Has Competition in the Japanese Banking Sector Improved?



competition, whereas θi,t = Si,t for Cournot competition. From (3) and (4), we obtain:6

1          CC

Pt (1--θi t) = rd + —iʌ + —

(5)


t               i,t        i,t d

ηt                      i,t      qi,t

On the other hand, if the banks are maximizing their joint profits, represented as
Pt (Qt)Qt +n (rib,tbi,t -ri,dtdi,t -Ci,t(qi,t,di,t)), the first order condition can be expressed as
i=1

equation (5) with θi,t = 1 . Furthermore, Si,tθi,t1 represents an array of repeated

game equilibria whose one-shot game is Cournot oligopoly.

Rearranging (5) and defining Ri,tPtqi,t to represent the revenue of bank i generated

by loans, we obtain the following equation:

θ d        ∂Cit      ∂Cit

Rlt = R- ridtqit + qit——+qit——.                   (6)

i,t                i,t       i,t i,t        i,t                    i,t

ηt                   di,t       qi,t

θt represents the degree of competition. Note that we have dropped out the subscript i

for θ in order to capture the industry average degree of competition.7

Because the marginal cost is not observable, we assume the translog cost function:

ln Ctt = ao+ aιln ⅛,t +1 a2(ln)2+a3ln di,t+1 a4(ln di,t)2+a5ln wt +1 aXln wt)2

,           _______, 2          ,__, 2 _______,________ , 2           ,           (7)

+a7(lnqi,t)(lnwi,t) +a8(lnqi,t)(lndi,t) +a9(lndi,t)(lnwi,t) +εiC,t,

where wi,t stands for the wage rate of bank i, εiC,t is an error term, and variables

with upper bars are the deviations from their means. Given equation (7), our regression

6 Equation (3) can be estimated instead of equation (5). Estimation of equation (3), however, did
not generate reasonable results. The reason for the poor results may be that, in the estimation, we
should use as
rib,t the data for securities including stocks, instead of the data for government bonds,
because the revenues from government bonds are not reported. In Japan, however, considerable
parts of stocks are held as the mutual holding, which is not determined based on the short-run
incentives. In addition, although bond yields should include capital gains and losses, they are not
incorporated in the reported data of
rb . Therefore, we present the results based on equation (5).

7 See Bresnahan (1989) for this point.



More intriguing information

1. The name is absent
2. The Challenge of Urban Regeneration in Deprived European Neighbourhoods - a Partnership Approach
3. LAND-USE EVALUATION OF KOCAELI UNIVERSITY MAIN CAMPUS AREA
4. Elicited bid functions in (a)symmetric first-price auctions
5. The name is absent
6. Perfect Regular Equilibrium
7. The name is absent
8. The name is absent
9. Revisiting The Bell Curve Debate Regarding the Effects of Cognitive Ability on Wages
10. Wirtschaftslage und Reformprozesse in Estland, Lettland, und Litauen: Bericht 2001
11. The name is absent
12. Cancer-related electronic support groups as navigation-aids: Overcoming geographic barriers
13. Momentum in Australian Stock Returns: An Update
14. Auctions in an outcome-based payment scheme to reward ecological services in agriculture – Conception, implementation and results
15. Name Strategy: Its Existence and Implications
16. The name is absent
17. Personal Income Tax Elasticity in Turkey: 1975-2005
18. L'organisation en réseau comme forme « indéterminée »
19. The name is absent
20. The name is absent