OBHCs f rom the the SP model. estimates however, than were 0.71 of estimates estimates, of scale shown (SE) of the r e s u l t i n g m e a n SEA or SES) s u g g e s t e d s m aller calculated As elasticities o bservations, equality were from the MP model. actual different. considerably subsamples. 3. in unit b a n k i n g scale the quite and OBHC sample, l i m i t i n g b a n k branching. using were entire in indicating of the returns earl i e r purposes, 3, observation the 97 p e r c e n t Clos e r v i e w i n g institutions at specific the e s t i m a t e d a verages elasticities on MBHCs. is c o n s i s t e n t w i t h each in Table for both MBHCs observations finding (1) shadow and actual in all scale calculated the MBHC reported than u n i t y that actual for for of Table to scale also of data for for b o t h SEs and SEA , are S E s and SEA are increasing at the were were reported for MBHCs and 20 In Panel (c) of Table 3, we have model of m e a n values (ASE) o r i g i n a t e d by Benston, e m p l o y e d by H T . direct effect ef f e c t ar i s i n g b a n k offices of an i n c rease from the utilized q -» rA estim a t e s scale is d e s i g n e d (1982) and also the indirect on the number of is d e f i n e d as , 31nC 31nB + 31nB 31nQ * (22) r e l a t i v e l y m inor and i n s i g n i f i c a n t the MP economies and the output ASE for to capture b o t h on cost of h i g h e r in production. indicate for MBHCs scale and H u m p h r e y in output effe c t estimates of a u g m e n t e d Hanweck, The ASE m e a s u r e ASE Our of a m e a s u r e inc l u d e d augmented e c onomies augmented economies of for OBHCs.1 2 Impact of The Technical impact examining tcha the Change of t e c hnical elasticity change on b a n k of cost w i t h costs resp e c t to is e x p l o r e d by the time trend, T. 31nC 81nT - <> + #TTlnT + ZjV^lnCkjP,) *T + 0QTlnQ + 0TBlnB + The t e c h n i c a l -change the right - side S|k i'lyi T i = L ,K . (23) e l a s t i c i t y of shadow cost T C H S is given by expression for the e l a s t i c i t y of actual cost TCH 21 in e q u a t i o n (23) advancement is g a uged by less our with than zero. est i m a t e s Panel (a) progress of within suggest Table 3, for MBHCs and suggests the MBHC p r o gress observations The role the of the bias previous greater economies showed from of the role dissemination the Once of again, the more is the an i n d i c a t i o n of re s t r i c t i v e shown (TCHS— -0.338 of t e c hnical is 91.6 p e r c e n t to compare to get is s t a t i s t i c a l l y finding technology. 3 can be u s e d change in line with E v a l u a t i o n of that the MP m odel of t e chnical is for T C H S. and SP models SP model This from H T . shown in technical However, larger b e n e f i t s in Table As of t e c h n o l o g y form exists. are of the data, progress. than for OBHCs. and the d i f f e r e n c e estim a t e s these m e a s u r e s p r o d u c e d by u t i l i z i n g to the the effect TCH~-0.306) suggest conclusion in the MP Technical at the m e a n s t e chnical at each data point presented Relative understate to w h i c h p r o d u c e d n e g a t i v e values of t e c h n o l o g y model. the found to realize results extent that organization technical were extent our results conf l i c t wit h MBH C s the term excluded. significant in di r e c t the last B a s e d on c a l c u l a t i o n s on average our prio r s the to and significant. Thus, chan g e may be misleading. The i n f luence of t e c hnical change economies e x a m i n e d by d i f f e r e n t i a t i n g the InT. of InT on u n a u g m e n t e d We focus elasticities. to InT, on the effect Differentiating and u sing equation (17) scale on scale elasticity with SEA in e q u a t i o n gives, is re s p e c t to scale (20) wit h respect 22 tchsea 3SEa a2lnCA ” 31nT " 31nC31nT (24) Differentiating SEs as d e f i n e d in eq u a t io n (21) gives, 3SES T C H S E s “ 31nT ” °QT» the (25) first term of T C H S E A in e q u a t i o n Using the grand m e a n data es ti ma te s in Table from 3 imply on averag e flattened time. in wh ic h important. Th es e re sults actually sug ge st s Thus, c o in ci des to b r e a k appears en h a nc ed it e s s e n t i a l l y findings it easier 1986), given over of scale. as a ba r r i er cost the output and make size d e c l in i n g price particularly that as curve has from other the p e r i o d That decreases to pr od u c e advances, the HT findings serves ave ra ge down scale b a rr i e r s of t e ch ni cal econ om ie s the pa ra me t e r led to m o de s t the with re as on a b l e ( Br es na ha n and of T C H S E A and T C H S E S change has t ec hn o l o g y makes co ntr as t with advantageous. our estimates of scale. This This te chn ol o gy for all banks te chn ica l i n c r ea se d e ff ic i e n c y computer had that is u s ed as a means less and SP model, in ec onomies over in du st ri es and the (24). analyzed. tech ni cal is, their it makes change results size more 23 Finally, derived. (19) the effect of t e c hn o l o gy For the sh adow cost on factors can also be differentiation function, shares of equation gives, 3M,S 3lnT ” y i T From Table 1, labor saving Again, this and, The of eq uivale ntly, found Demand in other and ow n-p ri ce constant, e la s t i c i t y the is re la t i ve ex pe cta ti ons to labor. and coin ci des with industries. of de mand for the input i, h o l d i n g Q elasticity 4 displays and OBHC Mi S(M1 l ) + y f Sl £1 _ 3Xj Pj = 3Pj X t - gr an d means the MBH C capital using in e qu a t i o n cr os s- p ri ce % at technology is given by u s i n g Xj as d ef in ed Table that Substitution _ 3X, P t E 3P7 x; The suggesting is in a gr eem en t wi t h wha t has b e e n Elasticities * -0.012 * ( (11) of demand for a c o n s t a n t output of the data is (28) Mis the results subsamples. (27) for these elasticities for all banks Because of the in the c a lc ula ted sample coefficient and for 24 restriction in our tw o - in p u t model, elasticities are equal in ab so lut e value Concentrating on the SP model, we re to two de cimal the same o w n - pr ic e el ast ic i ty in abs olu te value, d e ma nd the of labor cro s s - pr ic e inputs are presented mo d el with the gross Once s ig ni f i c a n t l y Estim at es also sho wn same ei t he r the Regulatory The Costs inputs that are g re ate r are tw o- in p u t from that this the more case, the are also reje cte d general model. elasticities of are give n by (29) than zero, substitutes, and OBHCs b a s e d and effects the own-price from the MP model pa rtial of the as and are v e r y n ea rly on the c o ef f i c i e n t estima te s the from SP or MP model. utilization u n de r than unity MiSMjS ( * 0^) = for MBHCs the indicating those usin g 4, less elasticity the the results (1966) The 1 + estimated for Re sults in Table a required are positive, from but than for estimates and OBHCs. o wn -pr ic e greater in sign. elasticity is ne ga t i ve re q u i re d ag ai n for the A l l e n sub sti tution, The As substitutes. in Table 4. differ input relatively e la st i ci ti e s and op p o s it e for MBHCs estimated demand. and cr o ss - p r ic e the own- pr ice places for each for capital b e i n g elasticity the own- are Factor Shares of r e g u l a t i o n on actual p r o d u c t i o n i nv es t i g a t e d by al ter na t iv e co mp ar i n g a ss u m p t io n s results of relative costs for pric e the and factor SP model e f f ic i e n cy and 25 ine fficiency. un d er Estimated relativ e price actual p r o d u c t i o n efficiency e s t i m a t e d p ar am ete rs of the are Results unde r us i n g es t i ma te d c oe f f i c i e n t s the kK — 0.562. shares, means The e st ima te s shown all banks, re la ti ve price (and average) r el at iv e the 5, of the SP model, MBHCs, e s t i m a t e d re g ul a t o ry efficiency. are ob ta ine d by inclusive and of factor at the grand and OBHCs. costs a ri sin g from great er total than w o u l d have This w o u l d p r e v a i l e d under tran sl ate on b a n k returns effects the to 1.8 p e r c e n t amount effect regulation-induced inefficiency are b a s e d on c a lc u l a t i o n s p r o d u c t i o n costs greater price for all banks, shares except w i t h kR - kL - 1 of total p r o d u c t i o n costs i n e f f i ci en c y price significantly The in Table of the data sets For r e la ti ve and factor c a l c u l a t e d by us i ng SP model, imposed. costs are about into a on e qu ity the same or assets. for MBHCs and OBHCs. As sta te d earlier, misleading el e men ts w hi c h seen w h e n Viewing e va lu a ti ng share regulation when considered.1 3 increases, based to parti es the effect in Table s i g n i f i c a n t l y with the in te re s t e d of capital whil e ba n k ma n a ge rs constraints. in e v a l u a t i n g of r e g u l a t i o n a c tu al ly of this on factor shares. share the decreases ac tu al l y econom iz e of capital, is of capital to 25 percent. with shadow cost o b ser ve d capital cost examp le r e g u l a t i o n - -from 17 the more me a n i n g f u l Thus, in s i g n i fi c an tl y A r e ve a l i n g 5 for CA the on their p e r c e i v e d price regulatory CA could result in fluence behavior. the results in cr ea s es However i n f o r m at io n a n al y z i ng shares are share on p h ys i c a l i n cl u d i n g the capital costs of 26 Potential Explanations The labor, be for Price es t i ma te d v alu e ba nk e rs this alters obviou s pr ic e the q u es t io n exists: of capital c u s t om er s e xp la na tio ns that Why may s pe nd sums theory" co incides wi t h f in a n c i a l w o u l d be capital r ais ed expenditures r e su lt b as is this the for p hy si c al on offic es cu rrent 1989, increase di sc us s ed how effective ex pl anations and v ia ble structure. This ratio may equity Kolari service f r eq ue ntl y co mpete and Zardkoohi). and teller machines. oc cu rre d as a direct on a no n- pri ce One means through a d di ti on al capital in ca pital wh ic h w o u l d not have ba nk ers if requirement. result Finally, and plush Second, then more the The "c om mitment architecture ge n e r at e d w i t h o u t requirement. alternative b a n k m a n a g e m e n t may com me rci al banks. are b i nd i n g An and po t e n ti a l long-haul, debt-equity r e g ul at e d capital (Evanoff to large to Although te nt at i v e o b s e r v e d ornate than w o u l d be sub opt im al the an imposing p h y s i c a l r eq ui r em e n ts of regulation, is the found at ma ny resulting without p e r c e i ve three to of b a n k management. is part of a long run st rategic plan. surroundings The to ge nerate the capital We have is a credible over relative inexpen si ve? to co nv inc e in the m ar k e t behavior exist, institution and will be large that, of p h ys i c a l do ban ke rs relatively First, the price true and the b e ha v i o r as b e i n g of f e r e d here. suggests than the m a rk e t price. cost elements alternative are less in B a n k i n g of kK».562 p er cei ve significantly Inefficiency capital of doing ex penditures 27 5. S U MM AR Y AND C ON CL US ION S Re s ea r c h e r s m ak in g l i t e r a t u r e have alternative scale the data sources, cost contribution st ructures is more basic. in these c o m p e t i t i v e markets the We into the influ enc es of She p ha rd 's Lemma. which "shadow" pri c es because based on this mod e l as a special neoclassical i.e., Using sampl e in the b a n k i n g the v a l i d i t y costs of banks indicated inf l ue nce Our that wh ic h may of re g ul a t o ry in co ntr a s t te chnical there is shares, The taking empirical a m o d i f i e d ve rsi on scale to past pr o g r es s subs um es from m arket The the n e oc l as s ic al su gg e s t that use for sample our of is rejected. economies Te ch ni c a l of costs cons tr ain ts . in ap p r op r i a te is assumption. d e via te results in efficiency. is b a se d on m i n i m i z a t i o n efficiency considered. but, in co rpo ra tes e mp iri ca l specification price of the a n e oc l a s s i c a l industry, and factor s had ow price b e h a vi o r relative Our c r iti qu e pr i ce of this are found for pr og res s was found findings, the results tended of for differen ces of techni ca l progress. the new methodo lo gy, su b st ant ia l, Each employs input prices case. of economies a s s u m p t i o n of cost m i n i m i z a t i o n The mo de l of the mod e l banks, is a fu n da m e n ta l are b a s e d on a mod el w h i c h re fl e ct to ac co u n t c h a r a c t e r i z e d by relative rea son to doubt acc o u nt measures and b r an c h banks. studies. es ti mat e ba n k p r o d u c t i o n re sul ts the It extent of r e g u l a t i o n significant means of unit to the b a n k cost efforts by u t i l i z i n g out pu t measures, fr a m e w o r k b as ed on the Given their and a l t e r n a t i v e m e t h o d o l o g y us ed cost c o n t r i bu t i o ns distinguished and scope, in the rec en t to d e c r e a se scale the to be of 28 ec on om ie s R e su lt s and was most b e n e f i c i a l from the a lt er n at i v e efficiency (the (the MP model) SP model) elasticities of demand subst itu tio n. inf lu en ce modeling re ve a le d and e s t i ma t io n the MP model was The ban k s results on average studies regulation, the attempt implications, reducing the pric e in own- that the results to me as u r e indicate should be and cro ss -p ric e the that the input costs, into the given that the test. sample Future bank i n du c e d by of the the for costs. i n e f f i c ie n c y to p o l i c y m a k e r s al te rin g the l i k e l i h o o d ratio of actual extent of incorporated in ef fi c i e nc i e s costs w i t h o u t relative price in ef fi c ie n cy cost of r e g u l a t i o n 2 percent for the companies. and A l l e n e l as t i c i t i e s r e j e c t e d under and pr ov id e re gu la t or y industry. input u n de r of l a r g e - b a n k p r o d u c t i o n is about should account relative d if fer en ces c o ns t r a in t s st ron gl y suggest es ti ma t io n s and under Ne vertheles s, of re gu la t or y mode l small for to m u l t i b a n k h o l d i n g and c o n c er n e d with saf et y of the 29 FOOTNOTES Shortcomings M in go (1978), M c Nu l ty 2See La wr e nc e di s c u s s i o n s fi na nc ia l ac c ou nt 3The of the (1983) and Shay of the impact industry. for FCA data are d i s c u ss e d Bell the effect in H e g g e s t a d and and G i ll i g a n and S m i r l o ck (1986), and Solom on (1987), of t e le c o m m u n i c a t i o n s and M u r p h y of te c hn o l o gy st a n da rd r e g u l a r i t y (1968) con di tio ns advances to effects include (1) (2) hom og en ei t y, (3) m o n o t o n i c i t y , and Jorgenson on the indirectly in banking. positivity, for a d i s c u s s i o n of the conditions. (1986) ^Empirical pr o bl e m s r e s u lt in g requirement in du st ri e s discussed to em b e dd ed e.g., recognized ac co unt generalized ig n o r ed the p o t e nt i a l but ignored and St anley studies (1974), ( 19 88) -- onl y r e cen tl y has marke ts Studies problem (1988) cost m o d e l - -At kinson and H a l v o r s e n in and C h r i s t e n s e n and have in the use (1983), this b e en in other and the p r o b l e m of re g u l at e d u t i l i t i e s N el s o n and Wohar See the po t e n ti al of the c o m p e t i t i v e for r e g u l a t i o n - i n d u c e d di st ort io ns i n p u t s - -Joskow Kowalewski McMullen While concavity. in the n e oc l a s s i c a l model. the r es u l t i n g biases, (1976). of b a n k costs have from v i o l a t i o n e xp li c i t l y estimati on , Green studies (4) for attempted on scale (1984). and a t te mpt ed of Is r ai l e v ic h incorporated (1980, 1984). into and a 30 5The idea that r e g u l a t i o n inputs has most regulation fuel the c om mon ly b e e n limited AH me n t io n ad j us t m e n t in v a l id it y prices. to to the f ina nc ial studies, we entirely to r e g ul at o ry occur sa tis ficing, attribute of the are d i s c u s s i n g pr ivate dealing w i t h the u n a v o i d a b l e of b a n k see AH (1984, Lau and Y ot op o l o u s 7 The fi nd w h e n to ma rk e t that errors input other cost of could account If r e g u l a t i o n wh i ch of m i n i m i z i n g are not that we is taken m a rk e t the factors such as emphasized r e su l t i n g pr iv a t e events p p . 648-54). (1971) d is c u s s e d the more mode l in this into ineffic ien cy total The AH mo d el bu ilds of rela ti ve example g e n e r a l i z e d model to and sh adow prices It is p o s s ib l e efficiency. cost in leading social cost operations. 6 1 so A the of m a rk e t ex t e r n a l i t i e s pr iv a te company the p o t e nt i a l It shou ld also be the For C o n s i s t e n t with previous data m e a s u r e m e n t m ar k e t utility r e gu l a t i o n s respe ct is restriction. of n o n c o m p e t i t i v e b e h a v i o r difference. of return its u s e f u l n e s s discuss industry. effects. a cc ou nt by b a n k managers, is (1971) c o m b i n a t i o n of rate of e l ec t r i c L em ma wi th d i v er g e n ce or co ns is t en t some effectively the as a result for role and e n v i r o n m e n t a l of She pha rd 's the form of r e g u l a t o r y G r e e n b a u m and H a y w o o d regulation which this However, the p o te n t i a l clauses the op t im a l associated with of public utilities. o b v i o u s l y not example, di storts price upon efficiency. are p r e c i s e l y what we is estimated. 31 8T o the re s p e c t extent to their 9T o check and MP models r e le va n t with the cost e v a l ua t e d in the equations. s p e c if i c a ti o n s case r e s t ri c t io n s the Obviously, The can be the share imply s y m m e t ry the two -i np ut for the invar ian ce their model. 1 The 0 studies as the of results are d ep osi t and H u m ph re y (1965). o u tp ut assets, Se a l e y Others as the sums value c om b i n a t i o n and Li nd le y (1977), the coef fi cie nt s) case fully of in accounts ex pl a i n for 159. controversy or cre di t deposits, of these ba la nce of recent output card (1982), and Bi nk ley (1981, A nu mb e r and m e a s u r e d "i nt er m ed i a r y Shaff er share t w o - i n pu t and H u mp h r e y of loans, tw o- eq u at io n of the accounts, an for that all s u b s t i t u t i o n m a tr i x of ba n k output. Pavel have u t i l i z e d dollar or some (1987), fact across in pu t-price of the Hanweck SP invariant. imposed on the have u t i l i z e d a " p r o d u ct i o n approach" of loans, the of the 9, p. a pp ro pr iat e m e a s ur e s number were that HT could not fully aware a c c o u n t s --e .g ., Benston, Hanweck of property See H T , footn ot e authors concerning This matrix reestimated from the row and column (i.e., case. the e q ua ti ons --i .e . , with r e st r i c t i o n s u b s t i t u t i o n mat rix equations we results follows sy m m e tr y zero with is appropriate. and c a p i t a l -share tw o-i np ut including are h o m o g e n e o u s the b u r d e n of regulation, eq ua t io n deleted. p a r a me t er subsystem, and te chn ol o gy the e st i m a t i o n l a b o r - sh ar e In v ar i a n c e firms the of identical k. v al ues restriction the that (1987) Berger, and Bensto n appr oa ch" me as ur in g securities, sheet 1984, iterns--e .g ., 1985), HT (1986), 32 and Ko l a r i and Za rdkoohi t a k e n a m ul t i p r o d u c t (1987). approach p r o d u c t i o n using va rio us output. the The assets H T , we that separable. The ev en more com ple x value of in t an gib le s (e.g., see measure bank issue. follow in ba nk Evano ff output 1988). large progress. are em phasis banks. is actually of service cost structure d i s t or t i o n s we use evaluating is the presence levels in the like or dollar e x a c e r b a t e d by For c o m p a r i s o n purp os es As Also funds unit on r e g u l a t o r y to that u s ed by others for using output m e a s u r e i n t e re s t e d and approach. of r a i s i n g such as d i f f e r in g We special further case of "production than de c i d in g b e t w e e n p h y s ic a l is in renders Sealey this the it for of an a p pr o p r ia t e The p r o b l e m si milar of b a n k i n g a compelling to use pr e c l ud e studies have for jo i n t ne s s theory the p r o d u c t i o n pr ocess of large banks with t e c h ni ca l to be rec en t and d ol lar val ue me asures on the if one w a n t e d choice measures. unit Like HT we data li mit at ion s assume tested an u n s e t t l e d co ns i d er as outputs. in H T , even approach", ph y s i c a l output m e a s ur e L i n d l e y make what many discussed and have lack of a c on se ns us appropriate earning A d di ti ona ll y, and an output these aspects of costs. nThe first s a t i s f i e d by re q ui r e s condition prices. p ri ce that two r eg u l a r i t y all the requires In the two factor that the input (footnote imp os ed the r e s t ri c ti o n s elasticity c on dit io ns (16). in case fu nc tio n be this is n on ne gat ive , c o n d i ti o n i.e., are Monotonicity shares be positive. cost 3) The co nc a v e is met sec on d order in the input if the Allen 33 i j 1 + The m a x i m u m value > 0 for M^l-Mj) - 0.25, satisfaction of the Yj £ -0.25. From our results, in (16), - -y^ = -.033. condition occurs is met ca p it al to imagine share greater that the SP model this requirement 1 2 It shou ld be scale se rvi ce co u l d e co no my d ema nd simp ly Benston, d InB/ d InQ regression Using elasticity the ef fe ct auxiliary HT r e g r es s io n rea li sti ca lly , It is to have a labor observations that this only a l t e r n a ti v e to note wh i ch violate to the e xt ent dispersed; operations estimated for mea su re that ba nk o t h er w i s e at one coeffic ie nts InQ, of of an a u xi li ary HT obtai ne d an e s ti ma te wi th respect to output. We from the c o e f f i c i e n t In B - a0 + ajlnQ + a2lnT. correlation of the in the au xi lar y of 0.725 for es ti ma te d on InQ in the This s p e c i f i c a t i o n was m o t i v at e d by co nc ern s term Follow ing + OijlnQ + o 2(1/2 ) (InQ) 2. t on br a nc h e s error the ba nk site. alternative it is or i n te r e s t i n g ob t a i ne d an estimate of br a n ch es of output and, This is the MP model. InB — a 0 g r a n d - m e a n data the likel y i n d e p e n de n t given by .97. com me rci al ba n k w i t h is g e o g r a p h i c a l l y from the than ~ in our sample is a pp r o p r i a t e et aL (1982), second order is less emphasized set up the than 97 percent. is less than co n d i ti o n r e qu ir es Thus a large global and the r e s t r i c ti o n s so long as for each data point difficult of s ec on d order therefore, over regression. serial Our 34 estimated the e la sti ci ty el a s t i c i t y was a ss um e d ^Alt ernatively, shares with For can be re s p e ct the which, seen by respect to be eff ec t to Q was — differentiation (SPS) function, -y^, de cl in es wi th from one. As in H T , of r e g ul a t i on on fa ctor of e q ua ti ons with (17) regulation and (19) regulation. d i f f e r e n t i a t i o n of - k« i since y 0.70. constant. to k t- -r ec o g n i z i n g kK decreases as kR de c rea se s the sha dow cost 3k< of B w i t h (17) generates 35 Table 1. E st im a ti on Re sult s Coefficient MP Model -0.930 00 (0.97) SP Mode l -1.104 (1.16) 0.321 (11.1)** (4.06)” 0.573 (4.04)” (1.88) 0.020 (1.87) (30.6)” 0.033 (10.2)” 0.0075 ^LQ (38.1)” 0.048 yL L (23 .4 )” 0.020 Pqq 0.679 0.577 Pq (45.4)” 0.456 Pk yLH ( kK l ) ' ^ the MP and SP Mod el s 0.544 Pl ^IB for (8.31)” 0.0051 (6.5 6) ” -0.0003 (0.52) -0.00005 (0.11) -0.0034 (1.88) -0.0020 (1.53) -0.018 0BB -0.012 (2.49)* 0.106 (1.75) 0.033 % (14.0)” 0.151 yn (8. 56 )” (4.30)” 0.025 (3.28)” 0.070 (0.39) 0.072 (0.40) -0.078 (0.59) -0.132 (1.01) -0.253 (11.4)** PH -0.237 *TT (10.7)” ®QB -0.0090 (1.75) -0.0044 (0.86) ®QH 0.023 (1.63) 0.021 (1.44) ®BH -0.062 (6.27)” -0.056 (5 .5 7 )” 0QT 0.024 (10.7)” 0.027 (2. 62 )” ®TB 0.0018 (0.29) 9th -0.084 (4.50)" 1.00 kK Generalized R2 Note: -0.0012 -0.085 0.562 .947 P a re n t h e t i c a l entries are r a t i o s - ■* (**) in di c a ti n g (0.19) (4.55)** (7.77)** .948 absolute va lues of a s ym pt oti c tsi gn i f ic a n c e at the 5% (1%) level. 36 Table 2. Results of H y p ot h e si s Tests Numb er of R es t r i c t i o n s Null Hypothesis Market-Price Model Shadow-Price (J) R e st r i c t i o n s LR X2(J) 1 kK = 1 27.30 6.63 Model 6 PQ * 1 27.30 16.81 7 QT 0 138.69 15.09 ®B Q 127.65 13.28 with: (a) C on s ta n t Returns to S ca le y L< r “ (b) Homogeneity 5 p q< t 0QB“ y Lq“ Homotheticity 4 ®QH “ PQ Q ®QB= (c) V ®QH: 0QT” 0 = 0 Note: the LR denotes the l ik el i h o o d - r a t i o l o g - l i k e l i h o o d ratio); for J de grees of freedom, X2(J) at the test denotes the on e- pe r ce n t st at i s ti c c h i - s qu a r e (-2 times value significance level. 37 Ta ble 3. Panel a. Scale E la st i ci ti es and Measu re s of T e c h n i c a l Change Sh ado w- Pri ce Model _________________________________________ Number SES Banks of Obs . All banks 2624 0.924 MBHC s 1685 (7.47)" 0.935 OBHCs 939 (5.87)" 0.906 TCHS TCH a 0.921 -0.338 -0.329 0.931 ( 2 3. 8 7 ) ” -0.392 -0.384 0.902 (2 3.71)" -0.241 -0.232 SE a (6.81)" Panel b . S ha do w- Pri ce O b s . such Number Banks (1 4.20)" Model that SES < 1 Perce nt O b s . such Number that T C H S < 0 P e rce nt All bank s 2580 98.3% 2404 91.6% MBHC s 1651 98.0% 1596 94.7% OBHCs 939 100.0% 808 86.0% ASE TCH Panel c . Ma rk e t -Price Model Number of O b s . Banks SE All banks 2624 0.916 (8 . 37 ) ” MBHCs 1685 0.928 (6.52)" OBHCs 939 0.895 (7.67)" Note: 0.978 (2.34)* -0.306 (22.67)" 0.972 (2.66)** -0.359 (2 2.53)" 0.991 (0.74) -0.210 (12.73)" Esti mat es are d er iv e d u s in g the g r a n d - m e a n value s of the obs er va ti o ns (Obs.) for all banks, m u l t i - b a n k h o l d i n g compa ni es (MBHCs), and o n e- ba nk h o l d i n g co mp an i e s (OBHCs). Pa re n th e t i c a l entries are absolute value s of asymp to tic t-ratios. Tests of the eq uality of the e s t i m a t e d scale c el as t i c i t y me as u re (SE vs SE ) and effect of technical change (TCH vs TCH ) in the two models were rejected; the ca l c ul at e d t-ratios were 8.5 and 6.1, respectively. 38 Table 4 E st ima te d E la st i c i t i e s Price Elasticity of D em and and S u b s t i t u t i o n Re lative Price E f f i c i e n c y Not Im posed in E s t i m a t i o n (S h a d o w - P r i ce Model) R e l a t i v e Price Effici enc y I m p o s e d in E s t i m a t i o n ( M ar ke t-P ri ce Model) All banks -0.1123 -0.5601 0.1123 0.1774 0.6284 0.5601 -0.1114 -0.1763 -0.6280 -0.5603 -0.1139 -0.1794 -0.6292 1K fK * -0.1774 -0.6284 T \l -0.5597 MBHCs \ k OBHCs % E l a s t i c i t y of Substitution OBHCs °LK % 0.7375 ( 4 2. 7 1 ) ” 0.7393 0.7366 (8.48)” MBHCs °IK 0.7407 (8.65)’ * All banks ( 4 1. 9 3 ) ” 0.7431 (8. 95) ” Note: E sti ma tes are for all banks 0.7391 ( 4 4. 0 5 ) ” de ri v e d usin g gr a n d -m e a n v a lu e s of the and the MBHC and OBHC subsamples. data 39 Tab l e 5. Es ti ma t ed Costs and Fa ctor Shares Fitte d value wi t h F i t t e d value with r el at i v e - p r i c e i n ef f i c i e n c y r e l a t i v e - p r i c e effici enc y _________ kK » .562_________________________ k = 1 ________ All banks Tot a l cost Lab o r sh are ML A 75.91% 82.92% MLS 84.84% 82.92% share MKA 24.09% 17.08% M kS 15.16% 17.08% Ca pi tal CA $77 ,84 4.3 5 $ 76 ,45 2. 69 M B H C s $7 1 , 23 3 . 44 Total cost La bor share M L A 76.05% m ls $6 9, 94 8 . 1 1 84.93% Cap i ta l CA share MKA 23.95% M s nK 83.01% 15.07% 16.99% OB HC s Total cost Lab o r share ML A 75.66% m ls 84.67% Ca pi ta l CA $89 ,70 7.3 6 24.34% M s nK Note: share MRA $8 8, 12 4 . 9 1 82.74% 15.33% 17.26% Esti mat es are de ri v e d from the g r a n d - m e a n v al ues of the data for all banks and the MBHC and OBHC subsamples. 40 LI T E R A T U R E CITED Atk ins on, Scott E., and R o b e r t Halvorsen. "P a ra m e t ri c E ff ici en cy Tests, Economies of Scale, and Input D e m a n d in U.S. Electric Power Generat ion ." International E c o n o m i c R e v i e w 25 (October 1984), 647-62. ____________________ . "A Test of Re la t i ve and A b s o l u t e Ef f i c i e n c y in R e g u l a t e d Utilities". The Review E c o n o m i c s a n d S t a t i s t i c s 62 (F ebruary 1980), 185 - 96 Bell, Fr e d r i c k W. , and Neil Banking: Its A Relation Boston: Quantitative to B a n k B. Analysis Regulation. Federal R es e r v e Murphy. of Bank Costs Behavior Price of in C o m m e r c i a l and R e s e a r c h Re p or t No. 41, Ba nk of Boston, 1968. Benston, George J. "Econo mi es of Scale and M a r g i n a l Ba nk i ng Operati ons " .T h e N a t i o n a l B a n k i n g R e v i e w 2 (June 1965), 507-49. Costs in ____________________ . "Econo mi es of Scale of F i n a n ci a l In s tit ut ion s. " Journal of M o n e y , Credit, a n d B a n k i n g 4 (May 1972), 312-41. Benston, George J., G e r a l d A. H a n w e c k and Da v id B. Humphrey. "Scale Economics in Banking: A R e s t r u c t u r i n g and Re a ss e ss m e n t . " J o u r n a l o f M o n e y , C r e d i t 5 a n d B a n k i n g 14 (November 1982, Part 1), 435-56. Berger, A l l e n N., G e ra l d A. Hanweck, and D av id B. Humphrey. "C o mp et iti ve V i a b i l i t y in Banking: Scale, Scope, and Product Mix Economies." J o u r n a l o f M o n e t a r y E c o n o m i c s 20 (December 1987), 501-520. Bre snahan, Timothy. "M ea su r i ng the Sp illovers From Te ch nic al Advances: M a in f ra me Comp ut ers in Fi na ncial Servi ce s." American Economic Review 76 (September 1986), 742-55. Ch r ist en sen , Laurits R., and W i l l i a m H. Greene. "Economies Scale in U.S. El ect ri c Power Genera ti on. " Journal of P ol i t ic a l E c o n o m y 84 (August 1976, Part 1), 655-76. of Diewert, W. E. " Ap pli ca tio ns of Du ality Theory" in M. In tr il iga to r and D. Kendricks, editors. F r o n t i e r s in Q u a n t i t a t i v e E c o n o m i c s , vo lume 2. Amsterdam: North Holland, 1974. Evanoff, Douglas D. "Bank D e pos it Rate D e r e g u l a t i o n Servi ce Levels." J o u r n a l o f C o n s u m e r A f f a i r s 23 (F o rt h co mi n g 1989). and Customer ____________________ . "Branch B a nk i n g and Service A c c e s s i b i l i t y " J o u r n a l o f M o n e y , C r e d i t , a n d B a n k i n g 2 0 (May 1988) , 191-202. 41 Gilligan, J oi n t Journal Thomas, and M i ch a e l Smirlock. "An E m p ir i c a l P rod uc tio n and Scale Economies in C o m m e r c i a l o f B a n k i n g a n d F i n a n c e 8 (March 1984), 67-76. Study of Banking.” Gilligan, Thomas, M i ch a el Smirlock, and W i l l i a m Ma rshall. and Scope Economies in the M u l t i p r o d u c t B a n k i n g Firm." J o u r n a l o f M o n e t a r y E c o n o m i c s 13 (May 1984), 393-40 5. Green bau m, Stuart the F ina nc ial Credit* a n d I., and Charles F. Haywood. Services I n d u s t r y . ” J o u r n a l B a n k i n g 3 (May 1971), 571-89. "S ecular of "Scale Change in Money* He g ges ta d, A rn ol d A., and J o h n J. Mingo. "On the U s e fu l n e ss F u n c t i o n a l Cost A na l y s i s D a t a . ” J o u r n a l o f B a n k R e s e a r c h 8 (Winter 1978), 251-56. Hotel lin g, H. "Edgewo rth 's T a x a t i o n Parad ox and the Nat ur e D e m a n d and Supply F u n c t i o n s . ” J o u r n a l o f P o l i t i c a l E c o n o m y 40 (October 1932), 577-616. of of Hunter, W i l l i a m C., and St ep he n G. Timme. " T ec hn ica l Change, O r g a n i z a t i o n Form, and the Structure of Ba nk Produ ct ion ." J o u r n a l o f M o n e y * C r e d i t * a n d B a n k i n g 18 (May 19 86) , 152-66. Isr ail evi ch , Philip, and K. J. Kowalewski. "The Ef fect of R e g u l a t i o n on Ohio E le ct ri c U t ili ti es. " Economic R e v i e w , Federal R e se r ve Bank of Cl e ve l a n d (Quarter 1, 1987), 10-19. Jo rge nso n, Dale W. "E co n o me t r i c Methods for M o d e l i n g Pr oducer Behavior.” in Z. Gr ili c h es and M. D. I n t r i l i g a t o r ( e d s . ) , H a n d b o o k o f E c o n o m e t r i c s , Volume III. E l se vi er Science Publishers, 1986. Joskow, Paul L. "I nf la t io n and E n v i r o nm e n t al Concern: Structural Change in the Process of Public U t il i t y Price R e gu l at io n." J o u r n a l o f L a w a n d E c o n o m i c s 17 (October 1974), 291-327. Kolari, and Lau, James, and A sg h a r Zardkoohi. . L e x i n g t o n Mass.: Performance Bank D.C. Costs* Structure* Heath, 1987. L aw r en c e J., and Pan A. Yotopoulos. "A Test For R e lat ive E f f i c i e n c y and A p p l i c a t i o n to Indian A g r i c u l t u r e . " American Economic Review 61 (March 1971), 94-109. Lawre nce , Colin, Innovation* and Ro be r t Regulation* Cambridge, Mass.: and the Shay, editors. Monetary Economy Ballinger, Technological . 1986. Lovell, C. A. Knox, and R ob i n Sickles. "T esting E f f ic i e n cy H y p o t h e s e s in Join t Production: A P a r a me t r i c Ap pr ach." R e v i e w o f E c o n o m i c s a n d S t a t i s t i c s 6 5 (February 1983) , 51-58. 42 McN ulty, James Proceedings of Competition. E. a "Economies Conference on of Scale: Bank Federal R e s e r v e Structure D i s c u s si o n . " and Bank of Chicago (May 1983) 456-61. Nelson, Ra nd y A., and M a r k E. Wohar. "Regulation, Scale and P ro d u c t i v i t y in S t e a m - E l e c t r i e G e ne r a t io n . " International E c o n o m i c R e v i e w 24 (February 1983), 57-79. Economies, Pavel, Christine, and Paula Binkley. "Costs and C o m p e t i t i o n Bank Credit Cards." E c o n o m i c Perspectives. Fed er al Re se r ve Bank of Chica go (March 1987), 3-13. Sea l ey C. W. Jr., and James T. Lindley. "Inputs, Outputs, T h eo r y of P r o d u c t i o n and Cost at D e p o s i t o r y Fi n a n ci a l Institutions. J o u r n a l o f F i n a n c e 32 (September 1977), 1251-66. Shaffer, Sherrill. of F irm Sizes." 467-76. "Competition, ____________ . E co nom ie s in M u l t i p r o d u c t R e s e a r c h 36 (1984), 51-58. Bulletin of "Scale Economic Applied Ec onomies of Scale, 17 (June 1985), in and a and Diversity Economics Firms." ____________ and Ed mond David. "Economies of S u pe r s c a l e and In te r st a te Ex pansion." U n p u b l i s h e d m a nu sc ri pt , Federal Reserve Bank of New York, (October 1986). Shephard, R. W. Princeton: Solomon, The Theory Cost and Production Functions. P ri n c e t o n U n i v e r s i t y Elinor H. Public of Policy Electronic Issues. Funds Boston: Press, Transfers and 1970. Payments: K l u w e r - N i j h o f f , 1987. Tod a Yasushi. " S u b s t i t u t a b i l i t y and Price D i s t o r t i o n in the Demand for Factors of Production: An Empi ri cal Es t im a t i on . " Applied Economics 9 (Septe mb er 1977), 203-17. _____________ . "E st im a ti on of a Cost F u nc ti on W h e n the Cost is Not Minimum: The Case of Soviet M a n u f a c t u r i n g Industries, 1958-1971. " T h e R e v i e w o f E c o n o m i c s a n d S t a t i s t i c s 5 8 (August 1976), 259-268. Appendix 1. Testi ng the A p p r o p r i a t e n e s s U s i n g A l te r n a t i v e of the R e s t r i c t e d Model Ou tput M e a s ur e s k Out p ut Me as ur e t hlr a Total Assets .554 6.93 36.08 .02(10)-7 Total Loans .520 7.02 35.98 .02(10)-7 Total Loans .569 6.46 31.36 .02(10)-6 Securi tie s .581 6.05 27.59 .02(10)-5 the t-ratio and Deposits Total L o a n s ,D e p o s i t s , and In v es t me nt Note: t denotes as ymp t o t i c h y p ot h es is of k equal h o o d ratio test of are to unity; statistic; in the Ch i- sq u a re a sym pt oti c s ta nd ar d for is t e st i n g the and a is the test. errors. the null likeli s i g n i f ic a n c e Parenthetical entries level