Assessing Systemic Importance With a Fuzzy Logic Inference System

• DOI: http://doi.org/10.1002/isaf.1371
• Date: 01 January 2016
• Published date: 01 January 2016

ASSESSING SYSTEMIC IMPORTANCE WITH A FUZZY LOGIC

INFERENCE SYSTEM

CARLOS LEÓN,

a,b

*CLARA MACHADO

a

AND ANDRÉS MURCIA

a,c

a

Monetary and International Investment Division, Banco de la República (Central Bank of Colombia), Bogotá, Colombia

b

CentER, Tilburg University, Tilburg,Netherlands

c

Representative Office for the Americas, Bank for International Settlements, México D.F., México

SUMMARY

Three metrics are designed to assess Colombian financialinstitutions’size, connectednessand non-substitutability as

the main driversof systemic importance: (i) centralityas net borrower in the money marketnetwork; (ii) centrality as

payments originator in the large-value payment system network; and (iii) asset value of core financial services. An

aggregated systemic importance index is calculated based on expert knowledge by using a fuzzy logic inference

system. We use principal component analysis to calculate a benchmark index for comparison purposes. Overall

similarities between both indexes put forward that expert knowledge aggregation is consistent with that based on a

purely quantitative standard approach. Specific non-negligible differences concur with the nonlinear features of an

approach whose intention is to replicate human reasoning. Both indexes are complementary and providea comprehensive

relative assessment of each financial institution’s systemic importance in the Colombian case, in which the choice of

metrics pursues the macroprudential perspective of financial stability.Copyright © 2015 John Wiley & Sons, Ltd.

Keywords: financial stability; fuzzy logic; macroprudential; systemic importance; systemic risk

1. INTRODUCTION

An institution may be considered systemically important if its failure or malfunction causes widespread

distress, either as a direct or indirect impact, where the main criterion for assessing systemic imp ortance

relates to their potential to have a large negative impact on the financial system and the real economy

(IMF et al., 2009). Hence, defining whether a banking or nonbanking financial institution is systemi-

cally important (or not) is critical for overseeing, supervising and regulating financial systems, and

for preserving financial stability.

To be able to assess systemic importance may assist financial authorities in focusing their attention and

resources—the intensity of oversight, supervision and regulation—where the systemic severity resulting

from a financial institution failing or near failing is estimated to be the greatest. Identifying systemically im-

portant institutions may also help financial authorities in policy-making (e.g. prudential regulation, over-

sight and supervision) and decision-making (e.g. resolving, restructuring or providing emergencyliquidity).

The literature converges towards the existence of three main key criteria for assessing and identi-

fying the systemic importance of financial institutions and financial market infrastructures: size,

* Correspondence to: Carlos León, Monetary and International Investment Division, Banco de la República, Bogota, Colombia.

E-mail: cleonrin@banrep.gov.co or carlosleonr@hotmail.com

Copyright © 2015 John Wiley & Sons, Ltd.

INTELLIGENT SYSTEMS IN ACCOUNTING, FINANCE AND MANAGEMENT

Intell. Sys. Acc. Fin. Mgmt. 23, 121–153 (2016)

Published online 17 June 2015 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/isaf.1371

connectedness and substitutability (IMF et al., 2009; Manning, Nier, & Schanz, 2009; Basel Com-

mittee on Banking Supervision, 2013).

1

According to Basel Committee on Banking Supervision

(2013), size is related to how difficult it is to replace the activities of a distressed or failing financial

institution; connectedness is related to how financial distress at one financial institution can materi-

ally increase the likelihood of distress at other institutions by means of the network of contractual

obligations; whereas substitutability is related to what extent a financial institution provides infra-

structure services (e.g. payments, custody services) to the financial system.

Despite the intuitiveness of these concepts, assessing and identifying systemic important institutions

within an indicator-based approach

2

remains a nontrivial task that implies several challenges. Two

challenges are particularly demanding. First, designing indicators or metrics for connectedness and sub-

stitutability may require, as acknowledged by recent literature, nonstandard data sources and tech-

niques, such as financial infrastructures’data and network analysis respectively.

Second, choosing a methodology capable of robustlyagg regatingthe metrics designed for the three afore-

mentioned concepts into a systemic importance index may be intricate. As put forward by IMF et al. (2009),

a high degree of judgment in the form of a qualitative framework is required to integrate the different com-

ponents of systemic importance. In this vein, concurrent with Mezei and Sarlin (2014), independent of the

rigor of the analytical tools, expert knowledge is always a central part of informed policy decisions.

The Basel Committee on Banking Supervision introduced an early approach to both challenges

(Basel Committee on Banking Supervision, 2011), which was later updated in Basel Committee

on Banking Supervision (2013). Regarding the first challenge, the Basel Committee on Banking Su-

pervision proposal relies mainly on traditional balance sheet data, against growing agreement on the

convenience of using other data sources and technical approaches (León, Machado, Cepeda, &

Sarmiento, 2012; ECB, 2010). About the second challenge, somewhat divergent from IMF et al.

(2009) concerns and suggestions, the proposal by the Basel Committee on Banking Supervision

(2013) employs an equal and fixed weighting scheme for aggregating five categories (i.e. each one

assigned a 20% weight), where the relevance of each constituent metric does not seem to follow

any technique –quantitative or qualitative. Furthermore, the equal and fixed weighted scheme pro-

posed by the Basel Committee on Banking Supervision may yield undesired results, such as biasing

results towards the most volatile categories (Hurlin & Pérignon, 2013) or naively assuming that all

financial systems are similar to each other.

This paper introduces two approaches to the aforementioned challenges. Regarding the first chal-

lenge, three metrics are designed to assess Colombian financial institutions’size, connectedness and

non-substitutability as the main drivers of systemic importance: (i) centrality as net borrower in the

money market network; (ii) centrality as payments originator in the large-value payment system (LVPS)

network; and (iii) asset value of core financial services. Unlike the Basel Committee on Banking Super-

vision (2013) proposal, our metrics are not limited to traditional balance sheet data and employ network

analysis methods (i.e. hub centrality) to assess the global importance of each financial institution within

1

The Basel Committee on Banking Supervision (2013) suggests adding two criteria (i.e. cross-jurisdictional activity and comple-

xity) in order to attain banks’global systemically importance and the difficulty of resolving a systemic event. Because this paper

focuses on nonglobal banking and nonbanking institutions’systemic importance, and as derivatives and other complex instru-

ments are rather scarce in the Colombian market, the criteria are limited to size, connectedness and substitutability, as originally

suggested by IMF et al. (2009). However, the proposed aggregation method is able to consider these two (or other) criteria.

2

There is an alternative to indicator-based approaches as the one proposed here: a model-basedapproach, which uses quantitative

models to estimate financial institutions’contributions to systemic risk. However,as highlighted by Basel Committee on Banking

Supervision (2013: 5): ‘models for measuring systemic importance of [financial institutions] are at a very early stage of develop-

ment and concerns remain about the robustness of the results; [for instance, the] models may not capture all the ways that a

[financial institution] is systemically important (both quantitative and qualitative)’.

122 C. LEÓN ET AL.

Copyright © 2015 John Wiley & Sons, Ltd. Intell. Sys. Acc. Fin. Mgmt., 23, 121–153 (2016)

DOI: 10.1002/isaf

the financial system, either as a borrower in the money market or as an originator of payments. In this

sense, this document contributes to the design of metrics that are closer to the macroprudential perspec-

tive of financial stability, in which the proximate objective is to limit financial system-wide distress and

in which prudential standards should be calibrated with respect to the marginal contribution of financial

institutions to system-wide risk (Borio, 2003).

About the second challenge, consisting of how to aggregate the different systemic importance met-

rics into a single index, we revise and implement the fuzzy logic-based systemic importance approach

suggested in León and Machado (2013). Fuzzy logic, in the form of a fuzzy logic inference system

(FLIS), is an engineering-type approach based on the deconstruction of expert knowledge into a quan-

titative method that imitates the way experts themselves think about the decision process regarding

what a systemically important financial institution is.

By implementing an expert knowledge-based model we try to exploit the fact that macroprudential

authorities (especially central banks) possess a variety of specialized domain intelligence and expertise

(see Mezei and Sarlin (2014)). In our case, the expert knowledge from the central bank is used to build

a set of inference rules that evaluates how the interrelations among the selected main drivers of

systemic importance determine their aggregation into a systemic importance index. In this sense,

paraphrasing Mezei and Sarlin (2014), we expect that the central bank’s acquired knowledge is able

to identify systemic importance metrics’interrelations that are hardly recognizable with only collected

historical data. We use principal component analysis (PCA), a standard (i.e. purely quantitative) dimen-

sion reduction method, to calculate a benchmark index for comparison purposes.

Results confirm that (i) credit institutions (i.e. commercial banks and other banking institutions) are

the most systemically important type of financial institution in the local market; (ii) despite being un-

important because of their size, some nonbanking institutions are systemically important because of

their centrality within the money market and the LVPS; (iii) similarities between PCA and FLIS indexes

put forward that expert knowledge aggregation is consistent with that based on a purely quantitative

standard approach; (iv) specific non-negligible differences between both indexes concur with the non-

linear features of an approach whose intention is to replicate human reasoning.

2. LITERATURE REVIEW: SYSTEMIC RISK AND SYSTEMIC IMPORTANCE

As presented by IMF et al. (2009), G20 countries use a general definition of systemic risk: the risk of

disruption to financial services that (i) is caused by an impairment of all or parts of the financial system

and (ii) has the potential to have serious negative consequences for the real economy. In respect of pay-

ment systems, the Committee on Payment and Settlement Systems and International Organization of

Securities Commissions (2012) defines it as the risk that the inability of one or more participants to per-

form as expected will cause other participants to be unable to meet their obligations when due.

Irrespective of which of these definitions is embraced, and despite there being no single definition of

risk that can be completely satisfactory in every situation (Dowd, 2005), it is common to think of risk

as a function based ontwo parameters: frequency and severity (Condamin, Louisot, & Naim, 2006), also

referred as likelihoodand impact respectively (Gallati, 2003). Although academic effort has traditionally

focused on systemic concerns based on the estimation of systemic risk (i.e. the product of frequency

and impact, as in Nor man et al. (2009)), there is a recent interest in focusing on systemic severity or

importance—see Tucker (2005), Rebonato (2007) and Taleb (2007).

Such increasing interest in the impact of systemic shocks—beyond the interest in their frequency—

results from the intrinsic characteristics of financial systems. As pointed out by Haldane (2009) and

ASSESSING SYSTEMIC IMPORTANCEWITH A FLIS 123

Copyright © 2015 John Wiley & Sons, Ltd. Intell. Sys. Acc. Fin. Mgmt., 23, 121–153 (2016)

DOI: 10.1002/isaf