Articulo 7 Oficial
Articulo 7 Oficial
Articulo 7 Oficial
Abstract
This paper analyses the financial stability implication of financial market development
in emerging markets. On the one hand, financial market development has enhanced
resilience and improved domestic financial stability by providing new tools to raise
funds and manage risks. On the other hand, high foreign participation in local
currency government bond markets, growing private sector foreign currency debt
levels and the growing role of non-bank financial institutions have increased external
vulnerabilities. These trade-offs raise several policy challenges such as developing
hedging markets, improving monitoring of FX flows and implementing
macroprudential tools and FX intervention. The paper concludes with some financial
stability implications of fintech.
JEL classification: D47, G23, G28, G38, O3.
Keywords: financial market development, financial stability risks, capital markets,
fintech.
In the aftermath of the crises in the 1990s, many EMEs sought to encourage FMD. As
a result, capital markets have deepened, becoming more liquid and resilient. On some
measures, equity and government bond markets in some EMEs are now comparable
in size with those in small open AEs. That said, financial intermediation remains heavily
bank-based, in particular for household lending. 1
As markets have developed, the private non-financial sector has become more
indebted. Corporate borrowing has been rising rapidly. And total credit to EME
households has almost doubled over the past decade. In several Asian EMEs,
household credit relative to GDP has now reached levels similar to, or even higher
than, those in the United States and some other advanced economies (AEs) (Graph 1,
left-hand panel).
Although banks maintain their dominance, lending by non-bank financial
intermediaries (NBFIs) has increased quickly, their financial assets almost tripling from
2007 to reach 86% of GDP in 2018. 2 Expansion was especially rapid in China, where
the assets of other financial intermediaries (OFIs), a sector dominated by investment
funds, soared from $270 billion in 2007 to $11 trillion in 2018 (Graph 1, centre panel).
Assets of pension funds have also grown in many EMEs, reflecting the strengthening
of existing voluntary pension systems or the launching of new mandatory schemes
(Graph 1, right-hand panel). But, except in a few jurisdictions, their size remains
relatively small.
1
See accompanying chartpack for details.
2
See FSB (2020).
Total credit to households at 2018 Lending by OFIs Funded pension assets of select
EMEs from 2010 to 2018 2
% of GDP % of GDP USD trn % of GDP
1
Aggregates of all jurisdictions. 2
Red arrow: an increase from 2010 to 2018; blue arrow: a decrease from 2010 to 2018.
Sources: FSB; IMF, World Economic Outlook database; BIS survey; BIS.
Cross-border capital flows have long been a major driver of financial stability risks in
EMEs. FMD over the past two decades has changed the currency composition of
EMEs’ assets and liabilities as well as the investor universe. These changes have
important implications for EMEs’ external vulnerabilities.
3
Furthermore, to reduce regulatory arbitrage between NBFIs and banks, some jurisdictions subject
NBFIs to the same regulatory provisions as banks or “right-size” regulatory measures based on the
scale and risks posed by the activities (eg Russia, Saudi Arabia or Singapore).
The deepening of local currency government bond (LCGB) markets has been one of
the most successful aspects of FMD in many EMEs. This has been particularly so in
Latin America, where the median share of government bonds denominated in
domestic currency rose from 43% at end-2004 to 85% at end-2019 (Graph 2, first
panel).
1
Box and whisker plots show median, interquartile range and range. 2 Argentina, Brazil, Chile, Colombia, Mexico and Peru. 3 China, Hong
Kong SAR, India, Indonesia, Korea, Malaysia, the Philippines, Singapore and Thailand. 4 The Czech Republic, Hungary, Israel, Poland, Russia,
Saudi Arabia, South Africa and Turkey.
Sources: Thomson Reuters Refinitiv; IMF, AREAER, Sovereign Debt and World Economic Outlook databases; IPE.
Policies to encourage foreign participation, coupled with the search for yield
amid global low interest rates, have been important drivers. Non-resident holdings
of EME LCGBs rose steadily from $800 billion in 2004 to almost $2 trillion in 2019
(Graph 2, second panel). Several FMD developments helped: first, as part of the efforts
to diversify the investor base, many EMEs relaxed capital account restrictions on
foreign participation in local bond markets (Graph 2, third panel). 4 Second, as the
size and liquidity of these markets grew, EME local government bonds emerged as a
global asset class, stoking demand from institutional investors. For example, both the
Bank of Mexico and the Central Bank of Colombia papers note that the inclusion of
LCGBs into a global index was supportive for the domestic bond market development.
4
Since 2010, an increasing number of EMEs have introduced conditional requirements to allow foreign
participation, coinciding with the start of what the World Bank termed the fourth wave of debt
accumulation (Kose et al (2019)).
5
Eichengreen et al (2003) coined the phrase “original sin” to describe the tendency for foreign
investors to lend to EMEs only in foreign currencies due to EMEs’ structural weaknesses and lack of
market credibility.
6
Argentina, Colombia, Hungary, Indonesia, Malaysia, Mexico and the Philippines.
7
At times of stress, an investor has the incentive to redeem from a fund earlier than others as the
liquidation value of fund shares declines when there are large withdrawals.
8
The survey for this meeting shows that about half of the central banks do not know whether foreign
investors have hedged their LCGB FX exposures or not. Those who do have information believe that
foreign investors have hedged only partially (Graph 6).
1
Shaded area denotes insignificant impact of changes in foreign holdings on LC yield spread. Expected appreciation and depreciation are
represented by the cross-economy historical 1st and 99th percentile, respectively, of changes in risk reversal. 2 Marginal effect of changes
in lagged FX expectation on LCGB yield spread when the share of foreign holdings exceeds 13%.
Despite the progress in the LCGB segment, most EME local currency non-financial
corporate bond markets have lagged behind those of AEs in terms of volume and
liquidity (see chartpack). Between 2009 and 2017, the median share of EME non-
financial corporates’ local currency issuance was only 35% (Graph 4, left-hand panel),
some 15 percentage points below the median in AEs. The exceptions are some large
Asian EMEs, where the share exceeds 70%. 9 One reason for their subdued
development could be the lack of foreign participation. The survey for this meeting
reveals that foreign participation is generally below 5% (Graph 4, centre panel), much
9
The note submitted by Brazil points out that the decline in longer-term yields associated with fiscal
adjustment has spurred local capital markets by encouraging domestically based corporations to
prepay their foreign currency liabilities and raise funds in local currency.
Share of local currency corporate Investor base of local currency Gross international bond issuance by
bond issuance1 corporate bond markets EME non-financial corporates
Per cent Per cent USD bn
1
By nationality. Average 2009–17.
Sources: CGFS (2019); BIS survey; BIS international debt securities statistics.
The latest surge of corporate external foreign currency debt issuance is likely to
have increased EME vulnerabilities.
First, corporates’ leverage and foreign currency debt service burdens have
increased, raising financial stability risks and potentially limiting policy space. 10 High
leverage makes EME corporates vulnerable to increases in borrowing costs. But even
a moderately levered firm may not be able to cover its interest expense, however low
that may be, if it is not profitable. For EME firms, a broad decline in earnings since
2010 has weakened their debt-servicing capacity (Beltran et al 2017). That said, the
maturity of bonds has lengthened. The share of bonds with a maturity of less than
five years has fallen from 40% in 2010 to 25% in 2018. This significantly reduces
rollover risks.
10
The note from the Bank of Russia stresses that high levels of foreign currency-denominated corporate
debt severely limited policy space during the 2008 and 2014 crises.
FMD has also contributed to a strong growth in FX markets for EME currencies in
recent years. 11 The average daily turnover of EME currencies, in both spot and
derivative transactions, rose by almost 60% between 2016 and 2019, to $1.6 trillion,
representing almost 25% of global FX turnover. Apart from the gradual
internationalisation of the Chinese renminbi, other general global macro and financial
developments, FMD played a role here. The broadening of FX trading systems and
their electronification have encouraged the participation of new players such as
hedge funds, proprietary trading firms and algorithmic traders in a market
traditionally dominated by inter-dealer trading among large banks (Patel and Xia
(2019)).
FX derivatives have expanded strongly. The latest data show that the volume of
trading in derivatives, on average, was twice that of spot transactions in April 2019.
The robust growth of non-deliverable forwards (NDFs) stood out.
The impact of higher turnover in FX markets on financial stability is not clear-cut.
On the one hand, higher turnover boosts market liquidity, and it can smooth price
adjustments, enhancing markets’ capacity to absorb shocks. On the other hand, it can
also amplify global market spillovers and increase exchange rate volatility. Some
central banks also worry that NDFs are not only used for hedging but for speculative
purposes when the domestic currency is not fully convertible, with potentially
destabilising effects.
Led by the robust growth in NDFs, offshore markets have grown in importance.
Offshore turnover is larger than onshore turnover for emerging Asia currencies as a
group and the ratio is above 3 for Latin America and some other EME currencies.
Views differ on the cost/benefit balance of the growing role of offshore markets.
In a textbook world, this should not matter as the location of trading is immaterial. 12
In practice, though, offshore trading can have benefits. It can increase liquidity, which
in turn reduces transaction costs, triggering a welcome feedback spiral. Overlapping
time zones can also allow agents to trade around the clock, helping them manage
11
There are exceptions. For example, the Central Bank of Israel note states that the recent increase in
the turnover of foreign institutions in the FX market has not been matched by an increase in their
involvement in domestic capital markets.
12
For currencies with deep and liquid markets, this may even be so in practice most of the time. But in
some instances, the location and time zone of trading matters, even for the most liquid currencies as
illustrated by the role of early Asian trading during the sterling and yen flash events in recent years.
Non-residents’ operating outside of the Asian session amplify market reaction Graph 5
Change in baht exchange rate vis-à- Trading volume in onshore market2 Turkish lira spot and Indonesian
vis the US dollar1 rupee onshore-offshore forward
spread3
% USD mn TRY/USD Pips
1
Calculated from sum of hourly changes of THB since 2017. 2 Trading volume is proxied by the average hourly volume of spot and forward
transactions of the onshore market since 2019 (data inception). 3 Asia movement: change in the THB/USD during onshore market trading
hours. London movement: change in THB/USD during the subsequent London trading hours.
Policy implications
FMD has changed the external financial stability risks in the past two decades, raising
several policy challenges.
Developing hedging markets. The survey reveals that it is rather common for non-
bank financials and non-financial firms in EMEs to hedge only a small portion of their
FX exposures. And that has not changed much over the past decade (Graph 6, left-
hand and centre panels). An often cited obstacle is the high hedging cost due to
underdeveloped derivatives markets.
Various central banks have employed different strategies to promote these
markets. For example, the Central Bank of Malaysia has launched a dynamic hedging
programme to allow foreign investors to enter into NDF contracts onshore. Bank
Indonesia has introduced mandatory corporate hedging requirements. A few central
banks (eg those of Indonesia, Mexico and Turkey) have started to conduct FX
interventions using NDFs settled in local currency to encourage a private sector
Currency hedging by resident non- Currency hedging by foreign Capital control measures1
financial corporations investors who hold EME local
currency government bonds As a share of respondents
Unchanged Unchanged
2018
2018
____________________________________________________ ____________________________________________________
not available
not available
2010
2010
CO
CO
CL
KR
PL
AR
BR
CZ
HU
IL
MY
TH
HK
SG
ID
PE
TR
RU
MX
BR
HU
ID
IL
MY
PE
AR
CZ
HK
KR
PH
PL
SG
TH
TR
ZA
RU
MX
To a large extent Not at all To a large extent Not at all
Some extent Unknown Some extent Unknown
1
For inflows, corresponds to restrictions on purchases locally by non-residents and restrictions on sales or issued abroad by residents. For
outflows, corresponds to sales or issued locally by non-residents and purchased abroad by residents.
Improving monitoring of FX flows. The broader investor base in EME debt and FX
markets has contributed to market depth and liquidity. But it could also generate
destabilising price dynamics, especially when transactions are speculative in nature.
In the current low global interest rate environment, FX carry trades involving EME
currencies have become increasingly popular. This raises the risk of sudden capital
outflows. 13
It is important for EME central banks to continue monitoring flows in the FX
market, including their size, composition and underlying drivers. But the increase in
non-bank activities raises data challenges. Bank Indonesia and the Bank of Israel have
recently introduced new reporting regulations, requiring market participants with a
daily turnover above a certain threshold to report transactions. Yet it remains difficult
to track data on offshore transactions, in particular when liquidity pools are
fragmented. In this context, sharing of market intelligence between central banks
could be an important way forward.
13
Some global investors have chosen to fund EME-focused carry trades with another highly correlated
EME currency to reduce risk. For example, investors may fund long positions of the Brazilian real with
the Mexican peso as both are perceived to be economies with commodity exposure. And this could
lead to further increase in short-term capital flows to EMEs.
14
For a detailed discussion, including how goals and objectives of FX intervention have evolved see the
2019 Emerging Markets Deputy Governors meeting, BIS (2019b).
15
Question remains about whether CFMs should be applied preemptively or on an ad-hoc basis. ASEAN
(2019) argues for a preemptive use of those CFMs that are aimed at ensuring financial stability.
Sources: BIS survey; IMF, Balance of payments statistics and World Economic Outlook database.
16
See chartpack for a definition of FMD. BIS (2019a), Chapter III provides a detailed discussion of Big
techs’ impact on credit provision as well as many other financial services.
17
At the first instance, these firms’ growth could pose risks to consumer and investor protection.
18
Take China as an example, a large tech firm has used securitisation to finance its microlending
business. The outstanding amount of these microloan-backed securities has grown from a negligible
amount in 2015 to a peak of CNY 470 billion in 2017.
ASEAN (2019): “Capital account safeguard measures in the ASEAN context”, February
2019.
Banerjee, R, M B Devereux, and G Lombardo (2016): “Self-oriented monetary policy,
global financial markets and excess volatility of international capital flows“, Journal of
International Money and Finance, vol 68, pp 275–97.
Bank for International Settlements (2018): Annual economic report, 2018.
——— (2019a): Annual economic report, 2019.
——— (2019b): “Reserve management and FX intervention”, BIS Papers no 104.
Beltran, D, G Keshav and A Rosenblum (2017): “Emerging market nonfinancial
corporate debt: How concerned should we be?”, Board of Governors of the Federal
Reserve System IFDP Notes, June 2017.
Carstens, A and H S Shin (2019): “Emerging markets aren’t out of the woods yet”,
Foreign Affairs, 15 March.
Chui, M, I Fender and V Sushko (2014): “Risks related to EME corporate balance sheets:
the role of leverage and currency mismatch”, BIS Quarterly Review, September.
Committee on the Global Financial System (2009): “Capital flows and emerging
market economies”, CGFS Papers, no 33.
——— (2019): “Establishing viable capital markets”, CGFS Papers, no 62.
Eichengreen B, R Hausmann and U Panizza (2003): “The pain of original sin”, in B
Eichengreen and R Hausmann (eds), Other people’s money – Debt denomination and
financial instability in emerging market economies, University of Chicago Press.
Financial Stability Board (2020): Global Monitoring Report on Non-Bank Financial
Intermediation 2019, January.
Fukao, M (1991): “Exchange rate movements and capital-asset ratio of banks: on the
concept of structural positions”, Bank of Japan Monetary and economic studies, Vol.9,
No.2
Gertler, M, and P Karadi (2011): “A model of unconventional monetary policy.” Journal
of Monetary Economics, vol 58, no 1, pp 17–34.
Hofmann, B, N Patel and S Wu (forthcoming): “Original sin redux: a model-based
evaluation”, BIS memo.
IMF (2019): “Lower for longer”, Global Financial Stability Report, October.
Kose, M, P Nagle, F Ohnsorge and N Sugawara (2019): Global waves of debt: Causes
and consequences, World Bank Group, December.
Morris, S, I Shim and H S Shin (2017): “Redemption risk and cash hoarding by asset
managers,” BIS Working Papers, no 608.
Patel, N and D Xia (2019): “Offshore markets drive trading of emerging market
currencies”, BIS Quarterly Review, December.
United Nations Conference on Trade and Development (2019): “Financing a global
green new deal”, Trade and Development Report, September.
On the back of rapid growth in local currency debt markets over the last two decades,
EMEs have reduced their reliance on external foreign currency borrowing – the so-
called original sin. But this has not eliminated their financial vulnerability entirely.
EMEs still rely heavily on foreign sources of funding, albeit in local currency, as their
bond markets have a less developed base of domestic institutional investors. Carstens
and Shin (2019) have termed this the “original sin redux”. They argue that this still
leaves EMEs vulnerable to capital flow reversals on account of currency mismatches
on the balance sheets of global lenders.
This annex discusses a model-based evaluation of the original sin redux and the
vulnerability of EMEs to foreign shocks. It finds that, while the original sin redux
reduces the vulnerability compared with original sin, it falls short of matching the
benefits that ensue from a large domestic investor base.
The model is a two-country new Keynesian DSGE setup featuring a small open
EME (the home country) and a large global economy (the United States). For ease of
modelling, the currency mismatches are assumed on banks’ balance sheets but results
would be similar if the mismatch affects other domestic balance sheets such as those
of firms or governments and banks are hedged. In particular, EME firms borrow from
domestic banks to finance investment. EME banks in turn obtain their funding from
global banks and deposits from domestic households. Both domestic and foreign
banks face a funding constraint that is governed by their net worth. 19 This gives rise
to a financial channel via the exchange rate. For instance, if the EME currency
depreciates, the value of loans, which are in local currency, declines relative to the
value of the liabilities, which are in foreign currency. This leads to a drop in net worth
for the EME bank in the case of the original sin, and for the global bank in the case of
the original sin redux. 20 In either case, there is a tightening in lending conditions that
affects the real economy.
Graph A shows the impulse responses of EME variables to a 100 basis point
increase in the US interest rate. The first panel shows that the shock leads to a
depreciation of the exchange rate, which triggers the financial channel. As a result,
GDP declines (second panel).
The decline is sharpest in the case of the original sin, where the currency
mismatch is on the balance sheet of EME banks. In comparison, the original sin redux
helps mitigate the impact on EME output and investment. In particular, while the
exchange rate still depreciates, there is no direct impact on the balance sheet of
domestic banks). As a result, net worth of domestic banks declines by less (Graph A,
third panel), tightening in credit spreads is lower, and lending to EME firms declines
by less than in the original sin scenario. 21 Finally, the impact on the real economy is
further mitigated in the presence of domestic sources of funding, as EME banks can
19
In this framework, banks could also be interpreted as asset managers more broadly.
20
The drop in net worth could be mitigated if the global bank also has liabilities denominated in the
EME currency. In the case where the EME currency liabilities and assets of the global bank are perfectly
matched initially, the shock would lead to an increase in its capital to asset ratio (see for instance
Fukao, 1991). In this model however, all liabilities of global banks are in dollars.
21
Conversely, the decline in net worth and output in the US is larger under the original sin redux
scenario compared to original sin (Graph A, fourth panel).
Real exchange rate GDP Net worth (domestic) Net worth (foreign)
% deviation from steady state % deviation from steady state % deviation from steady state % deviation from steady state
1
Impulse responses are in percentage deviations from the non-stochastic steady state. The exchange rate is defined as the number of EME
currency units per unit of foreign currency, so that an increase denotes an EME depreciation. “Original sin” denotes a scenario where the EME
borrows entirely from abroad in foreign currency. “Original sin redux” is a scenario in which the EME borrows entirely from abroad in local
(EME) currency. “Domestic” deposits denotes a scenario in which half of the EME borrowing is from domestic deposits and half is from abroad,
both in local (EME) currency.
Many state-owned enterprises have been in a better position than private firms to
reap the benefits of capital market development. First, their creditworthiness is often
perceived as stronger, due to explicit or implicit government guarantees. For similar
reasons, banks are more willing to lend to SOEs than to firms with a similar financial
position. A contributing factor is that these firms tend to be big, with tangible assets
that can serve as collateral. Second, funding for these large firms has benefited from
growing investor demand for sovereign and corporate bonds.
These factors have enabled EME SOEs to raise significant amounts of funding in
international and local bond markets (Graph B, left-hand panel). From 2005 to 2018,
the value of outstanding bonds more than doubled, surging to $450 billion. The
currency composition of debt, which has remained broadly constant over time, differs
substantially across regions. In 2018, the share of foreign currency bonds of Latin
American SOEs was around 80%, compared with 44% for Asian SOEs.
Amount outstanding of bonds1 Balance sheet indicators2 SOE-sovereign credit rating spread6
USD bn Per cent Grade points
1
Based on data from 34 SOEs. Latin America (eight SOEs): AR, BR, CL, CO, MX, PE; Emerging Asia (19 SOEs): CN, IN, ID, KR, MY, SG, TH; Other
EMEs (seven SOEs): CZ, RU, AE, ZA. 2 Non-tradable sector (16 SOEs) includes electricity, construction/real estate/ports, transportation and
utilities. Tradable sector (18 SOEs) includes oil and gas, aluminium, chemicals, and copper. 3 Debt over equity. 4 Net income over
equity. 5 Earnings before interest and taxes over interest expenses. 6 Based on S&P’s long-term foreign currency issuer rating of 31 SOEs
and their respective sovereigns.
As SOEs have accumulated debt and their leverage increased, their profitability
and debt servicing capacity have fallen (Graph B, centre panel). As a result, SOE credit
ratings have dropped (Graph B, right-hand panel). Non-economic factors have helped
to erode the financial strength of some major SOEs. For instance, weak corporate
governance has led to the inefficient use of resources, and revenues have been
diverted to government coffers. In 2015, some Latin American SOEs were on the verge
of bankruptcy due to political corruption scandals.
22
Another financial stability issue is the degree of domestic banks’ exposure to SOEs debt, including to
the firms that provide services to them. Although bank solvency is not a major risk, a rating
downgrade of the SOE could trigger higher capital requirements and increase credit costs.
23
In many jurisdictions, authorities also set limits on pension funds’ foreign currency exposure to
accompany the relaxation on their foreign securities holdings. In Hong Kong SAR, for example, there
is no limit on a mandatory pension fund scheme’s foreign investments but the fund must restrict its
foreign currency exposure to not more than 70% of its total assets (CGFS (2009)).
24
It should be noted that in both cases, Korean banks are building up debt vis-à-vis foreigners; but the
difference is that by signing an offsetting forward contract with a foreign bank branch, the banks’
debt are “collateralised” by the other currency leg.
Foreign assets of Israeli institutional Korean banks’ assets Korean banks’ liabilities
investors
As a percentage of total assets USD bn USD bn