非対称情報と公的金融

An empirical research of the behavior of government financial institutions with informational frictions

[Abstract]

In this paper, we investigate the behavior of government financial institutions(GFI) and private financial intermediaries(banks) empirically. Our basic model is one which was proposed in Inoue(1997). We assume that GFI applies average cost pricing and C(0)>0, C'>0, C">0. The major features of the model as follows.

　Each borrower(firm) has one investment project of unit size and some
collateral assets ε. The payoff of the project and the loan are

probability Return of project Loan payment

Good state q θ ｒ

Bad state 1-q 0 ε

　The cumulative distribution function ofε is F and F'>0. The present value of the firm's future business opportunities is v. There are informational asymmetries about the true states of the firms and the verification is costly. Because of limited liability of stockholders, there exists the following incentive compatibility constraint so that the firms would pay the gross loan interest rate r in good states.

ε+v ≧ r …①　 r= ε*＋v …①'

　Providing that the certain gross rate on the funds to banks is ρ and the bank loan market is oligopolistic, the expected profit of banks is

E（π）＝qrL＋（１－q）ε*（L）L－ρL－C（L） …②

ε*＝（ρ＋C'－qv）/（１－ηs） …③　,　η＝－（L/ε*）ε*'＞０　…④

　　The behavioral equation of GFI is

ε*P＝ρP＋CP/L－qv …⑤

　　Comparing Eq.③ and ⑤, it is clear that GFI can provide the fund to the firms excluded from the bank loan market because of less ε. Moreover, GFI's loan might be countercyclical.

　We adopt the quadratic cost function, and the estimated equations are Banks r ＝（１－η）－１｛α－ηv＋ρ＋βL＋（１－q）v｝　…⑥

GFI　 r ＝α＋β/２＊L＋γ/L＋ρP＋（１－q）v　…⑦

Our estimated results of Eq.⑥ and Eq.⑦ are favorable to the model.

Japanese version