Do Foreign Currency Accounts Help Relax Credit Constraints? Evidence from Nepal
Published date | 01 August 2018 |
Author | Nephil Matangi Maskay,Sven Steinkamp,Frank Westermann |
DOI | http://doi.org/10.1111/1468-0106.12183 |
Date | 01 August 2018 |
bs_bs_banner
DO FOREIGN CURRENCY ACCOUNTS HELP RELAX
CREDIT CONSTRAINTS? EVIDENCE FROM NEPAL
NEPHIL MATANGI MASKAY Nepal Rastra Bank
SVEN STEINKAMP*Osnabrück University
FRANK WESTERMANN Osnabrück University and CESifo
Abstract. We analyse a novel bank-level data set from Nepal, where domestic and foreign currency
(FX) deposits are reported separately on the liability side of commercial bank balance sheets. In a
panel regression analysis, we estimate semi-accounting-identities that allow us to identify the mar-
ginal sources of financing for various asset positions. We find that banks hedge against FX exposure
via their sectoral lending composition: banks with a large share of FX deposits primarily lend to
firms in traded-goods sectors. Loans to non-traded sectors are mostly financed by domestic deposits.
While earlier studies have documented a positive impact of FX accounts on financial development,
our analysis suggests that this does not need to imply that severely credit constrained sectors are
the main beneficiaries of this process.
1. INTRODUCTION
Foreign currency accounts have been shown to have a positive impact on finan-
cial sector development (see De Nicoló et al., 2005). In Nepal, the issue of for-
eign currency (FX) accounts has also gained attention in policy discussions, as
the country has been receiving increasingly large amounts of remittances from
the working population abroad in foreign currency, amounting to more than
28% of GDP in 2013. Recently, policy-makers have, therefore, suggested pro-
moting foreign currency accounts as a means to channel more remittances into
the domestic banking system (see e.g. the South Asian Association for Regional
Cooperation (SAARC), 2014; Pant, 2011).
However, does this policy really help to intermediate financial resources to
sectors of the economy that suffer most strongly from credit constraints? Could
it lead to a back-door introduction of foreign currency risk exposure in countries
with only partially liberalized capital accounts? In this paper, we address these
questions using a novel data set from Nepalese commercial banks’balance
sheets.
To motivate the exercise, we refer to the World Bank Enterprise Surveys
(WBES), which help to illustrate that credit constraints in Nepal are asymmetric
across sectors: small firms in non-traded sectors face severe credit constraints,
*Addressfor correspondence: Sven Steinkamp,Osnabrück University,Rolandstr. 8, 49069 Osnabrück,
Germany. E-mail: sven.steinkamp@uni-osnabrueck.de. We would like to thank the participants of
the annual conference of the Research Committee on Development Economics of the German
Economic Association (VfS) in Passau 2014, the VfS annual conference in Münster 2015, and, in
particular, our discussant Martin Brown for helpful comments and suggestions. Furthermore, we
are grateful for comments received at the presentation at the Nepal Rastra Bank, as well as several
interviews at the NRB, in particular Rajendra Pandit and Prakash Kumar Shrestha, the Himalayan
Bank, as well as the Remitters Association in Nepal.
Pacific Economic Review,••:•• (2016)
doi: 10.1111/1468-0106.12183
© 2016 John Wiley & Sons Australia, Ltd
while large traded-sector firms have relatively easy access to credit.1This obser-
vation confirms previous findings in the literature for the case of Nepal.2It is
consistent with the view that traded sector-firms, which have part of their reve-
nue in foreign currency, have easier access to bank credit, as well as alternative
external financing opportunities. In contrast, firms in the non-traded sectors are
usually small and rely more heavily on bank credit.
In the main part of the paper, we analyse a panel data set of commercial
banks’balance sheets. This unique data set is interesting for two reasons. First,
we are able to identify the foreign currency share of the banks’various asset and
liability positions, including the foreign currency (FX) deposits, which will be
our key variable of interest. Second, it also contains a detailed breakdown of
lending by sector, as well as other assets in domestic and foreign currency.
In this data set, we show that banks with large positions in FX deposits are
hedged against foreign currency risk in two ways. First, banks hedge their risks
directly by investing in non-resident and foreign exchange assets. Second, they
are also hedged via their sectoral lending composition: they lend more to firms
in traded goods sectors such as manufacturing. They also lend relatively less
to non-traded sectors, including the deprived sectors, which are recorded sepa-
rately and have been targeted by various support programmes. These indirect
hedging activities reduce the potentially positive impact of foreign currency ac-
counts on relaxing credit constraints, as foreign currency deposits are
intermediated primarily to the less constrained sectors.
In our panel regressions, we use the balance sheet data to estimate an account-
ing semi-identity that allows us to identify the marginal sources of financing of
various positions on the asset side of the banks’balance sheets. We demonstrate
that lending to the non-traded goods sector and to the deprived sectors of the
economy has largely been financed from domestic rather than foreign currency
deposits. This key finding is robust to alternative specifications that include in-
strumental variables (IV) regressions, different types of fixed effects, and
differencing the data. Looking at economic subsectors, we find that this result
is mostly driven by the construction, wholesale and retail-trade sectors, as well
as by direct lending to the deprived sectors. The manufacturing sector, on the
other hand, reacts significantly to changes in FX deposits, but not to domestic
deposits in most specifications.
We also illustrate that these hedging effects coexist with a positive indirect ef-
fect of FX deposits on total lending that results from an additional creation of
domestic deposits. This strong complementarity has its roots in the institutional
design of FX accounts in Nepal: as remittance-transfer companies are allowed
to hold FX deposits only for a limited time, a substantial share of the incoming
remittances are converted to domestic sector deposits. This indirect effect helps
1See Appendix I for a formal probit analysis. Traded goods sectors include, for instance, the
manufacturing sector, the mining sector and agriculture. The main non-traded sectors are construc-
tion and services.
2See, for instance, Tornell and Westermann (2005), Schneider and Tornell (2004), Beck et al. (2008)
and Brown et al. (2011).
N. M. MASKAY ET AL.2
© 2016 John Wiley & Sons Australia, Ltd
Pacific Economic Review
, 23: 3 (2018) pp. 464–489
doi:10.1111/1468-0106.12183
© 2016 John Wiley & Sons Australia, Ltd
DO FOREIGN CURRENCY ACCOUNTS HELP RELAX
CREDIT CONSTRAINTS? EVIDENCE FROM NEPAL
NEPHIL MATANGI MASKAY Nepal Rastra Bank
SVEN STEINKAMP*Osnabrück University
FRANK WESTERMANN Osnabrück University and CESifo
Abstract. We analyse a novel bank-level data set from Nepal, where domestic and foreign currency
(FX) deposits are reported separately on the liability side of commercial bank balance sheets. In a
panel regression analysis, we estimate semi-accounting-identities that allow us to identify the mar-
ginal sources of financing for various asset positions. We find that banks hedge against FX exposure
via their sectoral lending composition: banks with a large share of FX deposits primarily lend to
firms in traded-goods sectors. Loans to non-traded sectors are mostly financed by domestic deposits.
While earlier studies have documented a positive impact of FX accounts on financial development,
our analysis suggests that this does not need to imply that severely credit constrained sectors are
the main beneficiaries of this process.
1. INTRODUCTION
Foreign currency accounts have been shown to have a positive impact on finan-
cial sector development (see De Nicoló et al., 2005). In Nepal, the issue of for-
eign currency (FX) accounts has also gained attention in policy discussions, as
the country has been receiving increasingly large amounts of remittances from
the working population abroad in foreign currency, amounting to more than
28% of GDP in 2013. Recently, policy-makers have, therefore, suggested pro-
moting foreign currency accounts as a means to channel more remittances into
the domestic banking system (see e.g. the South Asian Association for Regional
Cooperation (SAARC), 2014; Pant, 2011).
However, does this policy really help to intermediate financial resources to
sectors of the economy that suffer most strongly from credit constraints? Could
it lead to a back-door introduction of foreign currency risk exposure in countries
with only partially liberalized capital accounts? In this paper, we address these
questions using a novel data set from Nepalese commercial banks’balance
sheets.
To motivate the exercise, we refer to the World Bank Enterprise Surveys
(WBES), which help to illustrate that credit constraints in Nepal are asymmetric
across sectors: small firms in non-traded sectors face severe credit constraints,
*Addressfor correspondence: Sven Steinkamp,Osnabrück University,Rolandstr. 8, 49069 Osnabrück,
Germany. E-mail: sven.steinkamp@uni-osnabrueck.de. We would like to thank the participants of
the annual conference of the Research Committee on Development Economics of the German
Economic Association (VfS) in Passau 2014, the VfS annual conference in Münster 2015, and, in
particular, our discussant Martin Brown for helpful comments and suggestions. Furthermore, we
are grateful for comments received at the presentation at the Nepal Rastra Bank, as well as several
interviews at the NRB, in particular Rajendra Pandit and Prakash Kumar Shrestha, the Himalayan
Bank, as well as the Remitters Association in Nepal.
Pacific Economic Review,••:•• (2016)
doi: 10.1111/1468-0106.12183
© 2016 John Wiley & Sons Australia, Ltd
while large traded-sector firms have relatively easy access to credit.1This obser-
vation confirms previous findings in the literature for the case of Nepal.2It is
consistent with the view that traded sector-firms, which have part of their reve-
nue in foreign currency, have easier access to bank credit, as well as alternative
external financing opportunities. In contrast, firms in the non-traded sectors are
usually small and rely more heavily on bank credit.
In the main part of the paper, we analyse a panel data set of commercial
banks’balance sheets. This unique data set is interesting for two reasons. First,
we are able to identify the foreign currency share of the banks’various asset and
liability positions, including the foreign currency (FX) deposits, which will be
our key variable of interest. Second, it also contains a detailed breakdown of
lending by sector, as well as other assets in domestic and foreign currency.
In this data set, we show that banks with large positions in FX deposits are
hedged against foreign currency risk in two ways. First, banks hedge their risks
directly by investing in non-resident and foreign exchange assets. Second, they
are also hedged via their sectoral lending composition: they lend more to firms
in traded goods sectors such as manufacturing. They also lend relatively less
to non-traded sectors, including the deprived sectors, which are recorded sepa-
rately and have been targeted by various support programmes. These indirect
hedging activities reduce the potentially positive impact of foreign currency ac-
counts on relaxing credit constraints, as foreign currency deposits are
intermediated primarily to the less constrained sectors.
In our panel regressions, we use the balance sheet data to estimate an account-
ing semi-identity that allows us to identify the marginal sources of financing of
various positions on the asset side of the banks’balance sheets. We demonstrate
that lending to the non-traded goods sector and to the deprived sectors of the
economy has largely been financed from domestic rather than foreign currency
deposits. This key finding is robust to alternative specifications that include in-
strumental variables (IV) regressions, different types of fixed effects, and
differencing the data. Looking at economic subsectors, we find that this result
is mostly driven by the construction, wholesale and retail-trade sectors, as well
as by direct lending to the deprived sectors. The manufacturing sector, on the
other hand, reacts significantly to changes in FX deposits, but not to domestic
deposits in most specifications.
We also illustrate that these hedging effects coexist with a positive indirect ef-
fect of FX deposits on total lending that results from an additional creation of
domestic deposits. This strong complementarity has its roots in the institutional
design of FX accounts in Nepal: as remittance-transfer companies are allowed
to hold FX deposits only for a limited time, a substantial share of the incoming
remittances are converted to domestic sector deposits. This indirect effect helps
1See Appendix I for a formal probit analysis. Traded goods sectors include, for instance, the
manufacturing sector, the mining sector and agriculture. The main non-traded sectors are construc-
tion and services.
2See, for instance, Tornell and Westermann (2005), Schneider and Tornell (2004), Beck et al. (2008)
and Brown et al. (2011).
N. M. MASKAY ET AL.2
© 2016 John Wiley & Sons Australia, Ltd © 2016 John Wiley & Sons Australia, Ltd
FOREIGN CURRENCY ACCOUNTS AND CREDIT CONSTRAINTS IN NEPAL 465
To continue reading
Request your trial