An earlier version of this note was submitted to the G20 International Financial Architecture working group. The present note includes additional information on developments and policy responses in non-G20 member countries.

The COVID-19 outbreak, in addition to dramatic implications for the health of people around the world, has triggered major economic and financial consequences: GDP is now expected to contract by 6% globally in 20201; trade could fall by 12% to 32% this year2; FDI flows are expected to fall by around 40%3; equity markets initially suffered sharp sell-offs before recovering somewhat in recent weeks, and financial conditions have substantially tightened.4 These developments in turn are significantly impacting global capital flows and countries’ external positions.5

The COVID-19 crisis and steep oil price declines have led to sharp swings in foreign exchange markets. As the COVID-19 outbreak has escalated into an unprecedented global crisis, accompanied by plummeting oil prices, exchange rates of key emerging market economies (EMEs) dropped substantially, notably those of the Brazilian real (BRL), Mexican peso (MXN), Russian rouble (RUB), South African rand (ZAR), and later the Indonesian rupiah (IDR) and the Turkish lira (TRY) (Figure 1). The currency depreciation accelerated between end-February and mid/end-March 2020. The currencies of advanced economies (AEs) have generally strengthened over the period, particularly the USD, JPY, EUR, and CHF (Figure 2). After a notable drop in the first half of March, the Canadian and Australian dollar rebounded.6

Since April 2020, the hardest-hit currencies have started to recover. This is particularly the case of the IDR, and the RUB, while the ZAR and MXN have stabilised. This rebound, and the heterogeneity in reactions across different emerging market currencies, may reflect various developments, including initially the extension by the US Federal Reserve of swap lines, which was coordinated with other G7 central banks and the Swiss National Bank in March7, and the opening of a repo facility (see section below) as well as the agreement announced by OPEC+ countries at the beginning of April 2020.

Portfolio investments, typically a volatile asset class, have reacted rapidly to the shock that the pandemic inflicted on the global economy. Economies that had entered the crisis with weaker positions have already experienced massive outflows of portfolio investments. This repeats a familiar pattern whereby international investors transfer capital back home or invest in safer assets during periods of uncertainty.

What appears exceptional about capital flow dynamics during the COVID-19 crisis are the scale and speed of the outflows. The Institute of International Finance (IIF) daily flows tracker estimates that around USD 103 billion were drawn from EMEs between mid-January and mid-May 2020, with equity inflows plummeting first, followed by debt flows. This sudden stop in capital flows has been faster and more incisive than observed during similar events in recent years, including during the 2008 Global Financial Crisis, the 2013 Taper Tantrum when the Fed announced a gradual exit to its quantitative easing programme, and the 2015 Chinese stock market sell-off (Figure 3).

A more complete picture of portfolio flow dynamics is provided by preliminary balance-of-payments data for the first quarter of 2020, which show the substantial drop in non-resident portfolio flows to EMEs, both debt and equity, and most strikingly in Brazil (Figure 4, left axis). While inflows to AEs generally strengthened in February, Japan and the US also experienced large drops in portfolio debt inflows in March (Figure 4, right axis). Portfolio inflows to Italy declined by more than USD 60 billion in March, in contrast to France and Germany, both of which recorded important debt inflows (Figure 4, left axis). Resident investors in some cases exacerbated the drop in non-residents flows by investing more abroad (Turkey, Japan), while in other cases they cut foreign investment, repatriating funds back home (Canada, US).

Higher frequency data point to a rebound of portfolio flows since April. According to the Emerging Portfolio Fund Research (EPFR) Global Navigator, which tracks mutual fund transactions, AE bond funds started to record important positive inflows at the end of April. EM equity funds, however, were still experiencing outflows at the end of May, hence recording 15 consecutive weeks of outflows since the beginning of the crisis. These however occurred at a much slower pace than in March, and on the bond side, EM funds stabilised in April and May.

The shock has not been uniform across countries. On the equity side, Brazil, India, Korea, Malaysia, Philippines, Turkey and South Africa funds were particularly affected, more so than China, Mexico, and some of the other oil exporters after the OPEC+ agreement. AE funds have generally resisted the global sell-off, with US, Japan and Europe starting to record inflows already in March. In Europe, a North (inflows)/South (outflows) divide appeared. On the bond side, Korea, Turkey and South Africa funds started to rebound in late April and May, while Thailand and South Africa recorded important outflows.

In parallel, global financing conditions tightened significantly, especially in global US dollar funding markets. The substantial increase in the cost of dollar funding is well reflected in recent movements in the “FX swap basis” - i.e. the difference between the interest rate for USD in the money market and the implied interest rate for USD from the Foreign Exchange (FX) swap market – which widened for most currencies. Since the beginning of April, USD funding pressures have abated significantly, partly thanks to the introduction of Fed swap lines.

These developments regarding dollar funding and the broader tightening of financing conditions have important implications for the corporate sector. The sharp reversals in profits and earnings expectations for companies triggered by the COVID-19 crisis comes after several years of corporate debt and leverage build-up, driven by historically low borrowing costs and various (including tax) incentives favouring debt over equity. In addition, a substantial part of this debt increase took risky forms.8 High indebtedness creates important vulnerabilities, as short term difficulties to service the debt – a liquidity issue – will over time transform into a solvency issue the longer business activity gets disrupted.9

Corporates face additional risks in the context of substantial currency depreciation. In recent years, EME corporates have increasingly been able to issue bonds in their local currency, with FX issuance now representing only 18% of total issuance.10 On the other hand, the share of foreign ownership has increased. A drop in a currency’s value against the USD will increase debt-servicing costs for FX bond issuers, likely amplifying liquidity issues. Local currency bond issuers may not be spared from the shock, either. A depreciation of EME currencies tightens AE investor risk constraints and may trigger sales and exit.11 This is apparent in the ongoing sudden stop in portfolio flows described above, which may reflect portfolio reallocation in the context of falling EME asset returns.

Corporates are not alike in terms of their funding needs and vulnerabilities.12 Differences relate to the types of instruments (loans or bonds), the share of foreign ownership of these instruments, and the nature of the indebted sector (banks, non-bank financial, or corporate). Loans are substantially larger than bonds only in China, India, Russia and Turkey. Chinese corporates have limited external debt,13 and a large share of their FX debt, as well as that of Russian and Turkish corporates, is with resident banks. Indonesia and the Philippines largely rely on resident USD deposits to fund FX loans, while Indian corporates rely on cross-border bank loans, and Turkish banks tend to borrow abroad to fund domestic FX loans. In Mexico and Turkey, FX debt has mostly been raised by non-financial corporations (NFCs). Non-bank financials have tended to rely less on FX borrowing, with the exception of Malaysia where they have held a large share of FX debt.

The macroeconomic and financial consequences of the COVID-19 outbreak largely follow dynamics that have been the object of a large body of research. The COVID-19 shock differs from past episodes of capital flow volatility because of its global shock on both demand and supply and its exogenous nature, but also by its speed and scale. This tempers the use of historical events as a guide for current policy. Nevertheless, the dynamics are overall well understood: Sudden stops in capital flows are typically associated with decreases in global risk appetite and with contagion proxies (e.g. financial linkages, trade linkages and regional proximity), although the influence of global risk on likelihood of sudden stops may have become more muted after the 2008 crisis.14 Countries with negative growth shocks are also more likely to experience sudden stops in capital flows. Finally, the behaviour of residents is crucial to understand capital flow dynamics in crises: drops in net flows often reflect as much a freeze in foreign investment (“stop” episode) as an increase in resident capital fleeing the country (“flight” episode).

A very large number of countries experienced sudden stops following the Global Financial Crisis, with banking flows playing the most crucial role. However, portfolio equity and debt flows led a number of stops in the early 2000’s and around 2015 (Figure 5). In comparison, BoP data confirm that the COVID-19 shock was a portfolio-driven sudden stop, with bank inflows appearing to compensate to a varying extent the drop in portfolio flows.

Countries have responded very differently in outflow/sudden stop situations, partly because a country’s position at the onset of the crisis largely conditions its policy response:

  • Monetary policy is a double-edged sword: easing rates protects domestic balance sheets where debt burdens are already high – but may accelerate outflows.

  • Sterilized FX intervention by central banks has generally been effective in delaying depreciation, giving time for the private sector to unwind FX positions. Its effectiveness, however, depends on the initial reserve stock and length of the shock.16

  • Macroprudential policy and currency-based measures have been eased in some countries, and tightened in others, especially macroprudential tools during episodes of capital flight (Table 1). Macroprudential policies are generally tasked with enhancing the resilience of the financial system to cope with shocks, and few tools are used counter-cyclically.

  • Countries have tended to ease inflow controls in stop and flight episodes, to allow new foreign capital to compensate for the outflow. Easing inflow controls has proven helpful in slowing exchange rate depreciation, e.g. in the case of India in 2013.17

  • Capital controls on outflows have rarely been tightened (Table 1) and are typically implemented as a last-resort tool in crisis circumstances. They appeared successful in Iceland in 2008 which was a typical crisis situation, but not apparently impacting outflows in Thailand in 1997 and Ukraine in 2008.18 In India in 2013, it appeared counterproductive, exacerbating depreciation pressures.19

Against this background, policy makers have relied on a wide range of policy tools to cope with the COVID-19 crisis and the associated financial shocks. Until now, governments have relaxed policies across the board20 and have made use of an array of fiscal and monetary policy tools. In particular, they have so far: eased monetary policy, conducted asset purchases, boosted USD liquidity through international swap lines, introduced fiscal stimulus, offered credit guarantees, relaxed prudential policies and engaged in regulatory forbearance. This section highlights selective measures that have a currency or residency dimension.21

In the context of fleeing foreign capital and rapidly depreciating currencies, countries tend to rely on FX reserves to defend their currencies and provide FX liquidity to domestic sectors. FX interventions intensified significantly since February 2020, reaching magnitudes comparable to those of the 2008 crisis, amid considerable turbulence and volatility in the FX market.

Chile’s central bank, for example, has expanded a FX intervention programme worth USD 20 billion, launched already in November 2019 and now extended until 2021. The Danish Central Bank has conducted FX interventions matching those that took place during the financial crisis in 2008. In light of the plunge in oil prices and the consequences of the pandemic, the Central Bank of Norway announced its readiness to intervene in currency markets for the first time in more than two decades to stabilise the NOK. The Swiss National Bank also intervened in the FX market, in the opposite direction, to limit further appreciation of the CHF.

Among EMEs, Brazil has conducted interventions in the spot currency and derivative markets, to support the BRL, which had depreciated by 15% since mid-February 2020. The magnitude of Brazil’s intervention amounted to USD 23 billion, corresponding to 6.4% of Brazil’s gross reserves, as of April 2020 (IMF, 2020). The Bank of Indonesia conducted a “triple intervention”, directly intervening to stabilize the depreciation of the IDR in the spot and domestic non-deliverable FX markets, and in the domestic government bond market. Colombia intervened with sales of USD through non-deliverable forwards (NDF): it sold on a monthly basis USD 1 billion in March, April and May, and it also sold USD 400 million of FX swaps in April 2020. Russia’s Central Bank used several facilities to sell FX under the fiscal rule: from 10 March to 28 May 2020 the aggregate FX sales based on the fiscal rule totalled USD 9.7 billion. Other countries like India, Mexico, and Peru have also recently intervened in the FX market, although full data are not yet available in all cases.

Overall, the capacity to use international reserves in times of crisis such as the COVID-19 crisis depends on the buffers built up over time, as well as the funding needs. Countries have in this respect not entered the crisis alike.

A few multilateral initiatives have been launched to prevent a dry-up of FX liquidity. On 15 March, the Fed established swap lines with four AE central banks – the Bank of Japan, the ECB, the Bank of England and the Swiss National Bank. And on 19 March, it also announced a temporary USD liquidity arrangements with the central banks of Australia, Brazil, Denmark, Korea, Mexico, Norway, New Zealand, Singapore, and Sweden, resuming a practice which was last used during the Global Financial Crisis. The Bank of Japan and the Bank of Thailand have also agreed on a bilateral swap agreement on 31 March, and the ECB agreed on swap lines with the Croatian National Bank and the Bulgarian National Bank on 15 and 22 April respectively.

In addition, on 31 March 2020, the Fed took the unprecedented step of establishing a new temporary repo facility for foreign and international monetary authorities, allowing them to enter into repurchase agreements with the Fed using US Treasury holdings as collateral. This would provide liquidity for central banks that have not been part of the swap agreements, but depends on the size of US Treasury holdings (Figure 6).

Beyond FX intervention, the policy mix has so far not relied too much on Capital Flow Management (CFMs) measures. Argentina has relaxed CFMs on outflows. Three EMEs, China, India and Peru, have, however, relaxed CFMs on inflows.

In Argentina, the central bank has added an option to allow withdrawals abroad in foreign currency using local bank accounts in local currency, subject to the limits for the purchase of foreign currency (USD 200, per month per person). The possibility of withdrawing foreign currency in cash through local credit or purchase cards, using overdrafts from the card issuer, was extended to USD 200 per transaction in non-bordering countries. Remittances to accounts abroad have been allowed from local accounts in foreign currency, up to USD 500 per month. The measures aim at helping Argentinian residents who are using foreign currency abroad, during the current emergency.

China has relaxed some controls on inflows by raising the macro-prudential adjustment coefficient of full-caliber cross-border financing from 1 to 1.25. China also removed restrictions on the investment quota of foreign institutional investors.

India raised the limit for foreign portfolio investors’ investment in corporate bonds to 15% of outstanding stock for 2020-21, a measure aiming at injecting liquidity into the Indian corporate bond market. Earlier in January, India had increased the investment cap of the Voluntary Retention Route (VRR) for investment in Indian debt by foreign portfolio investors, as well as increased limits on short-term investments by foreign portfolio investors. Secondly, India fully opened selected categories of government securities to non-resident investors, who were previously facing investment ceilings, in an additional move to support the bond market. Thirdly, the Reserve Bank of India extended the realisation period of export proceeds and repatriation, in order to allow domestic exporters to realise their receipts and repatriate funds within a longer time-frame, in view of their trading partners affected by the COVID-19 lockdown.

Finally, Peru sharply reduced its reserve requirements (encaje) for short-term FX liabilities to non-residents from 50% to 9%, to boost liquidity.

Policy makers have eased their stance on regulations targeting banks’ operations in foreign currency, falling into the category of Currency-Based Measures (CBMs).22 They have mainly relaxed CBMs that were already in place, and most of these measures are related to inflows, with few exceptions such as Turkey, which tightened some CBMs on outflows. The relaxation of measures intended to be used counter-cyclically has generally been welcomed. The most common instruments that countries have used include the following:

  • Differentiated reserve requirements: Countries that have reduced foreign-currency reserve requirements, or otherwise relaxed regulations related to FX reserve requirements include Indonesia, Turkey, and Peru. Indonesia reduced the FX reserve requirement ratio of all commercial banks from 8% to 4%, and lowered the local-currency denominated reserve requirement by 50 basis points for banks that finance import-export activities. Both measures aim at increasing forex liquidity and improve banks’ liquidity. Turkey also cut its FX reserve requirements by 500 basis points for banks, in all liability types and all maturity brackets for banks that meet real credit growth conditions within the context of the reserve requirement practice. Peru suspended the reserve requirements associated to credit in FX for the rest of 2020, and loosened the weekly limits and outstanding stocks floor in relation to reserve requirements for operations of selling of FX in exchange of domestic currency through derivatives (forwards and swaps). The latter measure was adopted to promote exchange risk coverage, in a context of higher financial volatility on the forex market. Also, the Chilean Central Bank has expanded on a temporary basis the eligible currencies for maintenance of the FX denominated reserve requirement, to include also EUR and JPY in addition to USD.

  • Differentiated liquidity ratios: Several countries have relaxed regulations on currency-differentiated liquidity ratios for banks. Sweden, for example, has temporarily allowed banks to fall below the required Liquidity Coverage Ratios (LCRs) for individual currencies and total currencies. Korea has also lowered its foreign-currency denominated LCR from 80% to 70% until May 2020. Hungary tightened from 15% to 10% the Foreign Exchange Coverage Ratio (FECR), which imposes a limit on the difference between FX denominated assets and liabilities of credit institutions as a percentage of total assets. The measure aims at further reducing the on-balance sheet currency mismatch to 10% of the balance sheet total, thereby reducing excessive reliance by Hungarian banks on the international swap market, in a context where the HUF reached record lows.

  • FX derivatives limits: Korea and Turkey took action in this area, although the former with an easing stance directed at capital inflows, and the latter with a tightening move toward capital outflows. Korea has relaxed the cap on FX forward positions for local banks, from the current 40% to 50% of their equity capital, and for foreign banks from the current 200% to 250%. The measure aims at boosting short-term debt and forward contracts denominated in FX. In a tightening move, Turkey cut the ceiling for FX swap, forward and option transactions local banks can conduct with foreign entities from 10% to 0.5% of the lender's equity. In practice, the measure constrains Turkish banks' ability to conduct operations in FX and lend TRY overseas to foreign entities. Turkey also cut the limit on TRY-denominated sell-side FX swaps, forwards and other derivatives with non-residents with a seven-day maturity to 1% of banks' equity; the new limit for these instruments with a 30-day maturity was set to 2%. Under these new rules, the volumes of TRY-denominated swaps Turkish banks can transact overseas would decrease.

  • Tax/levies on FX liabilities: Korea suspended its stability levy on financial institutions’ non-deposit FX liabilities from April to June 2020, and some instalments will be available for payments of the levy due in 2020. The relaxation of the measure is expected to increase financial institutions’ reliance on short-term FX funding.

  • Risk weightings for FX loans: Russia reduced to zero its risk weight add-ons for FX loans granted to pharmaceutical and medical supplies companies, in order to support the financing needs of these sectors, which may need to conduct operations in FX. In addition, it introduced a reduced risk ratio of 70% on bank’s RUB-denominated exposures to such companies.

  • Limits on FX operations of brokers and dealers: Brazil relaxed its limit on FX operations performed with clients by securities and stocks brokerage firms, securities and stocks dealer firms, and foreign exchange brokerage firms, from USD 100,000 to USD 300,000.

The COVID-19 outbreak has led to unprecedented capital outflows from EMEs, driven by sales of portfolio assets by foreign investors. The scale and speed of outflows in the current crisis have been about four times larger than during the 2008 financial crisis. As of Q2 2020 there are some signs of stabilisation, although the outlook remains fragile and global uncertainty continues to weigh on global investor confidence.

In the face of global dollar liquidity shortages, some EMEs central banks intervened in the FX market to support depreciating currencies, and several central banks have established or expanded swap lines. Countries have so far not seen a need to resort to capital controls on outflows, with responses in the area of capital flows largely focused on relaxing rules on inflows, easing liquidity, and increasing access to foreign funding. Currency-based measures have become an important part of EME’s policy toolkits, with a variety of tools used, including differentiated reserve requirements and liquidity ratios, and FX derivative limits.

It may be too early to assess the effectiveness of such easing actions, although the impact of such measures in previous crisis could provide a first indication. More comprehensive assessments should be conducted once the full nature of the current crisis becomes clearer, taking into account each country’s situation, and including a rigorous cost/benefit analysis of each measure, weighing the positive effects on domestic variables against the risks of externalities such as distortions, regulatory arbitrage, and fragmentation.

Countries should continue to closely monitor developments and risks, as the intensity of the pandemic varies across countries and regions, and as the impact of the crisis on the economies becomes clearer during the second half of 2020. The high level of non-bank FX debt, in light of expected downgrades in the near future, may need particular attention as a potential source of vulnerability in the coming months. In the longer term, local capital market development, including the development of a domestic investor base, may help increase resilience to outflows.

International co-operation will continue to be key. As some countries will continue to review and adjust their policy measures, they may learn from the successes (or failures) of their peers in using particular measures in different situations. International Organisations, and specialised groups such as the Advisory Task Force on the Codes (ATFC), could be further leveraged as a platform to exchange experience and conduct more in-depth analysis on the effectiveness of measures.

In today’s global financial markets, co-operation may also be helpful to avoid, as much as possible, negative spill-overs of one country’s measures on other countries. As unintended market fragmentation could make it more difficult for businesses to raise funding, countries have a strong interest in cooperating to find the most appropriate tools to address the crisis, without compromising the prospects for recovery.


← 1. See OECD (2020), OECD Economic Outlook, June 2020. Scenario where the outbreak is contained.

← 2. See WTO (2020), WTO Trade forecast, 8 April 2020.

← 5. These developments were also highlighted in Investment policy responses to COVID.

← 6. The significant appreciation of the USD, particularly against EM currencies, appears to fit historical patterns of crises - see for instance Corsetti and Marin (2020), The dollar and international capital flows in the COVID-19 crisis, VoxEu.

← 8. Such as lower-rated credit issued in the form of BBB bonds, non-investment grade bonds, and leveraged loans: see Global Financial Markets Policy Responses to COVID-19.

← 9. The OECD highlighted in 2019 that corporations in both AEs and EMEs were facing record levels of repayment requirements, with AE and EME companies needing to pay or refinance USD 2.9 tr and 1.3 tr, respectively, within 3 years, and warned that an economic downturn may increase the rate of downgrades in the BBB rated corporate bond segment. See Çelik, S., G. Demirtaş and M. Isaksson (2019), “Corporate Bond Markets in a Time of Unconventional Monetary Policy”, OECD Capital Market Series.

← 10. Çelik, S., G. Demirtaş and M. Isaksson (2019), above cited.

← 11. See for instance Hofmann, B., I. Shim and H. Shin (2020), Emerging market economy exchange rates and local currency bond markets amid the Covid-19 pandemic, BIS Bulletin No 5 and Hofmann, B., I. Shim and H. Shin (2019), “Bond risk premia and the exchange rate”, BIS Working Papers

← 12. See McCauley, R., P. Mcguire and V. Sushko (2015), “Dollar credit to emerging market economies 1”, BIS Quarterly Review and Aldasoro, I. and T. Ehlers (2018), “Global liquidity: changing instrument and currency patterns”, BIS Quarterly Review for further analysis of FX debt across sectors and countries.

← 13. Foreign ownership is estimated at around 1 to 1.6% of the total value of outstanding bonds, and the share of FX-denominated bonds represents only 6.4% of Chinese non-financial companies’ issuance. See Cerutti E. and M. Obstfeld (2018), “China’s Bond Market and Global Financial Markets”, IMF Working Paper; IMF (2019), The Future of China’s Bond Market; and Celik et al (2019), Corporate Bond Markets In A Time Of Unconventional Monetary Policy for data.

← 14. See the seminal paper by Forbes, K. and F. Warnock (2012), “Capital flow waves: Surges, stops, flight, and retrenchment”, Journal of International Economics, Vol. 88/2, pp. 235-251 and Forbes, K. and Warnock (2020), Capital Flow Waves-or Ripples? Extreme Capital Flow Movements Since the Crisis for an updated analysis post GFC.

← 15. See Lepers and Mehigan (2019), “The broad policy toolkit for financial stability: Foundations, Fences, and Firedoors”, OECD Working Papers on International Investment, 2019/02.

← 16. See Basu, S. et al. (2018), “Managing Capital Outflows with Limited Reserves”, IMF Economic Review, Vol. 66/2, pp. 333-374 for recent work modelling the conditions for the effectiveness of FX intervention.

← 17. See Sahay, R. et al. (2014), Emerging market volatility: Lessons from the Taper Tantrum for a general analysis of policy responses of EM during the 2013 Taper Tantrum.

← 18. See Saborowski, C. et al. (2014), “Effectiveness of Capital Outflow Restrictions”, IMF Working Papers for an empirical and historical review of capital controls on outflows.

← 19. See Basu, K., B. Eichengreen and P. Gupta (2014), From Tapering to Tightening - The Impact of the Fed’s Exit on India for an analysis of India policy reaction during the 2013 taper tantrum.

← 21. some of which have recently been tightened in a number of countries.

← 22. See De Crescenzio et al. (2015, 2017) for a detailed description of CBMs, defined as regulations discriminating on the basis of the currency of an operation – in other words, measures that apply a less favorable treatment to operations by financial institutions in a particular currency, typically foreign currencies.


This paper is published under the responsibility of the Secretary-General of the OECD. The opinions expressed and the arguments employed herein do not necessarily reflect the official views of OECD member countries.

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area.

© OECD 2020

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at