Financial Frictions, Financial Shocks, and Aggregate Volatility

• Published date: 01 September 2019
• Author: CRISTINA FUENTES‐ALBERO
• Date: 01 September 2019
• DOI: http://doi.org/10.1111/jmcb.12554

DOI: 10.1111/jmcb.12554

CRISTINA FUENTES-ALBERO

Financial Frictions, Financial Shocks, and

Aggregate Volatility

The Great Moderation was accompanied by an increase in financial volatil-

ity. We explore the sources of these divergent patterns in volatilities by

estimating a model with time-varying financial rigidities subject to struc-

tural breaks in the size of shocks, the monetary policy rule coefficients, and

the average size of the financial rigidity. Institutional changes are key in

accounting for the Great Moderation and in shaping the transmission mech-

anism of financial shocks. The increase in financial volatilities is accounted

for by larger financial shocks, but the vulnerability of the economy to these

shocks is significantly alleviated by the estimated changes in institutions.

JEL codes: C11, C13, E32, E44

Keywords: cyclical volatilities, financial frictions, financial shocks,

structural breaks, Bayesian methods.

SINCE THE MID-1980S,THE U.S. economy has been character-

ized by a decline in the volatility of aggregate economic activity, often referred to as

the Great Moderation. We document that, contemporaneous to the Great Moderation,

there was a widespread increase in the volatility of financial variables. Moreover,

during this period, there were substantial changes in financial factors in the United

States, including regulatory changes, the development of new financial products and

techniques, and periods of heightened financial stress. Weargue that a comprehensive

analysis of the drivers of the divergentpatterns in volatilities should incorporate finan-

cial factors, in addition to the traditional drivers of the Great Moderation: good luck

and good monetary policy. In this paper,we evaluate the role of each of these drivers

in accounting for the empirical evidence by means of an estimated dynamic stochastic

general equilibrium (DSGE) model with financial frictions and financial shocks.

I thank Frank Schorfheide, Jes´

us Fern´

andez-Villaverde, Maxym Kryshko, Leonardo Melosi, Raf

Wouters, and John Roberts for their comments and suggestions and Egon Zakrajˇ

sek for sharing the

estimated series for marginal bankruptcy costs. The views expressed in this paper are solely the responsi-

bility of the author and should not be interpreted as reflecting the views of the Board of Governors of the

Federal Reserve System.

CRISTINA FUENTES-ALBERO is in the Board of Governors of the Federal Reserve System, Division of

Research and Statistics (E-mail: Cristina.Fuentes-Albero@frb.gov).

Received September 22, 2016; and accepted in revised form July 31, 2018.

Journal of Money, Credit and Banking, Vol. 51, No. 6 (September 2019)

C

2018 The Ohio State University

1582 :MONEY,CREDIT AND BANKING

The model integrates the financial accelerator framework of Bernanke et al. (1999)

into a standard model with nominal and real rigidities as in Justiniano et al. (2010).

In our model, loans are extended at a premium over the risk-free rate because there

is asymmetric information between borrowers and lenders. This external finance pre-

mium depends upon the balance sheet position of firms and the size of the financial

rigidity, which is summarized by the marginal bankruptcy cost parameter. We intro-

duce financial shocks affecting the firms’ net worth and marginal bankruptcy costs.

In this environment, the institutional framework defining the debtor–creditor rela-

tionship and product innovation is captured by the size of the financial rigidity at the

steady state, while financial shocks capture changes in financial factors at business

cycle frequencies. To test the role of institutional change and changes in the size of

exogenous shocks in accounting for the divergent patterns in cyclical volatilities, we

estimate the model subject to structural breaks in the monetary policy coefficients, the

size of the financial rigidity at the steady state, and the size of shocks. Our economet-

ric approach generalizes the work of C´

urdia and Finocchiaro (2013) by incorporating

steady-state parameters in the set of parameters subject to structural breaks.

Unlike many articles in the literature, we conclude that there was no unique driver

of the Great Moderation. In particular, in our model, the slowdown in real and

nominal volatilities was driven by a combination of good luck, in the form of smaller

economic shocks, and better institutions, in the form of a more proactive monetary

authority and an easier access to credit for firms. The increase in the volatility of

financial variables was driven by larger financial shocks hitting the U.S. economy.

However, despite the increase in the size of financial shocks, we show that, thanks

to the institutional changes in the mid-1980s, the relative role of financial shocks in

driving nonfinancial variables has not increased and the transmission mechanism of

financial shocks is significantly more muted. Therefore, we conclude that the tighter

monetary policy regime and the improvements in the financial intermediation process

have safeguarded the United States from financial disturbances.

A vast literature has examined the sources of the Great Moderation. Most of the

contributions using structural vector autoregressive(VAR) models, such as Stock and

Watson (2002) and Primiceri (2006), attributethe slowdown in aggregate volatility to

good luck in the form of smaller innovations. However, using simulated data from a

DSGE model with a change from passive to activemonetary policy, Benati and Surico

(2009) estimate a structural VAR model and show that the good luck and good policy

hypotheses are close to observationally equivalent in the context of a VAR model.

Researchers using DSGE models, such as Clarida et al. (2000), Lubik and

Schorfheide (2004), and Boivin and Giannoni (2006), conclude that the main driver

of the Great Moderation is good policy in the form of a stronger response to inflation

by the monetary authority. However, Smets and Wouters (2007) and Justiniano and

Primiceri (2008) conclude that good luck is the source of the Great Moderation.

While the former constrained the parameter set at the boundary of the determinacy

region, the latter allows for indeterminacy of equilibria but introduces time-varying

volatilities in the structural innovations. Jermann and Quadrini (2006), focusing on

corporate debt and equity financing, and deBlas (2009), using a neoclassical model

CRISTINA FUENTES-ALBERO :1583

with limited participation and rigidities in the supply of capital, conclude that a reduc-

tion in the level of financial rigidities combined with smaller shocks can slowdown

the volatility of real variables.

This paper contributes to the debate about the sources of the Great Moderation by

analyzing jointly the role of shocks, the conduct of monetary policy, and financial

rigidities in the demand of capital. This paper also contributes to the literature by

documenting the increase in financial volatilities in the U.S. corporate sector and

extending the novel estimation methodology of C´

urdia and Finocchiaro (2013).

The paper is structured as follows. We present the empirical evidence that

motivates the paper in Section 1. We provide an overview of regulatory changes,

innovations in technology and financial engineering, and episodes of heightened

financial uncertainty or financial shocks beginning in the mid-1980s in Section 2.

We describe the model in Section 3. We explain the estimation procedure and model

evaluation in Section 4. We analyze the role of institutional change in accounting for

the empirical evidence and in the transmission of financial shocks in Section 5. We

conclude in Section 6.

1. EMPIRICAL EVIDENCE

The stylized fact motivating this paper is that there is a divergent pattern in the

evolution of cyclical volatilities of macroeconomic and financial variables during

the Great Moderation. In this section, we revisit the dating of structural breaks in

macroeconomic volatilities and provide evidence of the slowdownin macroeconomic

volatility contemporaneous with the increase in financial volatility.

To identify the existenceof and to date the structural breaks, we first run univariate

Bai and Perron (1998) tests, which allow for multiple breakpoints at unknown dates,

on unbiased estimators for the residual standard deviation of the cyclical component

of real and nominal variables1over the 1954–2006 period.2Table 1 shows that we can

reject the null of parameter constancy for real variablesaround 1984. For nominal vari-

ables, the null is rejected for two different periods: the early 1970s and the early 1980s.

We use a multivariate approach to decide whether to consider one or two structural

breaks. In particular, we test for multiple structural breaks as in Qu and Perron

(2007) in the covariance matrix of a VAR model for the cyclical component of

output, consumption, investment, hours worked, wages, inflation, the federal funds

rate, spreads, and corporate net worth and we identify two breaks: one in 1969:Q4

1. We run an autoregressive model of order 1 (AR(1)) on the cyclical component of a variable,

assuming that the error of the AR(1) model, εt, follows a normal distribution. Then, we can ensure that

|εt|√π/2 is an unbiased estimator for the residual standard deviation of the cyclical variable. Giventhat

the cyclical component of a series has zero mean by construction, testing only for breaks in volatility using

Bai and Perron (1998) is not subject to the size distortions put forward by Gadea-Rivas et al. (2014).

2. We use pre-Great Recession data to avoid distortions caused by nonlinearities induced by the zero

lower bound on the federal funds rate, binding downward rigidities, and upward pressures on financial

volatilities during recent years.