저자 : 주동헌

In this study we evaluate whether the introduction of financial frictions into the
new Keynesian DSGE model improves the model as a data generating process.
To this end, we construct a DSGE model which incorporates both of financial
frictions between the depositor and the financial intermediary and between the
financial intermediary and the firm. The latter is reflected in the credit risk premium
that borrowing firms have to pay depending on their net worth and the
former is reflected in the implicit credit constraint that is imposed on the firm
exogenously. Using the major macroeconomic data of the Korean economy, financial
frictions in the model are evaluated in an indirect way by DSGE-BVAR(ˆλ)
estimation, which provides two model misspecification measurement: marginal
likelihood of DSGE-BVAR model and optimal DSGE prior weight, ˆλ. It is found
that incorporating the firm’s financial accelerator mechanism improves the DSGE
model as a data generating process and modeling the financial intermediary’s balance
sheets effect in the same model does little harm.


Contents

1 Introduction 1

2 Model 3
  2.1 Households  4
  2.2 Financial Intermediaries  5
  2.3 Entrepreneurs  8
  2.4 Capital Producers  10
  2.5 Retailers  11
  2.6 Equilibrium Conditions  12
  2.7 Financial Accelerator Channels  13

3 Calibration and Estimation results 14
  3.1 Calibration  14
  3.2 Estimation  16

4 Evaluation of Financial Accelerators 19

5 Summary and Further Research Direction 25

A Equilibrium Conditions 31

B Log Linearized Model 33

C DSGE-BVAR Estimation Method 34
  C.1 Likelihood Function  34
  C.2 Prior Distribution  35
  C.3 Posteriors  37
  C.4 The Fit of DSGE Model with λ  38

D IRFs of the DSGE Model 39

E Comparison of Financial Accelerator 43

F Comparison of DSGE and VAR IRFs 44