Discussion of

Are International Banks Different?

…………………………………………………..

TOBIAS ADRIAN, IMF FINANCIAL COUNSELLORBIS ANNUAL CONFERENCE – JUNE 2018

What is the paper about?

Degree of bank internationalization

Performance

Cyclicality of lending in home

country

Cyclicality of lending in host

countries

Strategy

Some key findings

Greater bank internationalization

Worse performance

Lower share of noninterest

income

Less deposit funding

More cyclicality of AE banks in

developing host country

Less cyclicality in home country

Comments

• Cross-border banking versus local banking via subsidiaries• A few stylized facts on G-SIBs’ degree / mode of foreign presence

• Role of relative profitability at home / abroad• Role of funding model• Role of line of business

• Regulation on “international” banking in home / host countries• Could help strengthen identification• Current regulatory issues

3

Cross-border banking remains larger than local banking

Source: Avdjiev, Aysun, and Hepp (2017)

G-SIBs’ relative profitability across locationsis a key driver of foreign expansion

G-SIBs ROA: Foreign Subsidiaries vs Home Subsidiaries 2014-16 Average, Percent

Source: Caparusso, Chen, Dattels, Goel, and Hiebert (forthcoming IMF WP)

G-SIBs ROA Differential and Internationalness Index 2014-16 Average, Percent

China

Japan

EUR ex UKUK

USA

0

10

20

30

40

50

60

-60 -30 0 30 60 90

Deg

ree

of in

tern

atio

naln

essi

ndex

(201

4-16

ave

rage

s)

RoA differentials of foreign and domestic bank subsidiaries (2014-16 averages, basis points)

0.0

0.3

0.6

0.9

1.2

1.5

China USA Japan EUR ex UK UK

Foreign Domestic

G-SIBs have a profitability gap in emerging markets

6

ROA: G-SIB’s Foreign Subsidiaries vs Domestic Incumbents(2014-2016 average)

Source: Caparusso, Chen, Dattels, Goel, and Hiebert (forthcoming IMF WP)

Greater multinationalization<-> Larger consumer business

7

BPCE

RBSBARC

UBS

MSWFCSG

NORD

BOC

INGSAN

BNY

STT

ABC

CCB

CA

JPM

STANC

BAC

HSBC

DB

GS

CS

R² = 0.142

0%

20%

40%

60%

80%

100%

120%

0% 10% 20% 30% 40% 50% 60% 70% 80%

Degr

eeof

Sub

sidi

ariz

atio

n --

Bank

Global Consumer Intensity

Bubble size proportional to

degree of internationalization

Subsidiarization and Global Consumer Intensity

Source: Caparusso, Chen, Dattels, Goel, and Hiebert (forthcoming IMF WP)

Greater multinationalization<-> Greater local funding

8

BPCE

RBSBARC

MSWFC SG

ING SAN

BNY

STT

JPM

STAN C

BAC

HSBC

DB

R² = 0.4362

0%

20%

40%

60%

80%

100%

120%

0% 20% 40% 60% 80% 100%

Degr

ee o

f Sub

sidi

ariz

atio

n -B

ank

Localization of Foreign Office Funding

Bubble size proportional to

consumer intensity

Subsidiarization and Overseas Office Funding Localization

Source: Caparusso, Chen, Dattels, Goel, and Hiebert (forthcoming IMF WP)

Use changes in regulation on “international” banking operations (in foreign countries) as a source of exogenous variation ?

Share of Countries that Tightened Regulations on International Banking Operations between 2006 and 2014, by Region (Percent)

0.0

0.1

0.2

0.3

0.4

0.5

AE excludingeuro area

Selected euroarea economies

Other euro areaeconomies

Emerging marketeconomies

1. Home Countries

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

AE excluding euroarea

Selected euroarea economies

Other euro areaeconomies

Emerging marketeconomies

2. Host Countries

9Source: April 2015 GFSR

Effect of regulation on foreign banking

10

-50

-40

-30

-20

-10

0

10

20

Foreign Cross-border Local

Regulations on international operations of banksGeneral banking regulations such as capital requirementsOtherClaims to GDP growth

Contributions of Regulatory Changes to Growth in Claims-to-GDP Ratio

Typesof claims

Source: April 2015 GFSR

Comments on selected regulatory issues

• Ring-fencing• Specificities of a banking union• Internationalization of EM / LIC banks• Correspondent banking relationships

11

Conclusion: in search of identification

Degree of bank internationalization

Performance

Cyclicality of lending in home

country

Cyclicality of lending in host

countries

Strategy

Additional Background

13

Subsidiary or Branch Model?

• Host authorities try to minimize financial stability risks by ensuring that foreign banks’ affiliates maintain sufficient capital and liquidity buffers easier under a subsidiary model

• One size doesn’t fit all: • diversity of business models, • differences in regulatory and tax regimes, • varying stages of financial development in host countries

• From a financial stability viewpoint, neither model outperforms the other in reducing both probability and cost of a banking group failure

• Mechanisms to ensure effective oversight and orderly resolution are a more effective route to resolving the efficiency-financial stability trade-off

Diversity of business models

15

Source: Caparusso, Chen,Dattels, Goel, and Hiebert(forthcoming IMF WP)

G-SIBs have a different business model than domestic incumbents in EMs

16

Decomposition of ROA difference(2014-16 average)

Source: Caparusso, Chen,Dattels, Goel, and Hiebert(forthcoming IMF WP)

Changes in regulation on “international” banking by type

17

Share of Countries that Changed Regulations on International Banking Operations between 2006 and 2014

0

5

10

15

20

25

30

35

Any

regu

latio

n

Pres

ence

Activ

ity

Supe

rviso

rydi

scre

tion

Info

rmat

ion

Depo

sitor

insu

ranc

e

Oth

er

TightenedLoosened

1. Home Countries(Percent)

0

10

20

30

40

50

60

70

80

Any

regu

latio

n

Pres

ence

Activ

ity

Supe

rviso

rydi

scre

tion

Info

rmat

ion

Reso

lutio

n

Oth

er

TightenedLoosened

2. Host Countries

ff l l

Source: April 2015 GFSR

Ring-fencing

• What is needed to prevent ring-fencing:• Better information sharing between home and host authorities• Supervisory colleges joint risk assessments• Crisis management groups group and firm-specific resolution plans

• Until sufficient progress is made in these directions, “ex-ante subsidiarization” remains an understandable ‘temptation’ and, in the end, a preferable option than discretionary ring-fencing during a crisis

International banking in a banking union

• Even in highly integrated areas, the incentive to ring-fence remains strong, without a completely common safety net e.g. in the Euro Area, the banking union needs to be completed with common deposit insurance and public backstop…

• … but financial integration requires more than no-ring fencing:• less fragmentation/further harmonization: e.g. reduce fragmentation in bank

insolvency laws that allow national authorities to avoid a substantial application of BRRD/bail-in

• coordinate/harmonize ELA and eventually centralize• greater centralization of supervision including 3rd country branches,

investment firms• more convergence in supervisory practices (esp. on-site)

“International” banks headquartered in EMs / LICs

• “International” banks from EMs / LICs (e.g. pan-African banks):• facilitate economic and financial integration; • promote transfer of knowledge and technology; • foster competition, contributing to financial innovation; • enhance financial inclusion; • and support the financing of local infrastructure.

• But risks stemming from cross-border banking are also high: as these groups expand, new channels for transmission of macro-financial risks and spillovers across home and host countries may emerge

• Enhanced cross-border cooperation on regulation, supervision and crisis management is needed, in particular to support effective supervision on a consolidated basis and cross-border resolution

Correspondent banking relationships

-25%

-20%

-15%

-10%

-5%

0%

5%

Africa Americas(excl. North

America)

NorthAmerica

Asia Oceania Europe(excl.

EasternEurope)

EasternEurope

Number of active counterparty countries by region(average percent change across region)

Jan. 2011 - Jun. 2017 Dec. 2016 - Jun. 2017

Source: FSB, SWIFT Watch, National Bank of Belgium.

The decline in CBRs continues: overall concentration of correspondents has decreased since 2015

While it seems to have stabilized in some areas (e.g. Caribbean), financial fragilities remain in some countries

Drivers of CBR pressures remain the same (e.g. lack of clarity over regulatory expectations; weaknesses in regulatory and supervisory frameworks, including for AML/CFT); but also some financial integrity issues (e.g. corruption, transparency, sanctions)

Potential negative impact on financial inclusion and remittances requires continued attention

IMF multipronged approach:• Monitor risks• Assess macro-criticality of this issue• Provide targeted technical assistance &

training• Facilitate dialogue• Collaborate with other stakeholders