The financial network channel of monetary policy transmission: an agent-based model

Abstract

The purpose of this paper is to explore the impact of monetary policy shocks on a financial network, which we dub the “financial network channel of monetary policy transmission”. To this aim, we develop a agent-based model (ABM) in which banks extend loans to firms. The resulting bank–firm credit network is structured as determined by plausible behavioral assumptions, with both firms and banks being always willing to close a credit deal with the network partner perceived to be less risky. As our ABM succeeds in reproducing several key stylized facts of bank–firm credit networks, we then assess through simulations how exogenous shocks to the policy interest rate affect some key topological measures of the bank–firm credit network (density, assortativity, size of largest component, and degree distribution). Our simulations show that such topological features of the bank–firm credit network are significantly affected by shocks to the policy interest rate, with such an impact varying quantitatively and qualitatively with the sign, magnitude, and duration of the shocks.

Appendices

Appendix A Parameters and initial conditions

Symbol	Meaning	Value
Parameters:
$N^{F}$	Number of firms	500
$\eta$	Labor productivity parameter	3
$\psi _{w}^{min}$	Nominal wage adjustment parameter (Eq. 4)	0.5
$\psi _{w}^{max}$	Nominal wage adjustment parameter (Eq. 4)	1.1
$\psi _{l}^{max}$	Target leverage adjustment parameter (Eq. 6)	0.2
$l^{max}$	Firms’ maximum leverage	5
$l^{min}$	Firms’ minimum leverage	0.5
$N^{B}$	Number of banks	50
$\kappa$	Banks’ maximum leverage ratio	0.04
$\alpha$	Banks’ capital buffer sensitivity to financial fragility (Eq. 7)	0.5
$t_{D}$	Duration of debt (in periods)	10
$\lambda$	Probability of choosing a bank at random	0.1
$f^{min}$	Minimum fraction of credit to be supplied in each credit deal	0.2
$i^{B}$	Base interest rate	0.02
$\beta$	Interest rate markup parameter (Eq. 8)	0.25
$\gamma$	Risk premium parameter (Eq. 8)	0.2
$f_{RI}$	Fraction of investment randomly shared among firms (Eq. 1)	0.1
$\epsilon$	Investment distribution parameter (Eq. 1)	2
$\phi$	Sensitivity of firms’ markup to a change in market share (Eq. 11)	0.2
$\psi _{\rho }$	Propensity to consume parameter (Eq. 3)	0.02
$f_{RD}$	Fraction of the aggregate demand randomly distributed among firms (Eq. 13)	0.1
$\theta$	Sensitivity of firms’ market share to relative price (Eq. 15)	1
$s^{max}$	Maximum market share	0.04
$\delta$	Proportion of profits distributed as dividends	0.15
$\tau$	Tax rate	0.15
$\psi _{\zeta }$	Sensitivity of government spending to the growth rate (Eq. 19)	0.05
$\zeta ^{min}$	Minimum fraction of government spending (Eq. 19)	0.02
$g*$	Growth rate target (Eq. 19)	0.03
Initial conditions:
$NW_{i,0}$	Firms’ initial net worth	$NW_{i,0}\sim N(10,2)$
$\mu _{i,0}$	Firms’ initial markup	$\mu _{i,0}\sim N(0.15,0.03)$
$l_{i,0}^{*}$	Firms’ initial target leverage	$l_{i,0}^{*}\sim U(1.5,2)$
$NW_{j,0}$	Banks’ initial net worth	$NW_{j,0}\sim N(10,2)$
$R_{0}^{H}$	Households’ initial cash	1,000
$\Gamma _{0}$	Government initial surplus	1,000
$w_{0}$	Initial nominal wage	1
$\zeta _{0}$	Initial fraction of government spending (Eq. 19)	0.04

Appendix B Sensitivity analysis

See Figs. 16, 17, 18, 19, 20, 21 and 22.

Fig. 16

Sensitivity analysis concerning the parameter $\psi _{l}^{max}$

Fig. 17

Sensitivity analysis concerning the parameter $\psi _{w}^{min}$

Fig. 18

Sensitivity analysis concerning the parameter $\psi _{w}^{max}$

Fig. 19

Sensitivity analysis concerning the firms’ initial net worth

Fig. 20

Sensitivity analysis concerning the firms’ initial mark up

Fig. 21

Sensitivity analysis concerning the firms’ initial target leverage

Fig. 22

Sensitivity analysis concerning the banks’ initial net worth