Tracking Banks’ Systemic Importance Before and After the Crisis

• Published date: 01 June 2015
• Author: Piergiorgio Alessandri,Andrea Zaghini,Sergio Masciantonio
• DOI: http://doi.org/10.1111/infi.12068
• Date: 01 June 2015

Tracking Banks’Systemic

Importance Before and After the

Crisis

Piergiorgio Alessandri, Sergio Masciantonio

and Andrea Zaghini

Banca d’Italia, Rome, Italy.

Abstract

We develop a methodology to identify and rank ‘systemically important

financial institutions’(SIFIs). Our approach is free of discretion and

based entirely on publicly available data, thus filling the gap between

the judgment of the regulator and the views that market participants

may form with their own information. We apply the methodology to

three samples of banks (global, EU and euro area) for the years 2007–

2012 and find that both the nature of systemic importance and its

geographical distribution have changed significantly over time, with a

shift from advanced western economies towards emerging markets,

particular ly China. Moreover, a distinc tive consequence of the crisis is

theproliferationofSIFIsthatareatthesametimecomplexandpoorly

substitutable, a combination that seemed relatively rare at the onset of

the crisis and might not be entirely consistent with regulatory aspira-

tions for a less risky financial framework.

The authors would like to thank Olivier de B andt, Giorgio Gobbi, G iuseppe Grande and Stefa no

Siviero for helpful discussions an d suggestions. An earl ier version of the paper was c irculated under

the title ‘Everything you always wanted to know about systemic importance (but were afraid to ask)’.

The views expressed do n ot necessarily reflect those of the Bank of Italy.

International Finance 18:2, 2015: pp. 157–186

DOI: 10.1111/infi.12068

© 2015 John Wiley & Sons Ltd

I. Introduction

The term ‘systemic importance’entered the economic le xicon only recently. It was

the collapse of Lehman Brothers in 2008 that showed how the failure of a single

financial institution –one that is not n ecessarily large but deeply interconnec ted –

could endanger financial stability worldwide. While the ‘too big to fail’prob lem had

already been identified by both academics and regulators, the issue of effect ively

defining, measuring a nd modelling system ic importance gaine d attention only after

the crisis erupted.

The methodolog y proposed by the Ba sel Committee on Bank ing Supervisi on

(BCBS 2011, 2013) to identify global systematically important financial inst itutions

(G-SIFIs) carries important practical implications because all banks identified as G-

SIFIs will be subject to cap ital surcharges and en hanced prudential supe rvision

starting in 2019, in an effort to make the default of a system ically important

financial insti tution less likely. Effective regulati on requires that banks and ma rket

participants h ave a good understan ding of its procedures. T he proposed BCBS

methodolog y, however, is insufficiently transparent; it cannot be replicated owing

to its reliance on unreleased sup ervisory d ata, supervi sors’subjective discreti on and

important undisc losed procedural det ails.

To overcome these flaws, we con struct a systemic i mportance (SI) inde x using a

non-discretionar y procedure, and base d exclusively on public data. We therefore

provide a methodolo gy by which mar ket participa nts can directly asse ss banks’

systemic importa nce in a way th at is consiste nt and directly comp arable with th e

financial stabi lity board (F SB) assessment. Our index e xtends that of Mas ciantonio

(2013), who implements the BCBS procedure using only publicly available data, and

whose results closely resemble the official set of G-SIFIs for 2010 and 2011.

Building on this methodological contribution, we then study how systemic

importance varied between 2007 and 2012. The comparis on of pre- and post- crisis

data integrates the (l imited) informat ion disclosed by the FSB, which covers only

2010, 2011 and 2012. Our sample includes large banks, and we examine separately

global, European and e uro-area institut ions. Measuring system ic importance wit hin

specific regions as well as g lobally is importa nt for at least two reasons. First, bank s

may have a significant i mpact on their national fina ncial system even if they are not

systemically impor tant globally, and domest ic regulators have an obvious interest in

monitoring systemic impor tance at that level. Secon d, bank failures may cause

cross-border spillovers that create internat ional problems, and the l ine between

domestic and global importance is quite thin, particularly when dealing with large

economic areas like the EU (BCBS 2012; EBA 2014). Our results suggest that both

the nature and the geog raphic distribution of systemic i mportance change d signifi-

cantly after the c risis: systemic i mportance has shif ted from developed to emergi ng

markets, particularly China, and the most important determinant of systemic

importance has become complexit y, instead of size.

158 Piergiorgio Alessandri et al.

© 2015 John Wiley & Sons Ltd